AI-Driven Predictive Control for Dynamic Energy Optimization in Flying Cars

Abstract

This study presents an AI-driven energy management system (EMS) for a hybrid electric flying car, integrating multiple power sources—including solid-state batteries, Li-ion batteries, fuel cells, solar panels, and wind turbines—to optimize power distribution across various flight phases. The proposed EMS dynamically adjusts power allocation during takeoff, cruise, landing, and ground operations, ensuring optimal energy utilization while minimizing losses. A MATLAB-based simulation framework is developed to evaluate key performance metrics, including power demand, state of charge (SOC), system efficiency, and energy recovery through regenerative braking. The findings show that by optimizing renewable energy collecting, minimizing battery depletion, and dynamically controlling power sources, AI-based predictive control dramatically improves energy efficiency. While carbon footprint assessment emphasizes the environmental advantages of using renewable energy sources, SOC analysis demonstrates that regenerative braking prolongs battery life and lowers overall energy use. AI-optimized energy distribution also lowers overall operating costs while increasing reliability, according to life-cycle cost assessment (LCA), which assesses the economic sustainability of important components. Sensitivity analysis under sensor noise and environmental disturbances further validates system robustness, demonstrating that efficiency remains above 84% even under adverse conditions. These findings suggest that AI-enhanced hybrid propulsion can significantly improve the sustainability, economic feasibility, and real-world performance of future flying car systems, paving the way for intelligent, low-emission aerial transportation.

1. Introduction

The evolution of urban air mobility (UAM) and electric vertical takeoff and landing (eVTOL) vehicles necessitates cutting-edge advancements in power management. The primary challenge lies in efficiently distributing power among propulsion, avionics, and auxiliary systems while optimizing battery life and ensuring system reliability. A significant area of development is the integration of artificial intelligence (AI) for adaptive power management. By leveraging machine learning algorithms, real-time energy optimization, and hybrid energy storage systems (HESSs), power distribution can be more dynamic, ultimately enhancing the efficiency, safety, and sustainability of flying cars [1,2,3,4].

Current studies on power management in eVTOLs primarily focus on static energy distribution systems, which lack the adaptability needed to address real-time flight conditions. Traditional power allocation methods fail to optimize energy usage during varying flight phases, such as takeoff, cruising, and landing. Furthermore, while research on HESS has demonstrated substantial improvements in efficiency, these systems typically lack AI-driven optimization techniques to dynamically adjust power allocation. The integration of wide-bandgap semiconductors has been proven to reduce power losses in electronic components; however, their application in conjunction with machine learning for active power optimization in flying cars remains underexplored [5,6]. This research seeks to bridge these gaps by combining AI, hybrid energy storage, and advanced power electronics.

For larger vehicles, such as the U.S. Defense Advanced Research Projects Agency (DARPA) Transformer vehicle, dedicated propulsion systems are used for lift, flight propulsion, and ground propulsion. However, in smaller vehicles, employing separate electric or hybrid systems increases the vehicle’s weight, volume, and cost, presenting significant challenges in packaging and system integration. Moreover, using a single engine or motor for both vertical lift and horizontal cruising propulsion can lead to inefficiencies. A larger, heavier engine optimized for lift may not perform efficiently during cruising, resulting in excessive energy consumption and reduced range. Therefore, depending on the vehicle’s performance and range requirements, separate engines may be necessary for cruising and VTOL operations [7,8,9,10].

A simplified propulsion architecture for a flying car is implemented such that the engine supplies power for both ground and flight propulsion but does not contribute to VTOL capability. During ground propulsion, a clutch disengages the propeller, allowing the vehicle to operate like a conventional automobile. During takeoff and flight, the engine exclusively drives the propeller. A more advanced hybrid automotive propulsion system, designed for flying cars, integrates VTOL capability. The system uses an auxiliary power unit (APU) with an engine-driven generator and power converter to charge a battery, which powers the vehicle’s wheels for ground propulsion. For VTOL and flight propulsion, a turbine engine or internal combustion engine is employed. The engine drives the vertical takeoff fan during VTOL, transitioning to propeller-driven horizontal flight once cruising altitude is reached [11,12,13,14,15].

The concept of using AI-driven predictive controllers for dynamic energy management and optimization in aerospace systems is an emerging and promising area of research. As electric and hybrid electric propulsion systems evolve, the need for advanced energy management systems (EMSs) becomes more pressing. AI-based methods, particularly those utilizing neural networks (NNs), offer substantial improvements in predicting power demand and optimizing energy distribution across various sources, such as batteries, fuel cells, and renewable energy sources [1]. Neural networks excel in learning complex patterns in historical data, allowing the system to predict future energy demands with high accuracy. In aerospace applications, this ability enables real-time adjustments to power allocation, improving the overall efficiency of energy use [16,17,18,19,20].

The integration of renewable energy sources like solar power further complicates energy management, as these sources are intermittent and weather-dependent [21,22,23,24,25]. Predictive controllers play a crucial role in mitigating uncertainties by dynamically adjusting power distribution in response to fluctuating energy availability. Previous studies, such as those by [1], demonstrate the effectiveness of AI in optimizing energy allocation, particularly in aircraft power systems. Similarly, [5] proposed a predictive optimization algorithm that balances battery and fuel cell power in electric aircraft, achieving significant improvements in fuel efficiency.

Thermal stress management is another critical area where AI-based predictive controllers can offer benefits. Traditional energy management systems typically rely on instantaneous current thresholds to protect components, which may not account for long-term thermal stress accumulation. In contrast, AI-based methods can predict thermal degradation over time and adjust system operations to prevent overheating and ensure component longevity [7,15,21]. This is particularly important in aerospace systems, where thermal stress can significantly impact the performance and reliability of critical components, such as power distribution units, converters, and batteries.

AI-driven controllers also enhance fault detection capabilities, which are essential for ensuring the safe operation of aerospace systems. Machine learning algorithms can detect potential faults in power generation, distribution, or storage components early on, preventing catastrophic failures. The ability to detect and address issues proactively is especially important in aerospace systems, where failures can have severe consequences. Furthermore, the integration of adaptive control techniques enables the system to adjust to varying load conditions, such as higher power demands during takeoff or landing and more stable demands during cruise flight [22,23,24,25].

Despite the promising potential of AI-driven predictive controllers in aerospace systems, challenges remain in their implementation. The complexity of integrating AI with existing aerospace systems, which must meet stringent safety and reliability standards, is a major hurdle. Moreover, the computational resources required for real-time AI decision-making could be a limiting factor, especially in smaller aerospace platforms. Additionally, the need for large, high-quality datasets to train AI models poses another challenge. Given the high costs and risks associated with collecting flight data, obtaining the necessary training data can be difficult. However, advancements in simulation technologies and the increasing availability of real-time monitoring systems offer potential solutions to these challenges [7,26,27,28,29,30].

Looking ahead, future research could focus on enhancing the explainability and transparency of AI models to ensure that their decisions can be understood and trusted by operators. Additionally, hybrid approaches that combine AI with traditional control methods could provide a balanced solution, ensuring reliability while leveraging the flexibility of AI [31,32,33,34,35]. The sources under evaluation cover a wide range of energy management and optimization topics in many transportation domains, with a focus on airborne and maritime vehicles. With a special emphasis on reducing underwater radiated noise (URN), Ref. [36] addresses the coordinated operation of a multi-energy ship microgrid that integrates diesel generators, batteries, photovoltaics, and combined cooling, heat, and power (CCHP) units. In contrast, Ref. [37] highlights interpretability in energy forecasting by presenting an explainable multi-task learning framework for EV charging demand prediction. On the other hand, we focus on dynamic environmental models for aerial vehicles, which deal with energy variations brought on by environmental disruptions.

Additional sources explore several energy management techniques for electric and hybrid flying vehicles [38,39,40,41,42,43,44,45]. A deep reinforcement learning-based EMS for a hybrid flying automobile which improves exploration efficiency and resolves engine on/off problems is proposed in Ref. [38]. Power battery performance under VTOL profiles, lifecycle emissions, and energy costs for various propulsion systems are examined in Refs. [39,40,41]. While Ref. [43] evaluates the cost and economics of electric flying cars, Ref. [42] offers a more comprehensive viewpoint on sustainable transportation technology. Last but not least, Refs. [44,45] stress safety and cost factors while concentrating on predicting EV charging demand and the best battery charging techniques. While all of these studies contribute to our study, our work stands out due to its consideration of particular airborne energy difficulties and its emphasis on dynamic power demand calculations in aerial contexts.

Although the cited works offer insightful viewpoints on energy management in various settings, our study stands out because it focuses on dynamic environmental simulations for aerial vehicles. We discuss the particular difficulties that come with airborne energy systems, taking into account things like changing power requirements during different flight stages and how environmental disruptions affect renewable energy sources like wind and sun. This emphasis enables the creation of customized plans to improve the dependability and efficiency of aerial energy systems.

Figure 1 visually represents an AI-driven energy management system for a hybrid electric flying car, illustrating the key components and their interactions across different flight phases. It integrates multiple power sources, including batteries, fuel cells, solar panels, and wind turbines, to optimize power distribution across different flight phases (takeoff, cruise, landing, and ground operation). The AI-based control system dynamically adjusts power allocation to enhance efficiency, extend battery life, and reduce emissions. Arrows indicate energy flow between sources, propulsion systems, and regenerative braking mechanisms.

Figure 2 shows the key factors affecting the efficiency of a hybrid electric flying car. It categorizes these factors into mechanical (aerodynamics, weight, drag), electrical (motor and battery efficiency, regenerative braking), and environmental (solar and wind energy availability, weather conditions). The interaction between these parameters determines overall energy consumption, flight range, and operational sustainability. The figure highlights how AI-driven optimization can mitigate losses and enhance performance.

The primary objectives of this research are to advance energy management and optimization for flying cars. First, we aim to develop an AI-powered energy management system (EMS) that can dynamically allocate power based on flight conditions, battery health, and system demands, ensuring optimal energy usage throughout the vehicle’s operation. Second, the research focuses on designing an adaptive hybrid energy storage system (HESS) that integrates solid-state batteries (SSBs) and supercapacitors, providing peak power delivery for demanding flight phases while extending the overall range of the vehicle. Lastly, the research aims to enhance predictive maintenance capabilities through AI-driven fault detection and prevention mechanisms, which will reduce vehicle downtime and increase the overall reliability of flying cars.

2. Electrical System of Flying Car

The flying car’s electrical propulsion system is designed to efficiently balance energy sources for both aerial and ground travel. It primarily relies on electric motors for propulsion, powered by a combination of batteries, fuel cells, and renewable energy sources such as solar and wind power. During takeoff and flight, the high power demand is met through battery and fuel cell energy, while in cruise mode, supplementary power from solar panels and wind turbines helps reduce the load on the main power sources. The system incorporates regenerative braking during landing, capturing kinetic energy to recharge the battery, enhancing overall efficiency and extending operational range.

In a hybrid propulsion configuration, electric motors are utilized for both flight and ground propulsion, while a dedicated engine is used exclusively for generating lift during vertical takeoff and landing (VTOL). This hybrid approach optimizes power distribution, allowing efficient energy use during different operational phases. On the ground, the flying car can operate in a fully electric mode, using battery-stored energy to reduce emissions and operating costs. During flight, the engine provides lift, while electric motors handle propulsion, ensuring a smooth transition between aerial and ground operations with improved energy efficiency and sustainability. Figure 3 illustrates a hybrid propulsion system where an electric motor is dedicated to ground propulsion while a conventional engine or alternative power source handles airborne propulsion. The system enables efficient taxiing, takeoff, and landing operations by minimizing fuel consumption and emissions during ground movement. The figure highlights the role of an auxiliary power unit (APU) or battery pack in supplying energy to the electric motor, ensuring optimal energy usage and improved sustainability. Figure 4 represents a propulsion architecture where electric motors are responsible for both flight and ground propulsion, while a dedicated engine provides lift assistance (e.g., in a VTOL or hybrid flying vehicle).

The system enhances efficiency by allowing for electric propulsion for horizontal movement while using an engine selectively for vertical takeoff and landing. The figure emphasizes the integration of battery storage, regenerative braking, and AI-driven power management to optimize performance and energy efficiency.

3. AI-Optimized Energy Management

Our study proposes an AI-optimized energy management system for flying cars which uses an AI-based predictive controller to efficiently manage energy sources (battery, fuel cell, solar, wind turbine) based on the power demand during various flight phases (takeoff, cruise, landing, and ground operation). Additionally, regenerative braking is used during landing to recover energy and recharge the battery.

The goal of the system is to optimize energy usage, reduce dependency on non-renewable sources, and improve efficiency, particularly by utilizing energy recovery through regenerative braking. The power demand during the flight phases and the available energy sources are dynamically adjusted through AI-based predictions. The energy management of a flying car operates through four key phases. The takeoff phase demands the highest power, primarily supplied by the fuel cell and battery to provide the necessary thrust. During the cruise phase, power demand is moderate, with solar panels and wind turbines contributing to energy generation, while the battery or fuel cell supplements as needed. In the landing phase, power demand decreases, and if regenerative braking is implemented, a portion of the energy can be recovered to recharge the battery. Finally, the ground phase requires minimal power, which is typically supplied by the battery for essential operations such as taxiing and onboard systems.

The SOC represents the percentage of the battery’s total capacity that is available at any given time. It is updated by tracking the energy usage and recovery over time.

SOC Calculation with Regenerative Braking: The battery’s SOC at any time t is updated by subtracting the energy used by the battery and adding any recovered energy from regenerative braking.

$$SOC\left(t\right)=SOC\left(t-1\right)-{\displaystyle \frac{E_{used}}{E_{battery}}}$$

(1)

where E_used is the energy used by the battery (kWh). E_battery is the maximum battery energy (kWh). η_battery is the efficiency of the battery (95% in the code). Energy Used by the Battery: The energy used by the battery at each time stepis computed as

$$E_{used}={\displaystyle \frac{P_{battery}\left(t\right)·\mathsf{\Delta }t}{{3600·\eta }_{battery}}}$$

(2)

Energy Supplied: The total energy supplied by all sources, including the battery, fuel cell, solar, wind, and regenerative braking (if applicable), is

$$\begin{array}{cc}{E}_{total\_supplied}=& {\int }_0^t(P_{battery}\left(t\right)+P_{fuelcell}\left(t\right)+P_{solar}\left(t\right)+P_{wind}\left(t\right)\hfill \\ & +P_{regen}\left(t\right))dt\hfill \end{array}$$

(3)

Power Demand: The remaining power demand at time t after accounting for solar and wind power generation is

$$E_{needed}\left(\mathrm{t}\right)=P_{demand}\left(t\right)-(P_{solar}\left(t\right)+P_{wind}\left(t\right))$$

(4)

If P_needed(t) > 150 kW, the fuel cell contributes up to its maximum power of 150 kW, and the remaining demand is met by the battery. Otherwise, the battery provides the full remaining power. During the landing phase, regenerative braking recovers a portion of the power demanded by the vehicle. In the code, 40% of the power required during landing is recovered as regenerative braking power:

$$P_{regen}\left(t\right)=-0.4·P_{demand}\left(t\right)\left(duringlandingphase\right)$$

(5)

where P_demand(t) is the power demand at time t (kW), determined by the flight phase. P_battery(t) is the power supplied by the battery at time t (kW). P_fuelcell(t) is the power provided by the fuel cell at time t (kW). P_solar(t) is the power generated by the solar panels at time t (kW). P_wind(t) is the power generated by the wind turbine at time t (kW). P_regen(t) is the power recovered through regenerative braking during the landing phase at time t (kW).

The negative sign indicates energy being returned to the battery. SOC with regenerative braking can be calculated as follows:

$$SOC\left(t\right)=SOC\left(t-1\right)-{\displaystyle \frac{P_{battery}\left(t\right)·\mathsf{\Delta }t}{{3600·\eta }_{battery}}}+{\displaystyle \frac{P_{regen}\left(t\right)·\mathsf{\Delta }t}{{3600·\eta }_{battery}}}$$

(6)

SOC without regenerative braking can be calculated as follows:

$$SOC\left(t\right)=SOC\left(t-1\right)-{\displaystyle \frac{P_{battery}\left(t\right)·\mathsf{\Delta }t}{{3600·\eta }_{battery}}}$$

(7)

Here, P_regen(t) = 0.

The overall system efficiency is defined as the ratio of total energy used to the total energy supplied, expressed as a percentage:

$$\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}={\displaystyle \frac{E_{total\_used}}{E_{total\_supplied}}}\times 100$$

(8)

where total energy used:

$$E_{total\_used}={\int }_0^t{P}_{demand}\left(t\right)dt$$

(9)

E_{total_supplied} is total energy supplied includes all energy sources and recovered energy with or without regenerative braking. By comparing the efficiency values with and without regenerative braking, you can assess the contribution of regenerative braking to the overall system efficiency.

4. Neural Network Architecture

In the proposed AI-driven energy management system for hybrid electric flying cars, the neural network architecture is designed to handle dynamic power allocation across multiple energy sources while adapting to varying flight phases and environmental conditions. A simple feedforward artificial neural network (ANN) is developed for power demand prediction, consisting of two hidden layers with 10 and 5 neurons. The network is trained using the Levenberg–Marquardt algorithm with mean squared error (MSE) as the loss function. Training data are split into 70% for training, 15% for validation, and 15% for testing, with loss monitored and plotted over epochs to assess performance. The trained ANN predicts power demand, which is integrated into an energy management strategy for calculating the state of charge (SOC) of a battery, both with and without regenerative braking. The model’s effectiveness is evaluated through predicted vs. actual power demand plots and SOC analysis. The ANN likely takes historical power demand and operational parameters as inputs, using activation functions like ReLU or tanh. The prediction performance is refined through batch size tuning and epoch selection, ensuring accurate power demand forecasting for optimized energy management.

The architecture is structured as follows:

• Input Layer: The input layer consists of eight neurons representing key system parameters, including power demand, state of charge (SOC) of the battery, available solar and wind power, flight phase indicators (takeoff, cruise, landing, ground), and environmental factors like wind speed and solar intensity.
• Hidden Layers: The network includes three fully connected hidden layers:

  ○

  First Hidden Layer: 64 neurons with the ReLU (Rectified Linear Unit) activation function to capture complex nonlinear relationships.

  ○

  Second Hidden Layer: 32 neurons with ReLU activation, reducing dimensionality while retaining essential features.

  ○

  Third Hidden Layer: 16 neurons with ReLU activation, further refining the feature set and preventing overfitting.

• Output Layer: The output layer consists of four neurons corresponding to the power allocation decisions for the battery, fuel cell, solar power, and wind turbine. A softmax activation function is used to ensure that the total allocated power remains within the system’s demand constraints.
• Optimization Technique: The network is trained using the Adam optimizer, known for its efficiency in handling sparse gradients and adaptive learning rates. The initial learning rate is set to 0.001, with a batch size of 32 and a maximum of 100 epochs.

Figure 5 shows the architecture of the ANN-based AI-driven management system for flying cars.

The loss function is defined as a mean squared error (MSE) loss function is used to minimize the discrepancy between predicted and actual power allocation, ensuring accurate system performance.

• Training and Validation: The model is trained on a dataset generated through extensive flight simulations, including diverse environmental conditions and flight scenarios. The data are split into 80% for training and 20% for validation, and k-fold cross-validation is applied to enhance generalizability.

This architecture balances computational efficiency and prediction accuracy, making it well suited for real-time deployment on resource-constrained flight platforms.

Let us formally define the neural network architecture and its training process with equations:

1.

Input Representation:

The input vector x ∈ R8 consists of:

$$x={[P}_{demand},SOC,P_{solar},P_{wind},Takeoff,Cruse,Landing,Ground]$$

(10)

where P_demand is the power demand, SOC is the state of charge, and the flight phase indicators are binary variables.

2.

First Hidden Layer:

Let W₁ ∈ R64 × 8 be the weight matrix and b₁ ∈ R64 be the bias vector. The output of the first hidden layer is

$$h_1=ReLU(W_{1}x+b_1)$$

(11)

where ReLU is the activation function:

$$ReLU\left(z\right)=\mathrm{m}\mathrm{a}\mathrm{x}(0,z)$$

(12)

3.

Second Hidden Layer:

With weight matrix W₂ ∈ R32 × 64 and bias b₂ ∈ R32,

$$h_2=ReLU(W_2{h}_1+b_2)$$

(13)

4.

Third Hidden Layer:

With weight matrix W₃ ∈ R16 × 32 and bias b₃ ∈ R16,

$$h_3=ReLU(W_3{h}_2+b_3)$$

(14)

5.

Output Layer:

The output layer produces power allocation decisions y ∈ R4 for the battery, fuel cell, solar, and wind power:

$$y=Softmax(W_4{h}_3+b_4)$$

(15)

where

$$Softmax\left(y_i\right)={\displaystyle \frac{e^{yi}}{\sum _{j=1}^4{e}^{yi}}}$$

(16)

6.

Loss Function:

We use the mean squared error (MSE) to minimize the difference between predicted power allocation y and the actual allocation ŷ:

$$MSE={\displaystyle \frac{1}{N}}{\sum }_{i=1}^N{‖{\dot{\widehat{y}}}_i-y_i‖}^2$$

(17)

7.

Optimization:

The Adam optimizer updates weights and biases iteratively to minimize the loss:

$${\theta }_{t+1}={\theta }_t-\eta ·{\displaystyle \frac{m_t}{\sqrt{v_t}+ϵ}}$$

(18)

where η is the learning rate, m_t and v_t are estimates of the first and second moments of the gradients, and ϵ is a small constant to avoid division by zero.

This structure enables efficient, real-time energy management by balancing predictive accuracy and computational efficiency.

5. Energy Management Metrices

In energy management systems (EMSs), particularly for systems like AI-optimized power management in flying cars, various metrics are critical to assessing the system’s performance, efficiency, and overall effectiveness. Below are the key metrics you can calculate to evaluate the energy management system.

System efficiency indicates how efficiently the system utilizes its energy sources to meet the demand. The system efficiency equation is given by

$$\mathrm{S}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}={\displaystyle \frac{E_{total\_used}}{E_{total\_supplied}}}\times 100$$

(19)

A higher efficiency means that the energy supplied by various sources (battery, fuel cell, solar, wind, regenerative braking) is used more effectively to meet the power demand.

System efficiency in an energy management framework typically accounts for the efficiencies of individual components like the battery, fuel cell, regenerative braking system, and power converters. The overall efficiency should be the product of these individual efficiencies when power flows through multiple stages. I will revisit the formula used in this work, cross-check it with the power flow and energy calculations, and update it accordingly to reflect a realistic and consistent system efficiency model.

Let us carefully revisit and revise the system efficiency formula. A fixed battery efficiency of 95% was assumed, but real-world system efficiency involves multiple factors, like the following:

• Battery Efficiency (η_b): This varies with temperature, current load, and state of charge.
• Fuel Cell Efficiency (η_fc): This depends on hydrogen consumption and electrical output.
• Solar and Wind Power Efficiency (η_s, η_w): This is typically assumed constant but can be influenced by environmental factors.
• Regenerative Braking Efficiency (η_r): This is the efficiency of capturing and storing braking energy.
• Power Converter Efficiency (η_c): This is the efficiency of DC-DC or DC-AC converters in the system.

Revised System Efficiency Formula:

For the total power delivered to the load P_load, the system efficiency can be expressed as

$${\eta }_{system}={\displaystyle \frac{P_{load}}{P_{input}}}={\eta }_b·{\eta }_{fc}·{\eta }_s·{\eta }_w·{\eta }_r·{\eta }_c$$

(20)

Energy consumption is the total energy consumed by the flying car over the time period of operation. The energy consumption equation can be given by

$$\mathrm{Energy}\mathrm{Consumption}={\int }_0^{t}P\left(t\right)dt$$

(21)

where P(t) is the power demand over time. This metric shows the total amount of energy used during the flight and helps in assessing the system’s energy sustainability. The battery state of charge (SOC) metric tracks the charge level of the battery, which is important for ensuring that the battery operates within its optimal range.

$$SOC\left(t\right)=SOC\left(t-1\right)-{\displaystyle \frac{E_{used}\left(t\right)}{E_{batterymax}}}$$

(22)

where E_used is the energy consumed from the battery, and E_battery max is the battery’s maximum energy capacity. SOC reflects how much energy remains in the battery at any given time. A lower SOC indicates more energy depletion, and maintaining SOC within an optimal range is crucial for system reliability.

Fuel cell utilization is the total energy supplied by the fuel cell during the operation of the flying car.

$$E_{fuelcell}\left(t\right)=\underset{0}{\overset{T}{\int }}{P}_{fuelcell}\left(t\right)dt$$

(23)

This metric helps understand how much of the power demand is being met by the fuel cell, and if fuel cell usage is optimized, it can reduce the dependency on batteries.

Regenerative braking contribution is the total energy recaptured through regenerative braking during landing or deceleration phases. The regenerative energy equation is given by

$$\mathrm{R}\mathrm{e}\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}=\underset{0}{\overset{T}{\int }}{P}_{regen}\left(t\right)dt$$

(24)

The higher the regenerative braking energy, the more energy is recovered and reused, contributing to improved system efficiency and extending battery life.

The solar energy contribution is the total energy generated and utilized by the solar panels. Solar energy can be expressed as

$$\mathrm{S}\mathrm{o}\mathrm{l}\mathrm{a}\mathrm{r}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}=\underset{0}{\overset{T}{\int }}{P}_{solar}\left(t\right)dt$$

(25)

This metric indicates the effectiveness of the solar panels in providing supplemental energy and reducing the overall demand from other sources.

The wind turbine contribution is defined as the total energy provided by the wind turbine (in cruise mode).

$$\mathrm{W}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}=\underset{0}{\overset{T}{\int }}{P}_{wind}\left(t\right)dt$$

(26)

The contribution of wind energy is important in reducing the reliance on fuel cells and batteries, especially during cruise mode. Energy Recovery Efficiency (regeneration) is defined as the efficiency of the regenerative braking system in converting kinetic energy back into electrical energy. The Energy Recovery Efficiency equation can be expressed as

$$\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{R}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}={\displaystyle \frac{\mathrm{R}\mathrm{e}\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{d}}{\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s}\mathrm{d}\mathrm{u}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{d}\mathrm{e}\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}}\times 100$$

(27)

A higher recovery efficiency indicates that more of the energy that would have been lost during deceleration is being converted back to useful energy.

The peak power demand is defined as the maximum power demand encountered during the flight, often occurring during takeoff or rapid changes in flight mode. Peak power demand can be expressed as

$$P_{peak}=\mathrm{m}\mathrm{a}\mathrm{x}\left(P_{demand}\right(\mathrm{t}\left)\right)$$

(28)

Therefore, understanding peak power demand helps in sizing and optimizing the power generation and storage systems (e.g., fuel cell or battery).

Energy Storage Efficiency is defined as the efficiency with which the energy storage system (battery) stores and releases energy. It can be expressed in the following equation:

$$\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{S}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}={\displaystyle \frac{\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{D}\mathrm{e}\mathrm{l}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{d}\mathrm{t}\mathrm{o}\mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}{\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{d}\mathrm{r}\mathrm{a}\mathrm{w}\mathrm{n}\mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}}\times 100$$

(29)

A high Energy Storage Efficiency indicates that the battery is effectively storing and discharging energy with minimal losses.

System availability/up-time is the percentage of time that the system is available for use without failure or major issues (e.g., battery depletion). Therefore,

$$\mathrm{A}\mathrm{v}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{a}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}={\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{O}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{T}\mathrm{i}\mathrm{m}\mathrm{e}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{T}\mathrm{i}\mathrm{m}\mathrm{e}}}\times 100$$

(30)

A higher availability means the system is operating effectively with minimal downtime due to energy constraints.

Cost efficiency is defined the cost per unit of energy provided by the system, considering energy sources (e.g., battery replacement, fuel cell usage, solar panel maintenance). Therefore, the cost efficiency equation is given by

$$\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}={\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{P}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{S}\mathrm{u}\mathrm{p}\mathrm{p}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{d}}}\times 100$$

(31)

Cost-effective systems minimize the cost per kWh of energy produced, which can impact the overall economic viability of the flying car. The cost efficiency is estimated by evaluating the total operational cost savings achieved through the AI-driven energy management system (EMS) compared to a baseline system without optimization. This value is calculated by considering multiple factors, including the cost per kWh of energy from different power sources (solid-state batteries, Li-ion batteries, fuel cells, solar, and wind), the total energy supplied by each source, and the reduction in energy waste achieved by the AI’s predictive power distribution.

The formula used for cost efficiency is

$$\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{E}\mathrm{f}\mathrm{f}\mathrm{i}\mathrm{c}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{y}\mathrm{\%}=\left({\displaystyle \frac{\mathrm{B}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}-\mathrm{O}\mathrm{p}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{i}\mathrm{z}\mathrm{e}\mathrm{d}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}}{\mathrm{B}\mathrm{a}\mathrm{s}\mathrm{e}\mathrm{l}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}}}\right)\times 100$$

(32)

where Baseline Cost is the total operational cost of a non-optimized system where power sources are used without AI-based load prediction and power management. Optimized Cost is the total operational cost of the AI-driven EMS, where power usage is optimized based on real-time demand prediction and efficient energy distribution.

The cost per kWh for each power source was factored into the calculation, along with the total energy supplied by each source over the flight duration. The AI-driven EMS minimizes the reliance on expensive, high-emission sources like fuel cells and prioritizes renewable and efficient sources like solar, wind, and solid-state batteries. This results in significant cost savings and improved sustainability.

The carbon footprint is the total CO₂ emissions associated with energy production and consumption, particularly from fuel cells or any non-renewable sources.

The carbon footprint equation is given by

$$\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t}=\sum (\mathrm{E}\mathrm{m}\mathrm{m}\mathrm{i}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{o}\mathrm{f}\mathrm{S}\mathrm{o}\mathrm{u}\mathrm{r}\mathrm{c}\mathrm{e}\times \mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m}\mathrm{S}\mathrm{o}\mathrm{u}\mathrm{r}\mathrm{c}\mathrm{e})$$

(33)

A lower carbon footprint indicates a greener and more sustainable energy management approach, especially when using renewable energy sources like solar and wind.

The Power Source Contribution Ratio is the proportion of the total power demand met by each energy source (battery, fuel cell, solar, wind, regenerative braking). The Power Source Contribution Ratio is given by

$$\mathrm{P}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}\mathrm{S}\mathrm{o}\mathrm{u}\mathrm{r}\mathrm{c}\mathrm{e}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{i}\mathrm{b}\mathrm{u}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{R}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}={\displaystyle \frac{P_{source}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{P}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}\mathrm{D}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}}}\times 100$$

(34)

This metric helps assess the balance between different power sources and can indicate areas for optimization (e.g., increasing solar contribution during sunny days).

Charging/Discharging Cycles are the number of charge and discharge cycles the battery undergoes during the flight. Charging Cycles can be expressed as

$$\mathrm{C}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{g}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{C}\mathrm{y}\mathrm{c}\mathrm{l}\mathrm{e}\mathrm{s}={\displaystyle \frac{E_{batterytotal}}{\mathrm{B}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{y}\mathrm{c}\mathrm{a}\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{i}\mathrm{t}\mathrm{y}}}$$

(35)

More cycles can lead to battery wear and reduced lifespan. Minimizing the number of cycles through efficient power management can improve the system’s long-term performance.

5.1. Technology Comparison

To optimize a flying car’s energy system, we must compare different battery chemistries and motor types based on efficiency, weight, and lifespan.

Table 1. Battery technology comparison.

Battery Type	Energy Density (Wh/kg)	Efficiency (%)	Lifespan (Cycles)	Charge Rate (C-Rate)	Cost (USD/kWh)	Weight Impact
Li-ion (NMC)	180–250	95%	1000–3000	2 C	120–150	Moderate
Li-ion (LFP)	140–180	96%	2000–7000	2–3 C	80–120	More weight but durable
Solid-State Li-ion	300–500	98%	5000+	5 C	400+	Very lightweight
Lithium-Sulfur	400–600	85%	500–1000	2 C	200–300	Extremely lightweight
Supercapacitors	5–10	99%	1M+	100 C	High	Heavy, but great for regen braking

shows a comparison of battery technologies.

The best energy storage options include solid-state Li-ion, which offers high efficiency, lightweight design, and fast charging but comes at a higher cost. LFP Li-ion is a more durable and cost-effective alternative, though it has a slightly lower energy density. Additionally, supercapacitors can complement these batteries by enhancing regenerative braking efficiency, improving overall energy recovery and system performance.

Table 2. Motor technology comparison.

Motor Type	Efficiency (%)	Power Density (kW/kg)	Torque Density (Nm/kg)	Cooling Complexity	Cost (USD/kW)	Best Use Case
PMSM (Permanent Magnet Synchronous Motor)	90–95%	3–5	10–15	Moderate	50–100	Best for flying cars
SRM (Switched Reluctance Motor)	85–92%	2–4	8–12	Low	30–60	Robust and cheap
Induction Motor (IM)	85–90%	2–3	6–10	High	40–80	Heavy and less efficient
Axial Flux Motor	95–97%	6–8	15–20	Moderate	80–150	Lightweight and compact

shows the motor technology comparison.

The best motor options for a flying car include Axial Flux PMSM, known for its highest efficiency, lightweight design, and compact structure, and Switched Reluctance Motor (SRM), which offers cost-effectiveness and robustness.

When analyzing energy savings and cost-effectiveness in flying cars, two key factors must be considered. Energy savings can be optimized by selecting efficient battery technologies and motor systems, while regenerative braking using supercapacitors further enhances efficiency during landing or braking phases. Cost-effectiveness depends on the initial investment in battery and motor systems, with long-term savings influenced by battery lifespan, efficiency, and maintenance costs.

5.2. Key Cost-Effectiveness Metrics

A number of important indicators are taken into consideration when assessing the cost-effectiveness of energy systems, such as motors and batteries. The first is the initial cost of energy systems (motor + battery), which varies depending on the technology and kind of components chosen; more sophisticated or efficient systems are typically more expensive up front. The next important indicator, which has a direct effect on the vehicle’s running costs over time, is the Energy Cost per Mile (or km), which shows how much energy is used per unit mile while accounting for both battery and motor efficiency. When comparing various system options over the course of the vehicle’s lifetime, lifetime energy savings measures the difference in total energy consumption for the same operational activity; a system that uses less energy results in considerable long-term savings. Together, these metrics provide a comprehensive assessment of the economic and operational performance of energy systems in electric vehicles.

Energy savings come from optimizing battery technology and motor selection. The energy saving formula is given by

$$\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{S}\mathrm{a}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s}=\left({\displaystyle \frac{\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}-\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{N}\mathrm{e}\mathrm{w}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}}{\mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{u}\mathrm{m}\mathrm{e}\mathrm{d}\mathrm{w}\mathrm{i}\mathrm{t}\mathrm{h}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}}}\right)\times 100$$

(36)

Initial battery and motor costs can be calculated depending on the particular technology selected. By calculating energy consumption, including losses from inefficient motors and batteries, operational costs also have an impact on cost-effectiveness.

The calculation of cost per kWh is provided by

$$\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{k}\mathrm{W}\mathrm{H}={\displaystyle \frac{\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{B}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{y}+\mathrm{M}\mathrm{o}\mathrm{t}\mathrm{o}\mathrm{r}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{O}\mathrm{u}\mathrm{t}\mathrm{p}\mathrm{u}\mathrm{t}\mathrm{O}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{L}\mathrm{i}\mathrm{f}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}}}$$

(37)

Lifetime cost can be expressed as

$$\mathrm{L}\mathrm{i}\mathrm{f}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}=\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}\mathrm{U}\mathrm{s}\mathrm{e}\mathrm{d}\mathrm{O}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{L}\mathrm{i}\mathrm{f}\mathrm{e}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e}\times \mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{k}\mathrm{W}\mathrm{h}$$

(38)

To optimize cost savings and reduce the environmental footprint in flying cars, it is essential to consider how different battery types and motor technologies impact energy consumption and carbon emissions. Cost savings can be achieved by improving battery and motor efficiency, which reduces operational costs and minimizes the number of energy cycles needed for a given distance. The environmental footprint, particularly carbon emissions, varies depending on the manufacturing processes, energy consumption, and lifecycle emissions of batteries and motors. By evaluating the carbon intensity (grams of CO₂ per kWh) of different battery and motor combinations, we can assess their overall impact and identify the most sustainable solutions.

Table 3. Environmental footprint of batteries and motors.

Technology	Carbon Intensity (gCO2/kWh)
Batteries
Li-ion (NMC)	150–200
Li-ion (LFP)	120–150
Solid-State Li-ion	80–100
Lithium-Sulfur	120–160
Supercapacitors	50–70 (minimal footprint)
Motors
PMSM	20–30
SRM	20–25
Induction Motor	25–35
Axial Flux Motor	15–25

gives the environmental footprint of batteries and motors.

Cost savings from battery and motor optimization can be achieved by analyzing both energy efficiency and operational costs. By comparing the battery cost per kWh with energy consumption, the total operational cost of a flying car can be estimated. Additionally, carbon emissions savings can be determined by evaluating the energy usage of older technologies versus newer, more efficient alternatives over a given distance, highlighting the environmental and economic benefits of advanced battery and motor systems.

Cost savings and the environmental footprint can be given by the following formula:

$$\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{S}\mathrm{a}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s}=\left({\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{S}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}-\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{N}\mathrm{e}\mathrm{w}\mathrm{S}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{S}\mathrm{y}\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}}\right)\times 100$$

(39)

Carbon Footprint Savings are given by:

$$\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{S}\mathrm{a}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{g}\mathrm{s}=\left({\displaystyle \frac{\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}-\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{N}\mathrm{e}\mathrm{w}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}}{\mathrm{C}\mathrm{a}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{n}\mathrm{F}\mathrm{o}\mathrm{o}\mathrm{t}\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{n}\mathrm{t}\mathrm{o}\mathrm{f}\mathrm{O}\mathrm{l}\mathrm{d}\mathrm{T}\mathrm{e}\mathrm{c}\mathrm{h}}}\right)\times 100$$

(40)

6. Flowchart Structure

The simulation models an energy management system for a flying car, integrating battery storage, fuel cells, and solar power while optimizing power demand across different flight phases—takeoff, cruise, landing, and ground operation. The simulation dynamically allocates power from available sources and provides key outputs, including power distribution, battery state of charge (SOC), and system efficiency trends. Enhancements such as regenerative braking during landing and a wind turbine generator in cruise mode further improve energy utilization. These modifications enable energy recovery, reduce battery drain, and enhance overall efficiency by intelligently balancing power sources based on demand.

To further optimize energy management, an AI-based predictive controller using a neural network (NN) is introduced. The NN model predicts power demand based on flight phases, dynamically adjusting energy allocation across the battery, fuel cell, solar, and wind power sources. This predictive approach minimizes unnecessary battery drain, enhances energy efficiency, and ensures optimal power distribution before high-demand conditions arise. A comparative analysis of SOC values with and without regenerative braking demonstrates the impact of energy recovery, highlighting how regenerative braking extends battery life and contributes to system sustainability. Figure 6 shows the flowchart steps of AI energy management scheme.

A lightweight design is crucial to minimizing fuel or energy consumption. Flying cars require powerful engines or electric motors to generate the necessary thrust for lift. The power requirements are demonstrated through the following performance parameters for a flying car (UTD flying car).

Table 4. Power and energy requirements for lift and propulsion.

Parameter	Value
Cruise velocity	150 mile/h (67.06 m/s)
Cruising altitude	10,000 ft (3048 m)
Vertical lift duration	10 min (3600 s)
Vehicle mass	3000 lb (1360.78 kg)
Range	>300 mile (482.8 km)
Maximum ground speed	80 mile/h
Operation mode	Pure electric or hybrid vehicle
Electric-only range	40 mile

presents the power and energy requirements necessary for both lift and propulsion.

7. Results

Figure 7 outlines the energy demand across different flight phases, highlighting the peak power requirements during takeoff and cruise, while showing reduced demand during landing and ground operations. Figure 8 further illustrates this optimization by showing the individual participation of each energy source in real-time. It clearly visualizes the dynamic adjustments made by the AI system, with regenerative braking energy represented as a negative contribution, indicating recovered energy being fed back into the system. This negative sign emphasizes the vital role of regenerative braking in reducing net energy consumption and improving overall system efficiency.

Figure 9 highlights the direct impact of regenerative braking on battery state of charge (SOC), emphasizing its critical role in energy recovery. The figure shows a noticeable increase in battery SOC after approximately 250 s, directly corresponding to the onset of regenerative braking. This demonstrates how energy that would otherwise be lost during braking is efficiently captured and reused, enhancing the battery’s longevity and overall energy efficiency.

Figure 10 breaks down the energy supplied by different sources, showcasing the contribution of the battery, fuel cell, solar panels, and wind turbine. The recovered energy is shown also in this figure. Figure 11 provides system efficiency metrics, offering insight into the overall energy utilization effectiveness.

Figure 12 presents a bar plot of power demand across different phases, visually emphasizing the high energy requirements of takeoff and cruise. The maximum demand is 200 kW in case of takeoff phase. The power demand is 100 kW and 50 kW in the case of the cruise and landing phases, respectively. The lowest demand is about 20 kW and occurs at the ground phase. Figure 13 provides a comprehensive analysis of various efficiency metrics that collectively offer a holistic evaluation of the system’s overall performance. These metrics include Energy Storage Efficiency, Energy Recovery Efficiency, and cost efficiency, each highlighting a different aspect of system effectiveness and operational viability. Energy Storage Efficiency is calculated at an impressive 100%, suggesting a potential anomaly or overestimation in the stored energy relative to the energy input. This could indicate measurement discrepancies, energy recycling effects, or additional system dynamics that require further investigation. Energy Recovery Efficiency stands at 41.3793%, reflecting the proportion of stored energy successfully reclaimed for useful output. This relatively moderate recovery rate suggests room for optimization in the storage and discharge processes to minimize energy losses and improve overall system responsiveness. The cost efficiency, measured at approximately 23.4029%, evaluates the economic feasibility of the system by comparing the benefits of energy savings or recovery to the associated operational and capital costs. This metric underscores the importance of balancing technical performance with financial sustainability. Together, these efficiency indicators provide valuable insights into the system’s strengths and areas for improvement, guiding efforts to enhance performance, reliability, and cost-effectiveness.

Figure 14 analyzes the carbon footprint based on energy sources, highlighting the sustainability benefits of renewable energy integration. Figure 15 analyzes the carbon footprint based on energy sources, emphasizing the sustainability benefits of integrating renewable energy. The total carbon emissions are calculated at 8.6122 kg CO₂, based on specific carbon emission factors associated with each energy source. Grid-based battery charging contributes 0.5 kg CO₂ per kWh due to its reliance on conventional electricity production. Fuel cell systems, assuming gray hydrogen production, have a significantly higher emission factor of 3.0 kg CO₂ per kWh. In contrast, renewable sources exhibit much lower emissions: solar energy is nearly carbon-neutral at 0.05 kg CO₂ per kWh, though this accounts for the life-cycle impact from panel production, which typically emits 30–50 g CO₂ per kWh due to material mining and processing. Wind energy is even cleaner, with an emission factor of just 0.011 kg CO₂ per kWh, reflecting the minimal environmental impact of its electricity generation process. This breakdown highlights the critical role of renewable energy in reducing the overall carbon footprint and promoting sustainable energy solutions.

Figure 16 identifies different energy losses in the system, while Figure 17 and Figure 18 specifically compare battery and motor efficiency losses. Figure 16 provides a detailed breakdown of the different energy losses within the system, while Figure 17 and Figure 18 specifically focus on comparing the efficiency losses of the battery and motor. The total system losses amount to 1.5615 kWh, distributed among battery losses at 0.14528 kWh, power electronics losses at 0.52467 kWh, and motor losses at 0.89158 kWh. These losses highlight the critical need for improving component efficiency to enhance overall system performance. Advancements in battery and motor technology contribute significantly to energy and cost savings. Energy savings from improved battery technology reach 10%, while motor technology improvements yield an 8.3333% reduction in energy consumption. Figure 19 contrasts energy savings across different operational strategies, while Figure 20 examines lifetime cost implications. These efficiency gains translate directly into lower lifetime costs: the old battery system incurs a lifetime cost of USD 7.5, which decreases to USD 5.4 with the new battery technology—a 20% cost saving. Similarly, the old motor’s lifetime cost of USD 6 drops to USD 3.3 with the adoption of new motor technology, reflecting an impressive 40% cost saving. Together, these enhancements underscore the potential for both energy efficiency and economic benefits through technological upgrades.

Figure 21 and Figure 22 compare cost and carbon footprint savings, highlighting the long-term economic and environmental benefits of an AI-driven energy management system. Battery technology improvements lead to a 25% reduction in carbon footprint, while advancements in motor technology achieve a 16.6667% reduction. These savings are calculated based on the carbon intensity of different technologies, measured in grams of CO₂ per kilowatt-hour (gCO₂/kWh). Battery carbon intensity varies across different chemistries, with values ranging from 70 to 200 gCO₂/kWh, while motor carbon intensity ranges from 20 to 35 gCO₂/kWh. Assuming a total energy usage of 1000 kWh over a specific period, the system’s efficiency improvements translate directly into reduced emissions and operational costs. Battery technologies like Li-ion (NMC), Li-ion (LFP), solid-state Li-ion, Lithium-Sulfur, and Supercapacitors exhibit efficiencies between 85% and 99%, with supercapacitors leading in performance. Similarly, motor technologies such as Permanent Magnet Synchronous Motors (PMSM), Switched Reluctance Motors (SRM), Induction Motors, and Axial Flux Motors show efficiencies between 88% and 97%, with Axial Flux Motors being the most efficient. These insights emphasize the critical role of technology selection and AI-driven optimization in minimizing environmental impact and maximizing cost efficiency.

Figure 23 presents a comparative analysis of battery system costs, specifically between solid-state batteries and Li-ion battery systems. While it effectively highlights cost differences, a more comprehensive evaluation would include additional factors such as energy density, lifespan, and manufacturing scalability. These aspects are crucial in assessing the overall performance and feasibility of each battery technology. Additionally, specifying the cost metric (e.g., USD/kWh or total system cost) would enhance clarity and allow for a more precise comparison. Figure 24, Figure 25 and Figure 26 further expand the analysis by comparing energy density, battery lifespan, and manufacturing scalability, respectively, providing a broader perspective on the advantages and limitations of each battery system.

Figure 27 provides a comprehensive analysis of the life-cycle cost assessment (LCA) of key components in a hybrid electric flying car. Figure 27a presents a bar chart illustrating the total cost associated with each component, including the battery, fuel cell, solar panel, wind turbine, motor, power electronics, and maintenance. The breakdown includes initial purchase costs, replacement costs, and operational expenses over the vehicle’s lifetime. Figure 27b visualizes the cost contribution of each component as a pie chart, highlighting the relative financial impact of each subsystem. This analysis emphasizes the significant cost factors in the system, offering valuable insights into cost optimization strategies for future sustainable flying car designs.

The performance of the AI-driven energy management system (EMS) is highly influenced by sensor noise and environmental disturbances, such as fluctuating wind speeds, variable solar irradiance, and measurement inaccuracies in power demand and state of charge (SOC). To assess these impacts, the model was tested under simulated noise and disturbances in sensor data.

Sensor noise was represented as Gaussian white noise added to power demand, solar, wind, and SOC measurements, while environmental disturbances—such as sudden drops in solar power due to cloud cover or fluctuations in wind power—were introduced as time-varying stochastic processes. The AI-driven EMS was then analyzed under these conditions to evaluate its prediction accuracy, power distribution efficiency, and SOC stability. Figure 28 presents the power demand with noise, Figure 29 illustrates the SOC with noise and environmental disturbances, and Figure 30 displays the impact of environmental disturbances on renewable sources.

The results indicate that with a sensor noise level of 5% standard deviation relative to the measured signal, the AI-driven EMS maintained an overall system efficiency of 89.6%, with SOC deviations constrained to ±3%. However, when the noise level increased to 10%, efficiency dropped to 84.2%, and SOC fluctuations became more pronounced (±7%). These findings highlight the robustness of the EMS but also emphasize the necessity for advanced filtering techniques—such as Kalman filters or adaptive observers—to mitigate sensor noise and maintain optimal performance under varying environmental conditions. The presented figures collectively illustrate the performance, efficiency, and sustainability of the AI-optimized energy management system for a flying car.

8. Discussion

The results presented in Figure 7, Figure 8, Figure 9, Figure 10, Figure 11, Figure 12, Figure 13, Figure 14, Figure 15, Figure 16, Figure 17, Figure 18, Figure 19, Figure 20, Figure 21, Figure 22, Figure 23, Figure 24, Figure 25, Figure 26, Figure 27, Figure 28, Figure 29 and Figure 30 provide a thorough evaluation of the system’s energy management, efficiency, cost-effectiveness, and environmental impact. The analysis of energy demand and real-time energy source participation reveals the AI-driven system’s ability to dynamically adjust power distribution, with regenerative braking playing a crucial role in reducing net energy consumption and enhancing battery state of charge (SOC). Energy recovery from braking and the balanced contribution of multiple sources—including the battery, fuel cell, solar panels, and wind turbines—lead to optimized system performance. Efficiency metrics like Energy Storage Efficiency (100%), Energy Recovery Efficiency (41.38%), and cost efficiency (23.4%) reflect both the system’s strengths and opportunities for improvement, guiding further enhancements in performance and financial sustainability.

Moreover, the system’s environmental and economic benefits become clear through carbon footprint analysis and cost-saving evaluations. The integration of renewable energy sources significantly reduces total carbon emissions to 8.6122 kg CO₂, with solar and wind energy showing near-carbon-neutral performance. Improved battery and motor technologies contribute not only to energy savings (10% and 8.33%, respectively) but also to substantial lifetime cost reductions—20% for batteries and 40% for motors. These advancements also lower carbon intensity and enhance system efficiency, with battery and motor efficiencies reaching up to 99% and 97%, respectively. Ultimately, the AI-driven energy management system demonstrates remarkable potential for minimizing environmental impact and maximizing long-term cost efficiency.

In real-time implementation, the performance of the proposed AI-driven energy management system (EMS) may be affected by several practical factors, including sensor noise, communication delays, and computational constraints. While the simulation results show efficient power distribution and optimized energy usage, real-world conditions introduce uncertainties that could impact system behavior. For instance, sensor noise and environmental disturbances can lead to inaccurate state-of-charge (SOC) estimation and power demand prediction. Communication latency between the AI controller and power sources may delay decision-making, affecting the system’s dynamic response. Moreover, the real-time computational load of running complex neural networks must be optimized to ensure fast and stable operation. To mitigate these challenges, advanced filtering techniques, robust control algorithms, and hardware-in-the-loop (HIL) testing can be incorporated to validate the system’s real-time performance and enhance its reliability in practical applications.

Under moderate noise conditions, the AI-driven energy management system (EMS) maintains high efficiency and steady SOC levels, exhibiting great resilience against environmental disturbances and sensor noise. However, efficiency decreases and SOC fluctuations intensify with increasing noise levels, highlighting the significance of sophisticated filtering methods like adaptive observers or Kalman filters. The results demonstrate how the EMS can handle uncertainties and optimize power distribution, giving it a practical option for flying cars’ sustainable energy management. The data that are displayed give a thorough picture of how well it performs, highlighting the necessity for additional improvements to guarantee resilience in actual operational settings.

9. Conclusions

This study demonstrates the effectiveness of an AI-driven energy management system (EMS) in optimizing power distribution for a hybrid electric flying car. By integrating multiple power sources—including batteries, fuel cells, solar panels, and wind turbines—the system dynamically allocates power to enhance overall efficiency. The AI-driven control approach efficiently integrates regenerative braking during landing, reduces battery drain, and makes use of renewable energy sources. Regenerative braking is essential for prolonging battery life and lowering overall energy usage, according to status of charge (SOC) studies.

Important performance indicators like cost analysis, power losses, and system efficiency also emphasize the benefits of AI-optimized energy distribution. The findings show that AI-based control improves efficiency by lowering energy losses in motors, batteries, and power electronics. By drastically cutting emissions and lessening the impact on the environment, carbon footprint analysis highlights the sustainability advantages of incorporating renewable energy sources. Additionally, cost efficiency analysis confirms long-term savings, making AI-enhanced energy management a practical and economical solution for future flying car propulsion systems. These findings suggest that AI-based energy optimization can enhance the sustainability, reliability, and economic feasibility of hybrid electric flying vehicles.

AI-driven controllers mark a significant breakthrough in dynamic energy management for aircraft systems, going beyond short-term enhancements in energy distribution. These systems improve efficiency, safety, and dependability in aerospace operations by precisely forecasting energy demand, maximizing power distribution, and taking into account elements like thermal stress accumulation. Future aerospace energy management could be greatly enhanced by ongoing developments in AI algorithms, real-time monitoring, and computer resources, even though integration and data availability issues still exist. This study emphasizes how intelligent energy management systems have the potential to transform energy use in aviation and other energy-intensive applications, promoting cost-effective and sustainable operations.

The proposed EMS also contributes significantly to environmental sustainability by optimizing power distribution and reducing carbon emissions in electric flying vehicles. Through the effective integration of renewable energy sources, regenerative braking, and advanced predictive control, the system reduces reliance on high-emission power sources such as fuel cells. A life-cycle carbon assessment (LCA) highlights substantial reductions in the overall carbon footprint, especially when utilizing solid-state batteries and wind energy. By prioritizing environmental goals—such as lowering greenhouse gas emissions and enhancing energy efficiency—this research aligns closely with global sustainability efforts in the aviation sector. Additionally, a comparative analysis of battery system costs was presented, specifically between solid-state batteries and Li-ion battery systems. While this analysis effectively highlights cost differences, a more comprehensive evaluation should include additional factors such as energy density, lifespan, and manufacturing scalability. These aspects are crucial in assessing the overall performance and feasibility of each battery technology. Specifying the cost metric (e.g., USD/kWh or total system cost) would further improve clarity and allow for a more precise comparison. Future research could expand this analysis to provide a broader perspective on the advantages and limitations of different battery systems, incorporating key metrics like energy density, battery lifespan, and scalability in manufacturing.

Overall, this work demonstrates the transformative potential of AI-driven energy management in flying vehicles, paving the way for more efficient, sustainable, and economically viable aerospace energy solutions.