Strategy for Renewable Energy Consumption Based on Scenario Reduction and Flexible Resource Utilization

Abstract

With the growing global emphasis on renewable energy, the issue of renewable energy consumption has emerged as a hot topic of current research. In response to the volatility and uncertainty in the process of renewable energy consumption, this study proposes a renewable energy consumption strategy based on scenario reduction and flexible resource utilization. This strategy aims to achieve the efficient utilization of renewable energy sources through optimized resource allocation while ensuring the stable operation of the power system. Firstly, this study employs scenario analysis methods to model the volatility and uncertainty of renewable energy generation. By applying scenario reduction techniques, typical scenarios are selected to reduce the complexity of the problem, providing a foundation for the construction of the optimization model. At the same time, by fully considering the widely available small-capacity energy storage units within the system, a flexible cloud energy storage scheduling model is constructed to assist in renewable energy consumption. Finally, the validity and feasibility of the proposed method are demonstrated through case studies. Through analysis, the proposed scenario generation method achieved a maximum value of 26.28 for the indicator $I_{DBI}$ and a minimum value of 1.59 for the indicator $I_{CHI}$. Based on this foundation, the cloud energy storage model can fully absorb renewable energy, reducing the net load peak-to-trough difference to 1759 kW, a decrease of 809 kW compared with the traditional model.

1. Introduction

Since the Industrial Revolution, while society has developed rapidly, humans have developed and utilized conventional fossil energy sources on a large scale. Facing the multiple pressures of energy depletion, environmental pollution, and climate deterioration, it is urgent to vigorously develop renewable energy and achieve an energy structural transformation [1,2]. Renewable energy sources, represented by wind power and photovoltaic power generation, boast abundant reserves, low development costs, and pollution-free characteristics. Compared to traditional energy sources, renewable energy is the optimal choice from both ecological and resource perspectives, gradually becoming the preferred option for sustainable human energy development [3]. With large-scale and high-percentage integration of new energy into the power grid, controllable power sources such as thermal power plants need to frequently adjust their output and, when necessary, start or stop operations to track peak and off-peak load changes. Currently, thermal power units generally suffer from insufficient peak shaving depth and lack of initiative in peak shaving. When the operating state of a thermal power unit deviates from its rated operating condition, the power generation efficiency decreases with the reduction in load rate, leading to decreased operational economy [4].

In formulating dispatch decisions for high-penetration renewable energy power systems, the uncertainty associated with renewable energy output scenarios often leads to conservative decision-making and increased economic investments. Scenario reduction methods serve as an effective tool for characterizing this uncertainty. In recent years, substantial research has been conducted by scholars both domestically and internationally on the generation of typical wind/photovoltaic (PV) scenarios in power systems [5]. Among them, scenario analysis enables the analysis of renewable energy uncertainty based on potential wind/PV output scenario datasets. By selecting reasonable typical renewable energy output scenarios, it characterizes the stochastic nature of the regional renewable energy output, thereby mitigating the negative impacts of uncertainty [6,7] and providing a decision-making basis for grid dispatch planning and other related tasks. Existing methods for generating typical scenarios mainly encompass probability modeling, deep learning, and clustering-based scenario generation. Probability modeling primarily utilizes statistical methods in conjunction with sampling techniques such as Monte Carlo for joint density modeling, generating typical wind and PV output or load scenarios. Reference [8] employs Gaussian kernel functions to generate probability density functions for wind and PV output at each time interval, combined with Copula theory to construct typical wind–solar output scenarios. Reference [9] assumes that the renewable energy output and load forecast errors approximately follow normal distributions, simulating prediction error scenarios through Markov chain multi-scenario approaches. Reference [10] describes regional wind speeds based on the Weibull distribution and probabilistically models light intensity using the Beta distribution. Reference [11] integrates quantile regression theory with Gaussian distribution theory, leveraging the concept of the Cartesian product to construct a statistical scenario set for renewable energy. Reference [12] predicts wind speeds using an autoregressive integrated moving average model and combines Latin Hypercube Sampling to generate future wind power output scenarios. Reference [13] performs probability modeling of wind power, electricity, and heat demand using Monte Carlo simulation methods, conducting uncertainty analysis based on scenario methods. Reference [14] proposes a method for acquiring wind speed scenarios based on C-vine Copula theory, which significantly increases the number of modeled random variables due to its consideration of spatio-temporal tail dependence structures.

To efficiently solve planning and operation models, many scholars have adopted clustering analysis to reduce scenarios. This involves classifying scenarios based on their original characteristics to decrease the number of scenarios, thereby enhancing solution efficiency. Considering the uncertainty of distributed generation in distribution networks, Reference [15] constructs a multivariate time-series dataset based on wind, solar, and load power levels and employs clustering based on correlation features to obtain typical scenarios and their corresponding weights. In the field of typical scenario generation, traditional methods such as K-means clustering [16], the Analytic Hierarchy Process, and Fuzzy C-Means clustering [17] have matured. Among them, K-means clustering is widely used due to its good clustering performance and high computational efficiency. Reference [18] applies multi-scenario analysis and K-means clustering to address distributed generation uncertainty; Reference [19] generates typical scenarios to describe distributed generation uncertainty through an improved bi-level clustering method; Reference [20] proposes a bi-level clustering algorithm that considers both micro- and macro-grid structures to achieve the similarity analysis of distribution network structures; Reference [21] employs a multi-scenario analysis method based on improved FCM clustering to describe the uncertainty of wind and solar power output for solving distribution network reconfiguration problems; Reference [22] adopts a multi-scenario analysis method based on K-means clustering to describe the uncertainty of wind and solar power output. However, the above clustering methods rely heavily on the selection of initial clustering centers, which is subject to subjective factors, and the improper selection of initial clustering centers can increase the error in clustering results. Reference [23] utilizes the Density-based Spatial Clustering of Applications with Noise algorithm, which is a density-based clustering method. By processing the electricity consumption behaviors of different users, it can efficiently analyze large-scale user behaviors. Nevertheless, this method also depends on the initial parameter settings.

On the other hand, energy storage systems, due to their charging and discharging characteristics, can maintain real-time power balance and are therefore widely deployed in power systems. Addressing the issue of energy storage deployment in power grids, Reference [24] analyzes the impact of distributed generation integration into distribution networks on power quality and protective device operations, proposing an adaptive current protection method by altering wiring configurations, albeit with a relatively long computation time. References [25,26,27] consider and analyze the optimal deployment of multiple or various types of distributed generations into the grid, establishing a series of mathematical models. By applying data-mining algorithms, they develop optimized dispatch schemes and verify their accuracy. Reference [28] establishes a grid dispatch model incorporating distributed generations, calculating the differences in costs between distributed generations and market electricity prices. Through a specific case study, it investigates the influence of distributed generations on network losses and marginal node prices within the power system. Reference [29] presents an optimization method for the power and capacity of energy storage devices utilizing a hierarchical control approach. By analyzing various control methods, it identifies the optimal rated power and capacity of energy storage devices that minimize output errors and total costs. Reference [30] introduces a capacity optimization method for energy storage devices that considers the load demand and state-of-charge (SOC) partitioning. It translates different operating ranges into impacts on the energy storage device lifespan and wind curtailment, which are further converted into economic implications. An improved particle swarm optimization algorithm is employed to optimize and obtain the results. However, current research has yet to consider the configuration model when wind power, photovoltaic generation, and energy storage are jointly integrated into a distribution network within a specific region. Currently, some researchers are also embarking on studies of more efficient and environmentally friendly new transportation modes from the perspectives of engine performance and fuel properties. This, to a certain extent, has promoted the popularization and consumption of new energy sources, reduced greenhouse gas emissions, and is of great environmental protection value [31,32,33,34,35,36,37,38,39].

The innovative points of this paper can be summarized as follows:

(1)

This paper proposes a GAN-based scenario analysis method and scenario reduction technique to model and optimize the uncertainties associated with renewable energy generation. By selecting typical scenarios, the complexity of the problem is reduced, providing a solid foundation for the construction of the optimization model.

(2)

This paper constructs a flexible cloud energy storage scheduling model that leverages the widely available small-capacity energy storage units within the system. This model employs flexible scheduling strategies to assist in renewable energy consumption, enhancing the flexibility and reliability of the energy system.

2. Method for Characterizing the Uncertainty of Renewable Energy Output Based on Multi-Scenario Analysis

2.1. Modeling Uncertainty in Renewable Energy

(1)

Probability Model of Wind Power Generation

Wind turbines installed at the user side are typically small-capacity units. The output power of these turbines varies according to changes in wind speed, exhibiting characteristics such as random fluctuations. As a result, modeling the probability of wind turbine output can essentially be transformed into modeling the probability of wind speed. Typically, the two-parameter Weibull distribution is employed to model the probability distribution of wind speed, with its probability density function expressed as shown in Equation (1):

$$f\left(v\right)={\displaystyle \frac{k}{c}}{\left({\displaystyle \frac{v}{c}}\right)}^{k-1}\exp \left(-{\left({\displaystyle \frac{v}{c}}\right)}^k\right)$$

(1)

In the equation, v represents the wind speed, $k={(a_v/{\mu }_v)}^{-1.086}$ is the shape parameter, and $c=\left({\mu }_v/\left(\mathsf{\Gamma }((k+1)/k)\right)\right)$ is the scale parameter. Meanwhile, $a_v$ and ${\mu }_v$, respectively, represent the standard deviation and mean of the wind speed.

The corresponding mathematical model for the active power output of the wind turbine is shown in Equation (2). In this paper, a constant power factor model is adopted to model the wind power output, with its reactive power output expressed as shown in Equation (3):

$$P_{WTG}(v)=\left\{\begin{array}{c}0\\ k_{1}v+k_2\\ P_{WTG}^r\end{array}\right.\begin{array}{c}v<v_{ci}\cup v>v_{co}\\ v_{ci}<v<v_r\\ v_r<v<v_{co}\end{array}$$

(2)

$$Q_{WTG}=P_{WTG}\tan {\phi }_{WTG}$$

(3)

Here, $P_{WTG}(v)$ represents the active power output of the wind turbine; $P_{WTG}^r$ and $v_r$, respectively, represent the rated power and rated wind speed of the wind turbine; $v_{ci}$ and $v_{co}$, respectively, represent the cut-in wind speed and cut-out wind speed; $k_1$ and $k_2$ are constants; ${\phi }_{WTG}$ represents the power factor of the wind turbine.

(2)

Probability Model of Photovoltaic Power Generation

Numerous studies have shown that the irradiance over a certain period of time can be represented by a Beta distribution, as shown in Equation (4):

$$f(h)={\displaystyle \frac{\mathsf{\Gamma }(\alpha +\beta )}{h_{\max }\mathsf{\Gamma }(\alpha )\mathsf{\Gamma }(\beta )}}{\left({\displaystyle \frac{I_l}{I_{\max }^{ra}}}\right)}^{\alpha -1}{\left(1-{\displaystyle \frac{I_l}{I_{\max }^{ra}}}\right)}^{\beta -1}$$

(4)

Here, $\mathsf{\Gamma }(*)$ represents the gamma function; $I_{\max }^{ra}$ represents the maximum irradiance (W/m²); $\alpha$ and $\beta$ are the shape parameters of the Beta distribution, which can be calculated based on the mean and variance of the irradiance over a certain period of time.

As the irradiance increases, the output power of the photovoltaic power generation equipment gradually increases until it reaches the rated power $P_{PVG}^r$. The active power output of the photovoltaic system $P_{PVG}$ can be expressed by Equation (5):

$$P_{PVG}=\left\{\begin{array}{c}\begin{array}{cc}{\displaystyle \frac{P_{PVG}^r{I}_l}{I_r}}& I_l\le I_r\end{array}\\ \begin{array}{cc}{P}_{PVG}^r& I_l>I_r\end{array}\end{array}\right.$$

(5)

where $I_r$ represents the rated value of irradiance.

2.2. Introduction to Scenario Generation Method

In this paper, a Generative Adversarial Network (GAN) is employed to construct a scenario set for renewable energy output. A GAN is an unsupervised neural network that continuously fits a model by training on real datasets, implicitly learning the intrinsic probability distribution of the data. Its basic structure comprises a generator and a discriminator, as shown in Figure 1. GANs are widely used in generating large amounts of similar data and improving data quality, even when the quality of the original data is not high. They have been applied to solve problems such as generating power data samples for new energy scenario generation and unit fault diagnosis. In the operation of power systems, the advantages of using GANs to generate typical operating scenarios are primarily embodied in the following aspects: GANs are capable of automatically learning the complex data distributions within power systems, such as the power output characteristics of renewable energy sources like wind and solar, which are often characterized by high levels of uncertainty and volatility. Through the generator of GANs, it is possible to simulate operating scenarios that closely resemble real-world conditions, providing robust support for power system planning and dispatch. The samples generated by GANs possess high fidelity, enabling the creation of diverse and typical operating scenarios. These scenarios can encompass the various operational states of power systems under different conditions, facilitating the assessment of power system stability and reliability. In contrast to traditional scenario generation methods that may require vast amounts of historical data and complex computational processes, GANs, through their adversarial training mechanism, can generate a large number of high-quality scenarios in a relatively short time. This significantly enhances the efficiency of scenario generation and reduces computational costs. Furthermore, GAN models exhibit high flexibility, allowing for customization and optimization according to different requirements and application scenarios. For instance, they can be trained to generate operating scenarios that are more tailored to the specific characteristics of power systems in particular regions.

In Figure 1, the generator G is used to capture the data distribution, while the discriminator D is used to determine the probability that a generated sample comes from the real data. The two models are trained simultaneously. The training objective of the generator G is to deceive the discriminator D with generated data, maximizing the probability of the discriminator D making a mistake. The loss functions of the generator and discriminator can be calculated according to the following equations:

$$Los{s}_G=-\underset{z\in p_z(z)}{E}\left\{D[G(z)]\right\}$$

(6)

$$Los{s}_D=-\left\{\underset{x\sim p_{data}(x)}{E}[\lg D(x)]\right\}+\underset{z\sim p_z(z)}{E}\left\{\lg \left\{1-D[G(z)]\right\}\right\}$$

(7)

Here, E(·) represents the expectation function; G(·) and D(·) are the functions of the generator and discriminator, respectively; p_z(z) and p_data(x) are the distributions of the input noise z and the sample input x, respectively.

$$\underset{G}{\min }\max V(D,G)=\underset{x\sim p_{data}(x)}{E}[\lg D(x)]+\underset{z\sim p_z(z)}{E}\left\{\lg \left\{1-D[G(z)]\right\}\right\}$$

(8)

V(D,G) is a binary cross-entropy function. The ultimate goal of (8) is to minimize the Jensen–Shannon distance between the probability distribution of the generated samples and the real data distribution.

In the space of any functions G and D, the unique solution that allows G to reproduce the distribution of the training data is when D = 0.5. Throughout the training process, no Markov chain or unfolded approximate inference network is required. Additionally, when the generator G and discriminator D are composed of multi-layer perceptions, the entire system can be trained using backpropagation.

2.3. Typical Scenario Screening Method

In Section 2.2, the strategy for generating renewable energy output scenarios was introduced. However, while a large number of scenarios can accurately characterize the uncertainty of renewable energy output, they also increase the computational complexity of the model. Therefore, this section focuses on screening typical scenarios through a selection strategy. This approach not only accurately captures the uncertainty of renewable energy but also significantly reduces the computational complexity. Current methods for screening typical scenarios include the intra-period mean method, intra-period typical day method, backward reduction method, etc. In this paper, we primarily adopt the Clustering by Fast Search and Find of Density Peaks clustering method.

Density-based clustering methods classify data based on data density. Compared to partition-based clustering algorithms, density-based clustering algorithms can identify data with arbitrary distributions. The Fast Search and Find of Density Peaks clustering method focuses on selecting cluster centers. The cluster centers are chosen based on the principles of “having a high density and being surrounded by points with lower densities” and “being far away from other high-density points.” Points with a high product of local density and relative distance are selected as initial cluster centers.

This algorithm defines two parameters for each sample: the local density of the sample ${\rho }_i$ and the distance to points with higher density than the sample point ${\delta }_i$. A cutoff distance $d_c$ is selected as a hyperparameter input into the system. The local density ${\rho }_i$ can be approximately viewed as the number of sample points within a distance less than the cutoff distance $d_c$ from sample point i. There are two forms of calculating local density: Gaussian kernel and cutoff kernel. To make it less likely for data points with the same density to be clustered together, this paper selects the Gaussian kernel to determine the cutoff distance. The local densities ${\rho }_i$ are then arranged in descending order. The calculation formulas are shown in Equations (9)–(12). Among them, the formula for the cutoff kernel is given in Equation (9):

$${\rho }_i={\displaystyle \sum _{j}f(d_{ij}-d_c)}$$

(9)

Here, $d_{ij}$ represents the Euclidean distance between data points x_i and x_j. $d_c$ is a constant. The operational meaning of the function $f(·)$ is given in Equation (10):

$$f\left(x\right)=\left\{\begin{array}{cc}1\hfill & x<0\\ 0\hfill & x>=0\end{array}\right.$$

(10)

The formula for the Gaussian kernel calculation is given in Equation (11):

$${\rho }_i^{\prime }={\displaystyle \sum _j{e}^{-{\left(\frac{d_{ij}}{d_c}\right)}^2}}$$

(11)

Given that the algorithm is more sensitive to local density, it sets that only points with ${\rho }_i$ and ${\delta }_i$ relatively large distances to their neighboring points are considered as cluster centers. In other words, a sample can be regarded as a cluster center when both its local density and the distance to points with a higher density are relatively large, while the remaining sample points are considered as non-cluster centers. To adaptively determine the number of cluster centers, this paper selects a comprehensive consideration of the product parameter of ${\rho }_i$ and ${\delta }_i$, as shown in Equation (12):

$${\gamma }_i={\rho }_i{\delta }_i$$

(12)

Then, the initial cluster centers and the number of clusters are determined through a clustering decision graph. This graph sorts the data in descending order of ${\gamma }_i$, and the data points with relatively high values of ${\gamma }_i$ are selected as cluster centers by observation.

After the cluster centers are determined, the remaining sample points are assigned to the cluster of the sample with a higher local density and the closest distance to it. The cutoff distance $d_c$ represents the neighborhood radius, and the number of samples within the cutoff distance indicates the local density. The Fast Search and Find of Density Peaks clustering algorithm has a simple structure and requires fewer parameters. By selecting cluster centers through the clustering decision graph, it can overcome the shortcomings of the subjective selection of cluster centers in clustering algorithms such as K-means and Density-based Spatial Clustering of Applications with Noise. However, the selection of cluster centers relies on the local density ${\rho }_i$ and relative distance ${\delta }_i$, and these two parameters depend on the distances $d_c$ between data points. When the dataset is large, the computational complexity grows exponentially with the number of samples. In summary, the steps of the Fast Search and Find of Density Peaks algorithm are shown in

Table 1. The detailed steps of the Fast Search and Find of Density Peaks algorithm.

Step	Operation
Step 1	Input: dataset $X=\left\{x_1,x_2,\cdots ,x_n\right\}$, parameter $d_c$
Step 2	Calculate the Euclidean distance $d_{ij}$ between points and construct a similarity matrix
Step 3	Based on the similarity matrix and data attributes, a decision graph is plotted, and sample points with both relatively large ${\rho }_i$ and ${\delta }_i$ are selected as cluster centers
Step 4	Assign the remaining non-cluster center points to the cluster of the nearest point with a higher density than that point
Step 5	Remove noise points from the current cluster that are smaller than the boundary threshold
Step 6	Output: Clustering result Y

.

2.4. Evaluation Index

In the field of typical scenario generation for power systems, unsupervised learning methods are often employed for clustering due to the absence of class information or prior knowledge in the input scenario samples. To further validate the scientific selection of typical output scenarios and the effectiveness of the proposed improvement methods, the Davies–Bouldin index (DBI) and the Calinski–Harabasz index (CHI), which are commonly used evaluation metrics for clustering performance, are utilized to comprehensively assess the clustering results.

The DBI index quantifies the similarity between different clusters by measuring the Euclidean distance between clusters and the size of the clusters themselves. A larger cluster size combined with a smaller distance between clusters indicates a higher similarity between the two clusters. Conversely, a smaller cluster size with a larger distance between clusters suggests a lower similarity. A lower DBI value indicates a better clustering performance. The corresponding formula is expressed as follows:

$$I_{DBI}={\displaystyle \frac{{\displaystyle \sum _{w=1}^K{R}_w}}{K}}$$

(13)

$$\left\{\begin{array}{c}{R}_{wq}=\max {\displaystyle \frac{S_w+S_q}{{\xi }_{wq}}},\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}q=1,2,3,\cdots ,K;q\ne w\\ R_w=\max R_{wq},\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}q\ne w\end{array}\right.$$

(14)

$$\left\{\begin{array}{c}{\xi }_{wq}=d(m_w,m_q)={‖m_w-m_q‖}_2\\ S_w=\sqrt{{\displaystyle \frac{1}{N_i}}{\displaystyle \sum _{j=1}^{N_i}{\left|X_j-m_w\right|}^2}}\end{array}\right.$$

(15)

In the formula, I_DBI represents the calculated Davies–Bouldin index; K is the number of clusters in the clustering; $R_w$ is the similarity between the cluster center of the w-th cluster and the cluster center of the q-th cluster; $S_w$ corresponds to the diameter of the w-th cluster, which is the average distance between each sample within the cluster and the corresponding cluster center; and ${\xi }_{wq}$ is the inter-cluster distance, representing the Euclidean distance between the cluster centers of the w-th and q-th clusters.

Similar to the DBI index, the CHI is also a common parameter used to measure the effectiveness of clustering. The CHI index relies on the between-cluster and within-cluster dispersion matrices to characterize the degree of separation between clusters and the degree of cohesion within clusters, respectively. The ratio between these two quantities is used to represent the clustering quality. A higher value of the CHI index indicates that the distances between clusters are larger, resulting in a higher dissimilarity between different clusters, while the distances within clusters are smaller, indicating a higher similarity within the same cluster. The specific expression is as follows:

$$I_{CHI}={\displaystyle \frac{B_k}{K-1}}/{\displaystyle \frac{W_k}{n-K}}$$

(16)

$$B_k={\displaystyle \frac{1}{2}}[(n-K){\lambda }_k+(K-1){\overline{d}}^2]$$

(17)

$$W_k={\displaystyle \frac{1}{2}}[(n_1-1){\overline{d}}_1^2+\dots +(n_k-1){\overline{d}}_k^2]$$

(18)

$${\lambda }_k={\displaystyle \frac{1}{n-K}}{\displaystyle \sum _{i=1}^K(n_i-1)}({\overline{d}}^2-{\overline{d}}_i^2)$$

(19)

Here, n represents the number of input samples; $B_k$ represents the distance between the various clusters in the clustering result; $W_k$ represents the intra-cluster distance within each cluster in the clustering result; ${\overline{d}}_k^2$ represents the average Euclidean distance between all samples in the input sample set; ${\overline{d}}_j^2$ is the average Euclidean distance between the samples within the j-th cluster.

3. Scheduling Strategy for Renewable Energy Based on a Cloud Energy Storage Model

3.1. Concept and Working Mechanism of Cloud Energy Storage

Cloud Energy Storage (CES) represents a new form of shared service based on existing, established energy storage devices. It aggregates decentralized energy storage resources and enables their sharing through the cloud, leveraging internet technology. The cloud-based energy storage capacity is primarily provided by large-scale energy storage facilities, supplemented by distributed energy storage devices, allowing users to access storage resources anytime, on demand, without the additional costs of installing and maintaining their own energy storage equipment. Users interact with CES capacity similarly to physical energy storage devices, controlling charging and discharging according to their needs. This eliminates the burden of managing and maintaining energy storage infrastructure. CES service providers aggregate these dispersed storage devices, managing them centrally for maintenance, operation, and scheduling. By leveraging shared resources, they achieve economies of scale, improving the utilization of idle resources and reducing overall costs.

Given that users’ energy consumption patterns vary both within and between days, there is inherent complementarity in their storage needs. CES enables meeting these needs with minimal investment in storage equipment. The entities offering CES services are known as CES service providers. They set prices for cloud storage capacity and energy usage, and users purchase the right to use CES devices based on their storage needs. Information and financial transactions flow bidirectionally between CES providers and users through communication and financial systems, with energy exchanges occurring over the power grid. When the distributed renewable energy output cannot meet charging demands, users can purchase electricity from the grid at real-time prices to charge CES devices. If users utilize their own renewable energy for charging, no charging fees apply. Discharging from CES devices is also free. When users have surplus renewable energy and their net load is negative, they can sell electricity back to the grid, earning repurchase revenue. As a third-party platform, CES providers connect with both the grid and users, collecting critical information such as user loads, grid prices, and storage system configurations. Through comprehensive analysis and optimization, they provide users with optimal charging and discharging plans for cloud batteries. The typical working mechanism of CES is illustrated in the diagram below. As shown in Figure 2 below, this model benefits both users, who save on electricity costs, and grid companies, which can generate additional revenue. The peak-to-valley difference in power system loads, a crucial indicator for studying peak shaving measures and energy conservation, can be mitigated by harnessing the dynamic response capabilities of energy storage systems. By reducing this difference, CES systems significantly enhance grid efficiency and offer substantial application value.

3.2. Optimal Dispatching Model for Renewable Energy with the Participation of Cloud Energy Storage

Without energy storage, the power $P_{i,t}^{in}$ that user i needs to obtain from the grid during time period t is as follows:

$$P_{i,t}^{in}={(L_{i,t}-P_{i,t}^{renew})}^{+}$$

(20)

The power that user i still has left after meeting its own load during period t and sends back to the grid is as follows:

$$P_{i,t}^{out}={(L_{i,t}-P_{i,t}^{renew})}^{-}$$

(21)

In Equations (20) and (21), $P_{i,t}^{in}$ represents the load of user i during period t, and $P_{i,t}^{renew}$ represents the renewable energy power generated by user i during period t.

The rules for the functions ${(·)}^{+}$ and ${(·)}^{-}$ are defined as follows:

$${(x)}^{+}=\left\{\begin{array}{c}x\hspace{0.33em}\hspace{0.33em}x\ge 0\\ 0\hspace{0.33em}\hspace{0.33em}x<0\end{array}\right.$$

(22)

$${(x)}^{-}=\left\{\begin{array}{c}-x\hspace{0.33em}\hspace{0.33em}x\le 0\\ 0\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}x>0\end{array}\right.$$

(23)

By weighing the relationship among electricity price, load, and renewable energy output, the charging and discharging strategy for user i during period t is formulated as follows: Charge when the electricity price α_t is lower than the critical charging price threshold $a_t^C$ or when the renewable energy output $P_{i,t}^{renew}$ exceeds the load $L_{i,t}$; discharge when the electricity price α_t is higher than the critical discharging price threshold $a_t^D$ and the renewable energy output $P_{i,t}^{renew}$ is less than the load $L_{i,t}$, as shown in Equations (24)–(26):

$$\left\{\begin{array}{c}{P}_{i,t}^C=\min \left\{P_i^{\max },{\displaystyle \frac{E_i^{\max }-E_{i,t-1}}{{\lambda }_i^C\mathsf{\Delta }T}}\right\},a_t\le a_t^C\\ P_{i,t}^D=0\end{array}\right.$$

(24)

$$\left\{\begin{array}{c}{P}_{i,t}^C=\min \left\{P_i^{\max },{\displaystyle \frac{E_i^{\max }-E_{i,t-1}}{{\lambda }_i^C\mathsf{\Delta }T}},P_{i,t}^{out}\right\}\\ P_{i,t}^D=0\end{array}\right.,\hspace{0.33em}\hspace{0.33em}{a}_t^C<a_t<a_t^D$$

(25)

$$\left\{\begin{array}{c}{P}_{i,t}^C=\min \left\{P_i^{\max },{\displaystyle \frac{E_i^{\max }-E_{i,t-1}}{{\lambda }_i^C\mathsf{\Delta }T}},P_{i,t}^{out}\right\}\\ P_{i,t}^D=\min \left\{P_i^{\max },{\lambda }_i^D{\displaystyle \frac{E_{i,t-1}-E_i^{\min }}{\mathsf{\Delta }T}},P_{i,t}^{in}\right\}\end{array}\right.,\hspace{0.33em}\hspace{0.33em}\hspace{0.33em}{a}_t\ge a_t^D\hspace{0.33em}$$

(26)

Here, $P_{i,t}^C$ and $P_{i,t}^D$ are the charging and discharging power of user i’s energy storage device during period t, respectively. $P_i^{\max }$ is the maximum energy storage capacity of user i. $E_i^{\max }$ and $E_i^{\min }$ are the maximum and minimum energy levels, respectively, of user i’s energy storage device. ${\lambda }_i^C$ and ${\lambda }_i^D$ are the charging and discharging efficiencies of user i’s energy storage device, respectively. ΔT represents the unit time interval.

As can be seen from the above equations, the user’s charging power is composed of two parts: when the electricity price is low, the grid power is used to charge the cloud-based battery; and when there is excess renewable energy output, it is used to charge the cloud-based battery. It is stipulated that renewable energy generation first satisfies the user’s load, and any excess renewable energy output is then used to charge the cloud-based battery. If the charging demand cannot be met solely by renewable energy, the grid power will be used.

The charging and discharging demands generated by the user are responded to and executed by the cloud-based energy storage service provider. When the user uses grid power to charge the cloud-based energy storage device, there will be an electricity charge incurred, which is collected by the cloud-based energy storage service provider and ultimately settled with the grid. When the user discharges from the cloud-based energy storage device, there is no cost incurred. If the user still has excess renewable energy power after satisfying their own load and the charging demand of the cloud-based energy storage device, they can feed this excess power back to the grid, and the grid will pay a certain amount of repurchase electricity fee. This fee is also first collected by the cloud-based service provider and ultimately settled with the user. Without considering the service fees paid by the user for the cloud-based energy storage device, the electricity purchase cost of user i from the grid during period t is as follows:

$$M_{i,t}=a_t{(P_{i,t}^{out}-P_{i,t}^{in}+P_{i,t}^D-P_{i,t}^C)}^{-}-{\beta }_t{(P_{i,t}^{out}-P_{i,t}^C)}^{+}$$

(27)

In summary, the renewable energy consumption and user electricity purchase model based on cloud-based energy storage services can be described by the following Equations (28)–(33):

$$\min M_{i,t}$$

(28)

$$s.t.\hspace{0.33em}0\le P_{i,t}^C\le P_{i,t}^{C,\max }$$

(29)

$$0\le P_{i,t}^D\le P_{i,t}^{D,\max }$$

(30)

$$E_i^{\min }\le E_{i,t}\le E_i^{\max }$$

(31)

$$E_i^{\min }={\eta }_{SOC}^{\min }{E}_i^{\max }$$

(32)

$$E_{i,t}=(1-w_i)E_{i,t-1}+\mathsf{\Delta }T({\lambda }_i^C{P}_{i,t}^C-{\displaystyle \frac{P_{i,t}^D}{{\lambda }_i^D}})$$

(33)

The primary focus of this paper is on scheduling strategies, rather than planning strategies. For power system operators, the investment costs associated with renewable energy construction are primarily considered during the planning stage. In other words, in the discussion of power system scheduling in this paper, more emphasis is placed on technical feasibility, system stability, or environmental benefits, while economic feasibility is implicitly assumed or taken as given background information.

4. Case Study

4.1. Introduction to Test System Information

To verify the rationality of the proposed model and the feasibility of the algorithm, simulation analysis is conducted based on typical daily data from a regional power grid. The program in this paper is written in MATLAB (R2023a) and run on a computer with an Intel (R) Core (TM) 2.4 GHz CPU and 16 GB of RAM. Randomly selected, 800 users participate in the cloud-based energy storage service mechanism, and each user purchases a cloud-based energy storage device with a capacity and upper limit of 2500 kW and 5000 kWh, respectively [17]. The minimum state of charge (SOC) of the energy storage device is set to 10%, and the initial SOC is assumed to be 15%. Both the charging and discharging efficiencies of the energy storage device are 96%.

Figure 3 shows the real-time electricity price data for the regional power grid. The real-time electricity price exhibits a stepped characteristic, with prices significantly higher during peak load periods than during off-peak periods. This helps to guide users to purchase electricity during off-peak hours, thereby alleviating system balancing pressure. The critical charging price is set to be 10% lower than the average price of the day, while the critical discharging price is set to be 10% higher than the average price of the day. The grid repurchase price for electricity is set to the average price of the day. During peak load hours, there is a sharp increase in electricity demand, which may exceed the capacity of the power supply to meet all users’ needs. At such times, to maintain the stability and safe operation of the power system, electric utilities must implement a series of measures to ensure adequate power supply, including increasing power generation and adjusting grid operation modes. These measures, in turn, elevate the cost of the power supply, resulting in a corresponding increase in electricity prices. From the perspective of the electricity market, electricity prices are a direct reflection of supply and demand dynamics. When demand exceeds supply, prices rise to discourage some non-essential usage, thereby restoring market balance. This market mechanism encourages users to rationalize their electricity consumption schedules and reduce peak-hour demand [20].

4.2. Effectiveness Validation of Typical Scenario Selection

In this paper, the initial wind–photovoltaic scenarios are divided into four typical output scenarios. The wind power data used in this paper are historical records from the region. This is primarily because power systems have accumulated a vast amount of historical wind power data over long-term operations, which encapsulate seasonal and diurnal variations in wind power output [15]. These data are invaluable for analyzing the uncertainty of wind power. While real-time data can reflect the current actual wind power output, data from a single time point may fail to comprehensively represent the long-term trends and uncertainties of wind power output [17]. Furthermore, real-time data can be subject to instantaneous influences from various factors, such as sudden changes in meteorological conditions or equipment failures, which may cause significant fluctuations in the data, adversely affecting the stability and accuracy of dispatch decisions. Moreover, power system dispatch not only concerns the balance of power supply and demand in the current period but also necessitates consideration of power production plans for the coming period. Historical data provide a more comprehensive picture of wind power output, facilitating the development of more scientific and rational long-term dispatch plans. Through the analysis of historical data, the impact of wind power output uncertainty on power system operation can be assessed, enabling the formulation of corresponding risk mitigation measures. This risk assessment method based on historical data is more robust and reliable [20]. Based on this classification, the scenario partitioning is performed to obtain the typical scenario curves and their corresponding probability distributions. The occurrence probabilities of Typical Scenarios 1–4, are 0.383, 0.209, 0.195, and 0.213, respectively. The four typical renewable energy output scenarios are shown in Figure 4 below.

As shown in Figure 4, the quarterly average curves of typical wind power and photovoltaic output indicate that the four typical wind power scenarios in this region effectively describe the volatility of renewable energy output over time, encompassing temporal characteristics such as low output, high output, anti-peaking, and peaking. Among the four typical wind power scenarios after clustering, the variation trends of each typical output curve are significantly different. In the daily typical curves, the difference between the maximum and minimum normalized values of output on the same day can reach 0.42, indicating significant volatility. This demonstrates the rationality of the improved typical scenario processing method in describing the uncertainty of wind power.

The typical photovoltaic scenarios mainly describe the changes in daily peak output, with the overall curves showing a stepped distribution at different heights, indicating good reduction effects. This proves that the improved typical scenario processing method can reasonably describe representative photovoltaic output scenarios. Through correlation analysis of the typical scenarios, it can be seen that during early morning or nighttime when there is no sunlight, the wind power output can compensate for the lack of photovoltaic power, exhibiting good complementary characteristics. This effectively proves the scientific nature of the designed typical scenario processing method for modeling and screening combined wind–photovoltaic scenarios.

On the other hand, to further validate the effectiveness of the scenario selection method proposed in this paper, this section takes the quarterly output data as the analysis object and conducts a comparative analysis between the traditional K-means method (Method 1) [14], SOM-K-means method (Method 2) [24], entropy-weighted K-means method (Method 3), and the method proposed in this paper. A comprehensive comparison is made of the quality of the results obtained by each method. The effectiveness and quality of scenario selection are comprehensively judged by the $I_{DBI}$ and $I_{CHI}$ indicators mentioned above, and the relevant indicator values are shown in

Table 2. Average forecasting accuracy under different scenarios.

Index	Method 1	Method 2	Method 3	The Proposed Method
$I_{DBI}$	21.36	22.35	24.12	26.28
$I_{CHI}$	1.75	1.69	1.65	1.59

below.

Combining the two clustering effect indicators under the various methods mentioned above, a quantitative analysis of the obtained results is performed. Among the typical scenarios obtained by the method proposed in this paper, $I_{DBI}$ has the largest value, while $I_{CHI}$ has the smallest value. Considering that a larger $I_{DBI}$ indicates a better clustering performance, and a smaller $I_{CHI}$ indicator value also indicates a better clustering performance, the method proposed in this paper demonstrates a certain degree of effectiveness compared to the others. Data in power systems are often characterized by high dimensionality, nonlinearity, and complexity. Through the adversarial training process of its generator and discriminator, GANs are able to capture the nonlinear relationships within the data and generate new samples that embody these nonlinear features. In contrast, traditional clustering methods such as K-means may have certain limitations when dealing with nonlinear data. GANs automatically learn the intrinsic features of the data during the training process, which are of great significance for subsequent clustering analysis. With samples generated by a GAN, clustering algorithms can more easily identify key features and patterns in the data, thereby enhancing the clustering effect. K-means, a partitioning-based clustering method, relies on Euclidean distances between data points to calculate similarity. However, K-means is sensitive to the selection of initial cluster centers and susceptible to noise and outliers. Furthermore, it assumes that all data points are distributed spherically around the cluster centers, which may not hold true in power systems. The combination of SOM and K-means can improve clustering performance to a certain extent, as SOM can preserve certain data features through its topological structure. Nevertheless, SOM-K-means is still subject to the limitations of the K-means algorithm itself, such as the selection of initial cluster centers and the spherical distribution assumption. Entropy-weighted K-means improves the performance of K-means by introducing entropy weights, making the clustering results more stable. However, this method still relies on the basic framework of K-means and may have limitations when dealing with nonlinear data and complex distributions. Samples generated by a GAN exhibit a tighter distribution and higher separation degree, resulting in lower DBI values when traditional clustering methods are applied for subsequent analysis.

4.3. Optimality Verification of the Scheduling Model

To demonstrate the effectiveness of the scheduling model more intuitively, a typical scenario is selected and the detailed wind power and photovoltaic data are shown in Figure 5 below. Figure 6 below illustrates the charging and discharging power of a typical daily user for a cloud-based energy storage device. It can be observed that users tend to charge their devices during periods of low electricity prices and low load demand, while discharging them during times of high electricity prices and peak load demand. This strategy not only minimizes their electricity costs but also helps to flatten the load curve by reducing peak loads and filling in valleys, thereby optimizing overall energy usage.

Table 3. Comparison of optimization objectives of different models.

Cost/USD	Without Energy Storage	Traditional Energy Storage	Cloud Energy Storage
Electricity purchase costs	24,172	21,586	20,482
Compensation income from the grid	0	756	1425
Total cost	24,172	20,830	19,057

shows the comparison of electricity purchase costs before and after a user participates in cloud energy storage services. Due to the incentives of real-time electricity prices and the rapid load response capability of energy storage devices, users are able to modify their own load curves. After optimization, the user’s electricity purchase costs are significantly reduced, and they also receive a certain amount of compensation income from the grid for the electricity repurchased.

Additionally, we compared it with a traditional energy storage model (where the charging and discharging performance of the energy storage devices is the same as the cloud energy storage model, and the total energy storage capacity and upper limit of electricity for all devices are also 2630 kW and 4550 kWh, respectively). As can be seen from Table 3, the electricity purchase cost for the traditional energy storage model is higher than that of the cloud energy storage model. Although the sum of the energy storage capacities and electricity amounts owned by all users is the same as in the cloud energy storage model, since the energy storage devices are individually owned by each user and cannot be shared with others, there are instances where some users have idle energy storage capacities while others lack sufficient capacity. This demonstrates the more flexible and effective advantages of the cloud energy storage model in terms of capacity allocation.

Furthermore, the original load curve had a peak-to-trough difference of 2934 kW. In the absence of energy storage devices, the integration of renewable energy would expand this difference to 3206 kW, while still resulting in 926 kWh of renewable energy that could not be consumed [30]. The traditional energy storage model can only partially absorb renewable energy, with a net load peak-to-trough difference of 2568 kW. In contrast, the cloud energy storage model can fully absorb renewable energy, reducing the net load peak-to-trough difference to 1759 kW, a decrease of 809 kW. This alleviates the system’s peak shaving pressure. Therefore, whether from the perspective of user economic benefits or grid operation stability, the cloud energy storage model is superior to the traditional energy storage model.

5. Conclusions

This study proposes an effective renewable energy consumption strategy that integrates scenario reduction and flexible resource utilization. By modeling uncertainty and leveraging small-capacity energy storage, the strategy enhances renewable energy utilization while stabilizing the power system. Case studies demonstrate significant improvements in key indicators, underscoring the strategy’s validity, feasibility, and potential to substantially reduce peak-to-trough differences. Upon evaluation, the introduced scenario creation approach attained a peak score of 26.28 for $I_{DBI}$ and a nadir score of 1.59 for $I_{CHI}$. Building upon this framework, the cloud-based energy storage model demonstrates a comprehensive capacity to assimilate renewable energy sources, significantly narrowing the peak-to-trough discrepancy in net load to 1759 kW, marking a reduction of 809 kW compared with conventional models. In the future, with the continuous advancement of renewable energy technologies and the intelligent upgrading of power systems, the research findings of this paper are expected to gain even wider application and promotion.