The Analysis of Energy Recovered during the Braking of an Electric Vehicle in Different Driving Conditions

Abstract

The partial recovery of kinetic energy during braking allows the vehicle’s battery to be additionally charged and thus extends the range of an electric vehicle. Because of the different operating strategies of the braking energy recovery system, it is important to understand the factors influencing the level of recovered energy. The driving conditions at the place of use have a direct impact on the energy efficiency of an electric vehicle. The purpose of this paper was to analyze the energy recovered during braking in different driving conditions. The tests were based on the parameters of actual trips made along urban and suburban routes, and express roads. The collected actual speed profiles were used for the simulation studies. AVL cruise vehicle simulation software was used in the study. Simulation tests revealed that the levels of energy recovered during braking in an electric vehicle were the highest in urban conditions. The amount of energy recovered during urban driving can account for 20% of the total trip energy. In driving conditions characterized by different intensities caused by trips at different times of the day, similar values of recovered energy were recorded. When driving in the afternoon hours, the level of recovered energy per 1 km was about 2% lower than when driving in rush hour conditions. From the results presented in this paper, it can be concluded that driving conditions have an impact on the level of recovered energy. The type of road on which the electric vehicle drives is particularly important.

1. Introduction

Electric vehicles are charged with electricity from a power grid at specially dedicated charging stations. The charging method is an important research issue that is the subject of many studies [1,2,3]. The possibility of recovering energy while driving can significantly improve the energy efficiency of electric vehicles and at the same time increase their range. Electric-powered vehicles are equipped with a system called regenerative braking. Its purpose is to transform kinetic energy into electric power. When using the braking system, the electric machine acts as a generator that changes the direction of the torque and energy flow. The transfer of energy to the battery is controlled by the battery management system (BMS) [4]. The intensity of regenerative braking is connected with many different factors that can reduce the amount of recovered energy. These include battery-related parameters, such as the state of charge (SOC), temperature, and charging rate. Other important factors include engine power, speed, and torque, as well as its general efficiency [5]. The regenerative braking algorithm is based on signals from the measuring sensors (brake pedal position, speed of vehicle, and angular position derivatives of the pedals), while the intensity of the electrical braking depends on real-time calculations performed by the electrical system that controls this process.

The efficiency of the entire energy-recuperating system elements (electric machine, energy converters, and power source) is mostly affected by the operational parameters assigned to a given vehicle, such as the maximum range of the vehicle, its hill-climbing ability, and its accelerating performance, which in turn affect battery operation and lifespan. The situations of frequent acceleration and braking that are typical in heavy urban traffic and generate a high battery current during both recharging and discharging significantly affect the battery’s lifespan, which is expressed as the number of recharging and discharging processes. They also affect the level of energy recovered during braking.

Because of the limited power of the regenerative braking system, a mechanical braking system is still required. The cooperation of a regenerative braking system with friction brakes in electric vehicles causes the achievement of the optimum performance of regenerative braking to become complex. One of the main problems is the management of the division of the braking force between the regenerative and mechanical braking systems in a way that would enable the recovery of the highest possible level of energy from regenerative braking. Another problem is the method of distribution of the braking force to the front and rear wheels, to ensure effective and safe braking. Therefore, the effectiveness of the braking system in electric and hybrid vehicles significantly depends on the control strategy of the mechanical brakes and the power recuperation system. Special algorithms for the strategy of controlling energy recuperation in an electric vehicle are supposed to find such a combination of regenerative and friction braking that ensures the recovery of the maximum amount of energy, but simultaneously provides sufficient braking power needed by the driver, as well as vehicle stability and passenger safety. Two main types of regenerative braking systems in electric vehicles can be identified: a parallel system and a serial system.

In parallel braking systems, braking forces on a vehicle’s axes are generated in parallel and simultaneously by both the electrical and mechanical sub-systems. The combined effect is calculated as the resultant force of the action of all individual forces. The parallel strategy does not require the mechanical part of the braking system to be electronically controlled [6]. The algorithms of serial braking strategies are much more complex and require the cooperation of electronic control systems of both the electrical and mechanical sub-systems. When low-intensity braking is required, the system may only use the electrical machine. The achievement of the maximum available power by the electrical generator activates the mechanical brake [7]. In the serial strategy, functions performed by the mechanical brake system have to act together with the ABS system. The authors of [8] compared three regenerative braking strategies on the basis of simulation tests using an electric vehicle. The results of these tests indicate that the implemented strategy significantly affected the energy recovery index.

Regenerative braking in electric and hybrid vehicles has been the subject of many tests and analyses. Many papers demonstrate specific models of braking energy distribution control strategies. The braking energy distribution control strategy proposed in [9] involves the division of energy between the electrical and mechanical braking systems. The first phase of the braking process engages the electrical machine, which draws energy in order to achieve maximum power. The mechanical brake is then engaged and acts in accordance with the control curve presented in regulations ECE 13-H. When the critical deceleration value of 7 m/s² is exceeded, the braking strategy goes into emergency mode.

The authors of [10] developed a strategy for the distribution of braking energy between the friction brakes and the electric machine, based on the average value of deceleration during the braking of the vehicle. As a result, not only the deceleration, but also the maximum power demand was reduced by two-thirds. The proposed algorithm was applied to the BMS system of the electric vehicle and testing using a Brake Dynamometer. The test results revealed that the regenerative braking strategy proposed by the authors could increase the effectiveness of energy recuperation by as much as 18%.

Many papers present fuzzy logic algorithms implemented in regenerative braking control strategies. In the context of factors that affect regenerative braking, input signals may include factors such as the braking power required by the driver, vehicle speed, acceleration, battery power level, or battery temperature. The output signal is usually implemented as the ratio of the regenerative braking force to the total braking force. In [11], the fuzzy logic approach was applied to model the braking force distribution in an electric vehicle. According to the authors, using the acceleration and jerk of the vehicle and the road inclination, as well as using the appropriate model of fuzzy inference, it was possible to determine the regeneration coefficient during braking, without taking into account the information about the brake pedal. The work of [12] presents a model with the use of fuzzy logic to obtain a constant braking torque of the electric motor. The input data to the fuzzy logic model were the following three factors: SOC, speed, and brake strength. The authors of [13] described a regenerative braking strategy that uses fuzzy logic algorithms, by verifying the relation between the SOC state and battery temperature, and the effectiveness of regenerative braking. It was revealed that the implementation of this strategy extended the vehicle’s range by 26%. A 22% increase in power efficiency was also observed. The results presented in [14] show that the control strategy using fuzzy logic could provide greater energy for regenerative braking and increase the overall efficiency of the braking system.

Many papers have also demonstrated the effectiveness of regenerative braking during various road maneuvers. Other papers have analyzed the level of recovered energy and the energy recovery index in the context of the entire energy used during braking in simulated maneuvers such as hard braking [15], braking at low speeds [16], braking with specified deceleration [17], braking in a curve [18], braking on surfaces with different coefficients of adhesion [19], and braking with different loads [20]. Conducting tests of the braking system with an energy recovery system in this type of maneuvering makes it possible to assess the level of energy recovery and the effectiveness of the entire system.

The amount of energy recovered during braking can be influenced by the driver’s driving style. The driving style is the way the driver handles the vehicle and the driving behavior they display. Each driver has their own driving style, which is manifested in the manner of accelerating, braking, operating the accelerator and brake pedal, or exceeding the permitted speed [21,22]. The authors of [23] analyzed the influence of the force applied to the brake pedal during the braking of an electric vehicle. The results of these tests revealed that the amount of recovered energy was higher when the pressure on the brake pedal was lower. The results described in [24] revealed that, depending on the driver, the differences in braking energy per km may even be six-fold.

The authors of the above-mentioned papers discussed issues such as the strategy of the distribution of energy between the electrical system and the mechanical brake system during braking, as well as the properties of the electrical drivetrain components. Many works also deal with the issue of the efficiency of the braking energy recovery system in various road maneuvers in real tests or in simulation tests. There are few works in the literature focusing on the value of energy recovered in the aspects of using electric vehicles. It is also worth examining the influence of the value of the recovered energy on the total energy consumption or the electric range.

The aim of this paper was to expand the knowledge on energy recovered during the braking of electric vehicles. We decided to analyze the issue of energy recovered during braking in trips with different driving conditions. This paper presents the results of simulation tests aiming to determine the value of the energy recovered during the braking of an electric vehicle on different types of routes (urban, suburban roads, and motorway) and in urban traffic during different times of the day (traffic density).

The recovered energy levels presented in this analysis are important from the electric vehicle user’s point of view, especially considering the issue of range anxiety. The presented results may allow electric vehicle users to properly plan the route due to the type of road, taking into account the fact that the battery can be recharged during braking.

The structure of the paper is as follows. Section 2 describes the simulation methodology used to obtain the recovered energy during braking in different driving conditions. First, the parameters of the analyzed electric vehicle are given. Next, the software used is presented. Lastly, the driving cycles used for the tests are described. Section 3 shows the results of the electric vehicle simulation, presenting an analysis of the energy used and recovered by the electric vehicle during various driving conditions. In Section 4, the relationship between acceleration and initial braking speed to the level of recovered energy is discussed, as well as the effect of road type on the value of the recovered energy. Finally, Section 5 presents the main conclusions regarding the analyses carried out in the study.

2. Simulation Study

Simulation studies were used to assess the value of the energy recovered during journeys on routes with different traffic conditions. First, actual tests were carried out to collect the velocity profiles and acceleration profiles. Then, after verifying the collected data, the speed profiles were implemented into the AVL Cruise simulation program. The selected electric vehicle model was adjusted to reflect the Nissan Leaf. As a result of the simulation calculations, the value of energy consumption, the value of recovered energy, the course of the energy level in the battery, the course of the current intensity in the battery were obtained. A flow chart of the study is shown in Figure 1.

2.1. Vehicle Parameters

The simulation tests were carried out with a front-wheel-drive electric vehicle. The mass of the vehicle was 1400 kg, the drag coefficient was 0.284, and the frontal area was 1.97 m². The vehicle’s parameters are shown in

Table 1. The parameters of the vehicle used for the simulation.

Drivetrain Component	Parameter	Unit	Value
Electric machine in motor mode	Maximum torque	Nm	280
	Maximum power	kW	80
	Speed at maximum power	1/min	7000
	Speed at maximum torque	1/min	0.0001
Electric machine in generator mode	Maximum torque	Nm	280
	Maximum power	kW	75
	Speed at maximum power	1/min	5250
	Speed at maximum torque	1/min	3000
Battery	Nominal voltage	V	320
	Maximum voltage	V	420
	Minimum voltage	V	220
	Initial charge	%	95
	Maximum charge	Ah	10
	Number of Cell–Rows	-	5

.

2.2. Software Description

The simulation tests were carried out using AVL Cruise software. The application of this software enabled the performance of a series of analyses of the vehicle’s parameters and elements during its operation. These tests can be performed on electric vehicles, as well as on conventionally-powered vehicles. Different types of input data can be introduced, such as speed, acceleration, travel time, or route distance, in order to obtain detailed results that can be presented in a numerical or graphical form. Simulation tests carried out using AVL Cruise software allow for the calculation of values such as energy or fuel consumption, power flow, energy consumption during movement, or the emission of harmful exhaust dust/gas compounds [25]. The software has an extensive library of various vehicle category models with different drivetrain configurations. The user can also develop and add his/her own vehicle models and apply input parameters such as the ambient temperature, the initial battery state of charge (SOC), and the elevation of terrain [26].

The energy consumption values obtained during the simulation tests can be used to calculate the efficiency of the analyzed vehicle in given road conditions, without the need to carry out performance tests [27].

2.3. Driving Cycles

The simulation tests were based on speed profiles registered during actual trips. They were registered in various traffic conditions. The dynamic parameters of the vehicle were measured using a Corrsys Datron S-350^® Aqua optoelectronic sensor and a three-axis TAA^® acceleration sensor, which cooperated with a Datron uEEP12^® data acquisition station and a control tablet equipped with ARMS^® software v. 2.75. The following values were measured: time, instantaneous speed, and instantaneous acceleration.

2.3.1. Speed Profiles for Various Daytimes

The first research issue was to analyze the impact of traffic conditions along the urban route at selected times of the day. The changing traffic volume affected the travel time of the test section, the average speed, and the number of stops. The trips took place at three different times of day:

• during the morning rush hour,
• at noon,
• during the afternoon rush hour.

In any city, it is possible to identify periods during which traffic congestion on the road’s increases. There are often morning and afternoon rush hours, as well as a specific period between these times. Therefore, in this study, we decided to use these three times of the day. From previous studies, it was observed that different traffic conditions affected the way in which the test route was travelled. There were visible differences in values such as average speed, stopping time, and the number of stops [28,29,30]. Detailed parameters for the selected routes are presented in

Table 2. The parameters of the trips made during different daytimes.

Parameters	Morning Rush Hour	at Noon	Afternoon Rush Hour
Trip time, s	2248	1805	3002
Average speed, km/h	18.1	22.5	13.6
Maximum speed	59.2	56.7	52.1
Number of stops	46	31	73
Proportional stoppage time, %	25	17	36

. The recorded speed profiles are shown in Figure 2.

The test route was 11.45 km long. Heavy traffic, especially in the city center in the morning and in the afternoon, is typical for driving in rush hour traffic conditions. The average speed during these trips was much lower than when driving at noon. The average speed of the vehicle during the morning rush hour was 18.1 km/h, while during the afternoon rush hour it was recorded at 13.6 km/h (Table 2). Traffic density when travelling at noon was low. The average speed recorded when traveling at noon was 22.5 km/h.

2.3.2. Speed Profiles for Different Road Types

The second research issue involved an analysis of the energy consumption of an electrical vehicle in different road conditions, depending on the type of road. The generated speed profiles included trips along urban streets, in suburban driving conditions, and on express roads. The analyzed trips were approximately 20 km long. Figure 3 shows the speed profiles assumed for the purpose of the simulation tests.

Each of the roads selected for testing had different driving characteristics.

Table 3. The parameters of trips along different road types.

	Urban Road	Suburban Road	Highway Road
Length of route, km	17.2	17.1	19.2
Trip time, s	2976	1064	704
Average speed, km/h	20.8	57.3	102.1
Maximum speed, km/h	67.0	84.1	122.8
Maximum acceleration, m/s2	1.8	2.9	5.0
Maximum deceleration, m/s2	3.2	3.0	2.6

shows the parameters of these trips.

There were significant differences in speed in the urban cycle. The intensity of traffic and the urban infrastructure necessitate frequent acceleration, braking, and stopping. The vehicle’s traveling speed was in the range between 10 and 30 km/h.

The vehicle traveled at much higher speeds on a suburban route than on an urban route. The trip was quite smooth and did not involve any stopping. However, there were numerous situations where the speed had to be reduced. In this cycle, the vehicle had a speed between 45 and 70 km/h over 75% of the entire route.

When traveling along express roads, the vehicle’s speed was typical for this type of road and was usually between 105 and 120 km/h. There were no stops during this trip.

3. Results

This section presents the results of simulations using the real speed profiles discussed earlier and compares selected factors characterizing energy efficiency and recovered energy rates under different traffic conditions.

3.1. Energy Analysis of an Electric Vehicle during Trips at Different Times of the Day

Each of the three trips carried out at different times of the day had different driving characteristics. This was mostly due to the increased traffic caused by the morning and afternoon traffic peaks. Figure 4 shows the consumption of energy by an electric vehicle per 1 km.

The highest energy consumption levels were registered when driving in the afternoon rush hour. Because of the high intensity of traffic, this was the longest trip, which was characterized by low speeds and a large proportion of stoppage time in the total trip time. The lowest energy consumption levels were noted when driving at noon. This was the smoothest trip and had the lowest number of stops. Energy consumption per one kilometer when driving in the afternoon rush hour was proportionally 8.5% higher than when driving at noon and 6.4% when driving in the morning. Figure 5 shows the profiles of the instantaneous values of the SOC and battery current.

The trend for SOC values during the analyzed trips was quite even. These results were also confirmed by the battery depth of discharge (DOD) registered during trips in the analyzed times of day. In the context of changes in SOC values, we can see that the highest drop in the state of charge of the battery was registered when driving in the afternoon rush hour (Figure 5c). This was due to the high number of stops and the need to drive at low speeds. The DOD index following the completion of trips in the morning and at noon had similar values of 9.9% and 9.7%, respectively, while for the afternoon trip it stood at 10.4%.

The trend of electric current values allowed for monitoring the battery operation during the trip. Negative electric current values mean that energy is drawn from the battery. Positive values mean that the battery is being charged using energy from regenerative braking. Figure 6 shows the values of the energy recovered during regenerative braking.

The values of energy recovered during the trips were similar. The highest value of recovered energy per 1 km was noted when driving in the afternoon rush hour. This was due to the high intensity of traffic and the frequent speed changes. The lowest value of recovered energy was registered when driving in the morning. The differences in the values of energy recovered on the same route but in different traffic conditions were small, ranging from 0.8–1.6%.

3.2. Energy Analysis of an Electric Vehicle during Trips in Different Road Types

The results of energy consumption for an electric vehicle in different road conditions are shown in Figure 7.

The electric vehicle registered the lowest consumption of energy per 1 km in suburban driving conditions. Energy consumption in an urban cycle was 4% higher than in a suburban cycle. The highest energy consumption level was registered on a highway. In this case, energy consumption was 44% higher than in urban conditions. The instantaneous SOC profile and battery current values are shown in Figure 8.

For each trip, the initial battery state of charge was 95%. Following trips in urban and suburban conditions, the power level in the battery was reduced by approximately 12%. In the case of the express road trip, the battery discharge level was 24%.

Urban infrastructure and traffic density necessitate frequent stopping and driving at low speeds. Repetitive acceleration and braking affect the functioning of the battery and the power consumption level, as well as the level of recovered energy. The instantaneous trend of battery current and SOC values shown in Figure 8 demonstrates that the charging of the battery with energy recovered during braking was as frequent as the drawing of energy from the battery. During most of the trip, energy from the battery was used to overcome resistance to motion while driving. In suburban conditions, the speed was reduced several times, but the vehicle did not come to a complete stop. There was an instance of hard braking toward the end of the trip. This was accompanied by regenerative braking, as demonstrated by the increase in the battery state of charge (SOC) and the positive current value.

Figure 9 shows the level of energy recovered per 1 km in the analyzed driving conditions.

As mentioned above, urban driving conditions provide more regenerative braking opportunities. Simulation tests carried out on routes of a similar length but in different driving conditions revealed that it was indeed in urban conditions that the electric vehicle experienced the highest value of recovered energy. In the case of suburban driving, the value of energy recovered during the trip was 68% lower than in the case of urban driving. In a cycle with a speed profile corresponding to highway driving, the level of recovered energy was 86% lower than in a cycle reflecting urban driving.

4. Discussion

This paper analyzed the values of energy consumption and the values of recovered energy during trips in different traffic conditions. The results of the simulation tests demonstrated that the type of road affected both the level of energy consumption by an electrical vehicle as well as the value of energy recovered during braking.

Figure 10 shows the value of the recovered energy in relation to the deceleration and initial braking speed during trips made at different times of the day.

As demonstrated by the data shown in Figure 10, the level of recovered energy at low speeds was also low. This was especially evident during trips in heavy traffic conditions (Figure 11c), with many cases of short braking at low speeds. The highest amounts of energy were recuperated during braking at high speeds. The value of deceleration did not affect the amount of recovered energy in the same way.

Figure 11 shows the share of energy used to overcome the movement resistance and the proportional use of recovered energy during trips in urban, suburban, and highway conditions.

The results of the simulations made using actual speed profiles that reflect driving in urban, suburban, and highway driving conditions reveal that the levels of energy recovered during braking were the highest in urban traffic conditions. The proportion of recovered energy to the total amount of energy used in urban driving was 20%. In a driving cycle with a speed profile that reflected express road driving conditions, the proportion of recovered energy to the total amount of energy was only 2%. The relative amount of energy recovered during urban driving was 86% higher than in the case of driving on an express road. This seems obvious due to the presence frequent stops as a result of the character of the road infrastructure and traffic density. Driving on an express road generally involves traveling at a constant speed, with a low number of braking and stopping events. The value of energy recovered in these conditions is thus low [31,32].

The authors of this paper have also analyzed the effect of traffic density on the amount of energy recovered when driving in urban conditions. The lowest average speed was registered during the afternoon rush hour (13.6 km/h). The highest average speed was noted at noon (22.5 km/h). The results of the simulation tests revealed that the differences in the amount of energy recovered during the trips were approximately 2%. Similar conclusions were presented in the paper [33]. Analyses of the energy recovered during braking in actual traffic conditions in Gdańsk (Poland) have shown that trips made at low average speeds ensure the highest proportional amount of specific regenerative braking energy. The authors of [34] concluded that the highest level of energy efficiency of an electric vehicle is ensured when its average speed is between 25 km/h and 45 km/h. In such cases, energy recovered during frequent braking can significantly affect the energy efficiency of the engine. Recharging the battery using this recovered energy can extend the vehicle’s range, whereas too much traffic and travelling in traffic jams reduces the effectiveness of braking at low speeds [35]. As confirmed by the results described in [11], the recovery of energy at speeds below 16 km/h is not effective due to the specific nature of the electric engine. This was confirmed by the results of the simulation tests carried out during this research.

5. Conclusions

The purpose of this paper was to analyze the levels of energy recovered during the braking of an electric vehicle in different driving conditions. In the first part of this research, the authors analyzed the amount of energy recovered from braking during trips made at different times of the day. The driving conditions and fluidity of movement were affected by the intensity of traffic at different times of the day. The authors of this research used speed profiles that reflected driving during the morning and afternoon rush hours, as well as driving at noon. The results of these simulation tests revealed that the highest value of recovered energy per 1 km was registered when driving during the afternoon rush hour. When driving in the afternoon hours, the level of recovered energy per 1 km was about 2% lower than when driving in rush hour conditions.

In the next part of the research, the authors analyzed the energy consumption of an electric vehicle during trips made in urban, suburban, and express road conditions. The results of these tests revealed that the proportion of recovered energy to the total amount of energy used in urban driving was 20%. The characteristics of urban driving enable the recovery of energy due to frequent braking or deceleration events. The amount of energy recovered by an electric vehicle per 1 km on an urban route was 68% higher than on a suburban route and 86% higher than the energy recovered on an express road.

The percentage of electrically powered vehicles in the number of all currently used vehicles is high. The results presented in this paper complement previous research into the efficiency of vehicle energy recovery systems. The paper presents detailed values of the energy recovered under different traffic conditions. These results are relevant from the point of view of electric vehicle users. The conditions in which electric vehicles are used may affect energy consumption levels. Urban driving, due to frequent braking events, enables the recharging of the battery with the energy recovered during braking. This reduces energy consumption and extends the driving range of a vehicle.

The paper presents the results of simulation tests carried out only on three types of roads and at three times of the day. The authors realize that comprehensive analyses of the operation of electric vehicles require closer examination. In future work, we plan to extend the research to more accurately determine the impact of various road conditions on the energy efficiency of an electric vehicle. In addition to simulation tests, we also plan to use real electric vehicles.