Extended-range automobile energy management method combining working condition recognition and driving style

Abstract

The invention discloses a range-extending type automobile energy management method combining working condition recognition and a driving style, and relates to the technical field of range-extending type electric automobile energy management. Through an intelligent energy management strategy, the starting and output power of the range extender is dynamically adjusted according to the actual driving working condition and the driving style, invalid energy consumption is reduced, and the energy utilization efficiency of the whole vehicle is improved; the optimized energy management strategy can better respond to the intention of a driver, smooth and sensitive-response acceleration performance is provided, and meanwhile, the consistency of energy management and a driving mode is kept in an intelligent driving mode, so that the driving experience is enhanced; the optimized energy management strategy is beneficial to reducing the running cost of the whole vehicle, and maximization of economic benefits is achieved by reducing fuel consumption and battery replacement frequency; and moreover, the provided energy management system can adapt to different driving working conditions and driving modes, and efficient energy management can be realized no matter in urban congested road sections or expressways.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management of extended-range electric vehicles, and in particular to an energy management method for extended-range electric vehicles that combines working condition recognition with driving style. Background Art

[0002] An extended range electric vehicle (EREV) is a hybrid electric vehicle whose main power source is an electric motor. The vehicle is powered by a battery and is equipped with a small internal combustion engine as a range extender. Its main function is to charge the battery, thereby extending the vehicle's range. Unlike traditional hybrid vehicles, the internal combustion engine of an extended range electric vehicle does not directly drive the wheels, but charges the battery by generating electricity to ensure that the vehicle can continue to drive when the battery is low. This design can overcome the range anxiety of pure electric vehicles during long-distance driving while ensuring zero emissions and low noise;

[0003] The Energy Management System (EMS) of an extended-range electric vehicle plays a vital role in optimizing the vehicle's energy efficiency, extending its range, and balancing the use of the battery and the internal combustion engine (or range extender). Its core goal is to intelligently manage the battery's state of charge and the operation of the internal combustion engine according to different driving conditions, so as to minimize energy consumption and maximize battery life while ensuring driving performance.

[0004] However, the following technical problems still exist:

[0005] In terms of energy management, existing extended-range electric vehicles mainly rely on fixed parameters to control the start-up of the range extender and are unable to make intelligent adjustments based on actual operating conditions.

[0006] Existing extended-range electric vehicles lack consideration of driving intentions and driving environment in their range extender starting strategies, resulting in low energy utilization.

[0007] Existing extended-range electric vehicles have problems such as large power fluctuations and high energy consumption in terms of range extender output power control.

[0008] The existing energy management strategies for extended-range electric vehicles cannot effectively identify complex operating conditions, resulting in poor energy management effects.

[0009] In the intelligent driving mode of existing extended-range electric vehicles, the energy management strategy is out of touch with the driving mode, affecting the driving experience.

[0010] The existing energy management strategies for extended-range electric vehicles do not adequately protect battery life and can easily lead to degradation of battery performance.

[0011] The existing energy management strategy of extended-range electric vehicles cannot achieve real-time and accurate energy distribution, which affects the vehicle's range.

[0012] The existing energy management strategies of extended-range electric vehicles have poor adaptability when dealing with different operating conditions and are difficult to meet diverse needs.

[0013] Therefore, a new solution to the above problems needs to be proposed. Summary of the invention

[0014] The purpose of the present invention is to provide an energy management method for an extended-range vehicle that combines working condition recognition and driving style, which can be more based on the driving conditions and driver style identified by the working condition and the intelligent adjustment working point of the range extender under intelligent driving conditions, so as to achieve high efficiency and low energy consumption, so as to solve the technical problems raised in the background technology.

[0015] To achieve the above object, the present invention provides the following technical solution: a method for energy management of an extended-range vehicle combining working condition identification and driving style, comprising at least the following steps:

[0016] S1: Training working condition: collect working condition data and divide the working condition data based on LSTM neural network and deep learning neural network model. After division, use the trained deep learning neural network model for actual working condition, and finally determine the optimal working point of the range extender;

[0017] S2: Based on the training condition calculation of S1, further application is carried out in actual conditions to achieve actual extended-range vehicle energy management.

[0018] Furthermore, the S1 at least comprises the following steps:

[0019] S100: Collecting driving condition data, which is obtained by collecting signals from sensors on the test vehicle, wherein the driving condition data at least includes vehicle speed per unit time, power battery SOC, driver's accelerator pedal opening, brake signal, steering wheel angle signal and intelligent driving signal, and is then used for training recognition of driving conditions and subsequently for driver style classification;

[0020] S101: kinematically dividing the working condition data through an LSTM neural network, inputting a segment of speed under multiple unit times, and outputting a driving state, wherein the driving state at least includes starting, acceleration, deceleration, constant speed, and idling, and then classifying the working condition through a deep learning neural network, inputting the kinematically divided driving state and working condition data into a deep learning neural network model, and then dividing the working condition data again to obtain an environmental working condition division, wherein the environmental working condition division includes at least an urban working condition, a suburban working condition, and a high-speed working condition;

[0021] S102: inputting the working condition category data classified in S101, as well as the driving data, the driver's accelerator pedal opening, the steering wheel angle signal and the driver's brake signal into the deep learning neural network model, and then classifying the driver's driving style, wherein the driver's driving style is divided into an aggressive style, a normal style and a conservative style. The deep learning neural network model in S102 is used for actual working conditions;

[0022] S103: Determine the operating point of the range extender;

[0023] S104: Export and generate an output operating point mode diagram after the range extender is turned on according to requirements, for viewing and later correction.

[0024] Furthermore, the S103 at least includes the following steps:

[0025] First, the required power of the car in the training data is calculated. The required power of the car is calculated by the kinematic division of the working conditions and the environmental division of the working conditions. The calculation formula is as follows:

[0026]

[0027] Where: P e is the required power, η is the transmission efficiency, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance, A is the frontal area of ​​the car, V is the speed of the car, P a consumes power for other accessories;

[0028] Secondly, the required power is corrected, and the calculation formula is as follows:

[0029] P=γ0P E

[0030] Where: P is the corrected required power, γ0 is the correction coefficient of the required power according to the working conditions;

[0031] Finally, based on the battery SOC, operating condition category and required power P range, three groups of range extender output power points were developed through machine learning, with four optimal operating points in each group.

[0032] Furthermore, the S2 at least includes the following steps:

[0033] S201: Determine the actual operating condition of the extended-range electric vehicle and obtain the actual operating condition;

[0034] S202: Determine the driver style, and determine the driver style of the driver according to the deep learning neural network trained in S102;

[0035] After the driver's style is determined, the power required by the car is solved and corrected using the following formula:

[0036]

[0037] P=γ2P E2 =γ0P E

[0038] P=γ3P E3 =γ0P E

[0039] P=γ4P E4 =γ0P E

[0040] Where: P Ei is the required power, β is the driver's accelerator pedal opening or the loading signal size during intelligent driving, n is the motor speed, T MAX is the maximum torque of the motor at the current motor speed, η E is the motor efficiency, P a is the power consumed by other accessories, P is the corrected required power, γ2 is the correction factor for the required power calculation in the case of aggressive style, γ0 is the correction factor for the required power calculated according to the working conditions, γ3 is the correction factor for the required power calculation in the case of ordinary style, γ3 is the correction factor for the required power calculation in the case of conservative style, and P E2 is the power required for the vehicle in an aggressive style, P E3 P is the power required for the vehicle under normal vehicle style conditions; E4 is the power required for the vehicle under conservative style, P E is the reference power calculated according to the working conditions;

[0041] S203: Intelligent driving, the power required by the car is the same as S202, which is P Ei , and its correction formula is as follows:

[0042] P=γ1P E1

[0043] Where: P is the corrected required power, γ1 is the correction coefficient for the required power calculation under intelligent driving conditions;

[0044] S204: Determine the current SOC of the power battery;

[0045] S205: Determine which working condition is currently in use, by using the deep learning neural network trained in S101;

[0046] S206: Calculate the current required power, wherein the process of calculating the current required power adopts the power calculation formula in S202 and S203;

[0047] S207: Determine whether to perform intelligent driving, and if so, execute S209; if not, execute S208;

[0048] S208: Correcting the required power in S206 according to the correction formula in S202;

[0049] S209: Correcting the required power in S206 according to the correction formula in S203;

[0050] S210: The three groups of working points determined in S103 are determined in S205 to determine which group of working points they belong to. The working point of the current range extender output is found by using the power battery SOC value and the corrected required power value P.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] 1. The present invention uses an intelligent energy management strategy to dynamically adjust the start-up and output power of the range extender according to the actual driving conditions and driving style, thereby reducing ineffective energy consumption and improving the energy utilization efficiency of the entire vehicle;

[0053] 2. The optimized energy management strategy of the present invention can better respond to the driver's intention, provide smooth and responsive acceleration performance, and maintain the consistency of energy management and driving mode in intelligent driving mode, thereby enhancing the driving experience;

[0054] 3. The optimized energy management strategy of the present invention helps to reduce the operating cost of the vehicle and maximize economic benefits by reducing fuel consumption and battery replacement frequency;

[0055] 4. The invention can adapt to different driving conditions and driving modes, and can achieve efficient energy management whether it is a congested urban road section or a highway;

[0056] 5. The present invention ensures efficient energy distribution between the battery and the range extender through intelligent energy management, thereby improving the range of electric vehicles, reducing the number of charging times, and increasing travel convenience for users;

[0057] 6. The energy management strategy proposed in the present invention can be integrated into the existing EREV control system without large-scale changes to the vehicle structure, which is easy to implement and promote. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0059] Figure 1 The LSTM neural network of the present invention performs kinematic division on the driving condition and divides it into a schematic diagram of driving states such as starting, acceleration, deceleration, constant speed, and idling;

[0060] Figure 2 A schematic diagram of the deep learning neural network model of the present invention for classifying driving conditions into urban conditions, suburban conditions, and high-speed conditions;

[0061] Figure 3 The deep learning neural network model of the present invention classifies the driver's style into aggressive, normal, and conservative;

[0062] Figure 4 A schematic diagram of optimizing and solving the range extender operating point using the machine learning model of the present invention;

[0063] Figure 5 This is a diagram of the output working point mode after the range extender of the present invention is turned on;

[0064] Figure 6 This is a flow chart of the energy management strategy of the extended-range electric vehicle of the present invention under working condition identification including intelligent driving. DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0066] See also Figure 1 - Figure 6 , a method for energy management of an extended-range vehicle combining working condition recognition and driving style, comprising at least the following steps:

[0067] S1: Training working condition: collect working condition data and divide the working condition data based on LSTM neural network and deep learning neural network model. After division, use the trained deep learning neural network model for actual working condition, and finally determine the optimal working point of the range extender;

[0068] S2: Based on the training condition calculation of S1, further application is carried out in actual conditions to achieve actual extended-range vehicle energy management.

[0069] S1 includes at least the following steps:

[0070] S100: Collecting driving condition data, which is obtained by collecting signals from sensors on the test vehicle. The driving condition data at least includes vehicle speed per unit time, power battery SOC, driver's accelerator pedal opening, brake signal, steering wheel angle signal and intelligent driving signal, and is then used for training recognition of driving conditions and subsequently for driver style classification;

[0071] S101: Kinematic division of working condition data is performed through LSTM neural network. The training model diagram is as follows: Figure 1 , the input is a fragment of speed under multiple unit time, and the output is the driving state, which at least includes starting, acceleration, deceleration, constant speed and idling. The deep learning neural network is used to classify its working conditions. The training model is Figure 2 , input the kinematically divided driving state and working condition data into the deep learning neural network model, and then divide the working condition data again to obtain the environmental working condition division, which at least includes urban working condition, suburban working condition and high-speed working condition;

[0072] S102: Input the working condition category data classified in S101, as well as the driving data, the driver's accelerator pedal opening, the steering wheel angle signal and the driver's brake signal into the deep learning neural network model, and the training model is as follows: Figure 3 As shown, the driver's driving style is further divided into aggressive style, ordinary style and conservative style. The deep learning neural network model in S102 is used in actual working conditions;

[0073] S103: Determine the operating point of the range extender. The determination method is as follows: Figure 4 ;

[0074] S104: Generate an output working point mode diagram after the range extender is turned on according to the requirements, for viewing and later correction. Figure 5 Shown are a set of range extender output power points.

[0075] S103 at least includes the following steps:

[0076] First, the required power of the car in the training data is calculated. The required power of the car is calculated by the kinematic division of the working conditions and the environmental division of the working conditions. The calculation formula is as follows:

[0077]

[0078] Where: P e is the required power, η is the transmission efficiency, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance, A is the frontal area of ​​the car, V is the speed of the car, P a consumes power for other accessories;

[0079] Secondly, the required power is corrected, and the calculation formula is as follows:

[0080] P=γ0P E

[0081] Where: P is the corrected required power, γ0 is the correction coefficient of the required power according to the working conditions;

[0082] Finally, based on the battery SOC, operating condition category and required power P range, three groups of range extender output power points were developed through machine learning, with four optimal operating points in each group.

[0083] S2 includes at least the following steps:

[0084] S201: Determine the actual operating condition of the extended-range electric vehicle and obtain the actual operating condition;

[0085] S202: Determine the driver style, and determine the driver style of the driver according to the deep learning neural network trained in S102;

[0086] After the driver's style is determined, the power required by the car is solved and corrected using the following formula:

[0087]

[0088] P=γ2P E2 =γ0P E

[0089] P=γ3P E3 =γ0P E

[0090] P=γ4P E4 =γ0P E

[0091] Where: P Ei is the required power, β is the driver's accelerator pedal opening or the loading signal size during intelligent driving, n is the motor speed, T MAX is the maximum torque of the motor at the current motor speed, η E is the motor efficiency, P a is the power consumed by other accessories, P is the corrected required power, γ2 is the correction factor for the required power calculation in the case of aggressive style, γ0 is the correction factor for the required power calculated according to the working conditions, γ3 is the correction factor for the required power calculation in the case of ordinary style, γ3 is the correction factor for the required power calculation in the case of conservative style, and P E2 is the power required for the vehicle in an aggressive style, P E3 P is the power required for the vehicle under normal vehicle style conditions;E4 is the power required for the vehicle under conservative style, P E is the reference power calculated according to the working conditions;

[0092] S203: Intelligent driving, the power required by the car is the same as S202, which is P Ei , and its correction formula is as follows:

[0093] P=γ1P E1

[0094] Where: P is the corrected required power, γ1 is the correction coefficient for the required power calculation under intelligent driving conditions;

[0095] S204: Determine the current SOC of the power battery;

[0096] S205: Determine which working condition is currently in use, by using the deep learning neural network trained in S101;

[0097] S206: Calculate the current required power. The process of the current required power adopts the power calculation formula in S202 and S203;

[0098] S207: Determine whether to perform intelligent driving, and if so, execute S209; if not, execute S208;

[0099] S208: Correcting the required power in S206 according to the correction formula in S202;

[0100] S209: Correcting the required power in S206 according to the correction formula in S203;

[0101] S210: The three groups of working points determined in S103 are determined in S205 to determine which group of working points they belong to. The working point of the current range extender output is found by using the power battery SOC value and the corrected required power value P.

[0102] The above energy management strategy identifies the working condition, driver style, and required power, and then determines the optimal working point of the range extender. Finally, the range extender is operated at the optimal working point in the actual working condition. It can allocate power sources more accurately, optimize the use of the range extender and battery, and thus improve the energy efficiency of the whole vehicle. It can adapt to different driving conditions and driving modes, and can achieve efficient energy management whether it is a congested urban road or a highway. According to the actual driving conditions and driving style, the start-up and output power of the range extender are dynamically adjusted to reduce ineffective energy consumption and improve the energy efficiency of the whole vehicle.

[0103] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for energy management of an extended-range vehicle combining working condition recognition and driving style, characterized in that: At least the following steps are included: S1: Training working condition: collect working condition data and divide the working condition data based on LSTM neural network and deep learning neural network model. After division, use the trained deep learning neural network model for actual working condition, and finally determine the optimal working point of the range extender; S2: Based on the training condition calculation of S1, further application is carried out in actual conditions to achieve actual extended-range vehicle energy management.

2. The method for energy management of an extended-range vehicle combining working condition identification and driving style according to claim 1, characterized in that: The S1 at least comprises the following steps: S100: Collecting driving condition data, which is obtained by collecting signals from sensors on the test vehicle, wherein the driving condition data at least includes vehicle speed per unit time, power battery SOC, driver's accelerator pedal opening, brake signal, steering wheel angle signal and intelligent driving signal, and is then used for training recognition of driving conditions and subsequently for driver style classification; S101: kinematically dividing the working condition data through an LSTM neural network, inputting a segment of speed under multiple unit times, and outputting a driving state, wherein the driving state at least includes starting, acceleration, deceleration, constant speed, and idling, and then classifying the working condition through a deep learning neural network, inputting the kinematically divided driving state and working condition data into a deep learning neural network model, and then dividing the working condition data again to obtain an environmental working condition division, wherein the environmental working condition division includes at least an urban working condition, a suburban working condition, and a high-speed working condition; S102: inputting the working condition category data classified in S101, as well as the driving data, the driver's accelerator pedal opening, the steering wheel angle signal and the driver's brake signal into the deep learning neural network model, and then classifying the driver's driving style, wherein the driver's driving style is divided into an aggressive style, a normal style and a conservative style. The deep learning neural network model in S102 is used for actual working conditions; S103: Determine the operating point of the range extender; S104: Export and generate an output operating point mode diagram after the range extender is turned on according to requirements, for viewing and later correction.

3. The method for energy management of an extended-range vehicle combining working condition identification and driving style according to claim 2, characterized in that: The S103 at least includes the following steps: First, the required power of the car in the training data is calculated. The required power of the car is calculated by the kinematic division of the working conditions and the environmental division of the working conditions. The calculation formula is as follows: Where: P e is the required power, η is the transmission efficiency, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance, A is the frontal area of ​​the car, V is the speed of the car, P a consumes power for other accessories; Secondly, the required power is corrected, and the calculation formula is as follows: P=γ0P E Where: P is the corrected required power, γ0 is the correction coefficient of the required power according to the working conditions; Finally, based on the battery SOC, operating condition category and required power P range, three groups of range extender output power points were developed through machine learning, with four optimal operating points in each group.

4. The method for energy management of an extended-range vehicle combining working condition identification and driving style according to claim 3, characterized in that: The S2 at least comprises the following steps: S201: Determine the actual operating condition of the extended-range electric vehicle and obtain the actual operating condition; S202: Determine the driver style, and determine the driver style of the driver according to the deep learning neural network trained in S102; After the driver's style is determined, the power required by the car is solved and corrected using the following formula: P=γ2P E2 =γ0P E P=γ3P E3 =γ0P E P=γ4P E4 =γ0P E Where: P Ei is the required power, β is the driver's accelerator pedal opening or the loading signal size during intelligent driving, n is the motor speed, T MAX is the maximum torque of the motor at the current motor speed, η E is the motor efficiency, P a is the power consumed by other accessories, P is the corrected required power, γ2 is the correction factor for the required power calculation in the case of aggressive style, γ0 is the correction factor for the required power calculated according to the working conditions, γ3 is the correction factor for the required power calculation in the case of ordinary style, γ3 is the correction factor for the required power calculation in the case of conservative style, and P E2 is the power required for the vehicle in an aggressive style, P E3 P is the power required for the vehicle under normal vehicle conditions; E4 is the power required for the vehicle under conservative style, P E is the reference power calculated according to the working conditions; S203: Intelligent driving, the power required by the car is the same as S202, which is P Ei , and its correction formula is as follows: P=γ1P E1 Where: P is the corrected required power, γ1 is the correction coefficient for the required power calculation under intelligent driving conditions; S204: Determine the current SOC of the power battery; S205: Determine which working condition is currently in use, by using the deep learning neural network trained in S101; S206: Calculate the current required power, wherein the process of calculating the current required power adopts the power calculation formula in S202 and S203; S207: Determine whether to perform intelligent driving, and if so, execute S209; if not, execute S208; S208: Correcting the required power in S206 according to the correction formula in S202; S209: Correcting the required power in S206 according to the correction formula in S203; S210: The three groups of working points determined in S103 are determined in S205 to determine which group of working points they belong to. The working point of the current range extender output is found by using the power battery SOC value and the corrected required power value P.

Images