An intelligent energy management method for electric vehicles

Through intelligent energy management methods, dynamic programming and random model prediction algorithms are used to optimize the control temperature of motors and batteries, and the problems of high energy consumption and insufficient range of electric vehicles are solved, and the optimal allocation and range of the energy flow path of the entire vehicle are achieved. extension.

CN118107410BActive Publication Date: 2024-12-03JILIN UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410440012.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-12
Publication Date
2024-12-03
Estimated Expiration
2044-04-12

AI Technical Summary

Technical Problem

The existing electric vehicle energy management strategy does not take into account actual driving conditions and real-time road conditions information, resulting in high energy consumption and insufficient range.

Method used

Using intelligent energy management methods, obtain real-time vehicle information and road conditions information by obtaining navigation information, online map data and intelligent networking technology, and optimize the target control temperature of the motor and battery by using dynamic programming algorithms and random model prediction algorithms to achieve the maximum energy flow path Excellent allocation.

Benefits of technology

Reduce the energy consumption of the whole vehicle, extend the vehicle's cruising range, and improve driving economy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118107410B_ABST
    Figure CN118107410B_ABST
Patent Text Reader

Abstract

The present invention discloses an intelligent energy management method for electric vehicles, comprising: obtaining navigation information of users, using online map data to obtain real-time vehicle information and real-time road condition information to predict driving conditions, using intelligent network technology to obtain constraint information in the driving path, identifying travel characteristic parameters, constructing global driving conditions, and obtaining speed-time curves; formulating upper-layer control strategies and lower-layer control strategies based on hierarchical energy management strategies; inputting navigation information and speed-time curves into the upper-layer control strategy, optimizing the target temperatures of motors and batteries from the perspective of global energy minimization using a dynamic programming algorithm, and determining the target control temperatures of the motors and batteries; optimizing and controlling energy distribution using a random model prediction algorithm, and implementing real-time online adjustment of the target control temperatures of motors and batteries through information fed back to the upper-layer control strategy by the lower-layer control strategy, thereby completing intelligent energy management of electric vehicles.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of automobile energy intelligent management, and in particular relates to an electric vehicle intelligent energy management method. Background Art

[0002] In recent years, energy shortages and environmental pollution have become increasingly severe. The high-quality development of new energy vehicles is of great significance to reducing dependence on oil, improving energy structure, and reducing environmental pollution. At present, most of the energy management strategies of electric vehicles do not take into account the actual driving conditions and real-time road conditions, resulting in high driving energy consumption and insufficient driving range. Therefore, in order to ensure that electric vehicles can more reasonably distribute energy flow and improve driving economy, it is urgent to propose an intelligent energy management method for electric vehicles. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes an intelligent energy management method for electric vehicles to achieve optimal allocation of the energy flow path of the whole vehicle, reduce the energy consumption of the whole vehicle, and extend the vehicle's cruising range.

[0004] To achieve the above object, the present invention provides an electric vehicle intelligent energy management method, comprising:

[0005] Obtain the user's navigation information, use online map data to obtain real-time vehicle information and real-time road condition information to predict driving conditions, use intelligent Internet of Things technology to obtain constraint information in the driving path, identify travel characteristic parameters, construct global driving conditions, and obtain speed time curves;

[0006] Formulate upper-layer control strategies and lower-layer control strategies based on hierarchical energy management strategies;

[0007] Inputting the navigation information and the speed-time curve into the upper-level control strategy, optimizing the target temperature of the motor and the battery from the perspective of global energy minimization using a dynamic programming algorithm, and determining the target control temperature of the motor and the battery;

[0008] The target control temperatures of the motor and battery are input into the lower-level control strategy, and the energy distribution is optimized and controlled using a random model prediction algorithm. The information fed back to the upper-level control strategy by the lower-level control strategy is used to achieve real-time online adjustment of the target control temperatures of the motor and battery, thereby completing intelligent energy management of electric vehicles.

[0009] According to the electric vehicle intelligent energy management method provided by the present invention, before obtaining the user's navigation information, it is necessary to construct a driving condition fragment library, including:

[0010] Based on the historical driving data of electric vehicles, new travel characteristic parameters are constructed through principal component analysis;

[0011] The fuzzy C-means clustering algorithm is used to construct the initial driving condition segment library. The kinematic segments containing a preset number of categories in the driving condition segment library are obtained through silhouette coefficient evaluation to complete the construction of the driving condition segment library.

[0012] According to the electric vehicle intelligent energy management method provided by the present invention, the vehicle information includes vehicle position information and vehicle speed information; the road condition information includes average traffic flow speed, road conditions, traffic light status information and altitude information.

[0013] According to the electric vehicle intelligent energy management method provided by the present invention, the target temperature of the motor and the battery is optimized from the perspective of global energy minimization using a dynamic programming algorithm, and the method for determining the target control temperature of the motor and the battery includes:

[0014] According to the current vehicle information and the road condition information, system state variables and system decision variables are selected; according to the system state variables and the system decision variables, a state displacement equation for the whole vehicle is constructed; a target cost function is established from the perspective of global energy minimization; constraints are established in the dynamic programming solution process; the optimal objective function and optimal decision variables of each stage are obtained through inverse optimization solution; according to the optimal objective function and the optimal decision variables, a global optimal state variable solution is obtained through forward optimization solution to determine the target control temperature of the motor and the battery.

[0015] According to the electric vehicle intelligent energy management method provided by the present invention, based on the target control temperature of the motor and the battery, the energy distribution is optimized and controlled by using a random model prediction algorithm, and the method for realizing real-time online adjustment of the target control temperature of the motor and the battery through information fed back to the upper control strategy by the lower control strategy includes:

[0016] Based on the target control temperature of the motor and the battery, the system state quantity, the system control quantity and the system disturbance quantity are selected; a cost function is constructed with the goal of minimizing the error between the motor and battery temperature and the target control temperature in the prediction time domain; constraints are set according to the system state quantity, the system control quantity and the system disturbance quantity; a dynamic programming algorithm is combined with multi-step instant differential learning to complete rolling optimization of the target control temperature of the motor and the battery; the control strategy is adjusted in real time through feedback correction, the driving energy consumption of the whole vehicle and the energy recovery efficiency of the vehicle are obtained by calculation and periodically fed back to the upper-level control strategy, so as to realize real-time online adjustment of the target control temperature of the motor and the battery.

[0017] According to the electric vehicle intelligent energy management method provided by the present invention, the method for obtaining the profile coefficient is:

[0018]

[0019] Among them, a(i) is the average distance between the i-th sample and other samples in the same cluster category, and b(i) is the minimum distance between the i-th sample and each sample in other cluster categories.

[0020] According to the electric vehicle intelligent energy management method provided by the present invention, the method for constructing the state displacement equation for the whole vehicle is:

[0021]

[0022] Among them, T mot is the motor temperature, T bat is the battery temperature, P mot (t) is the motor power required at time t, P bat (t) is the battery power demand at time t, P AC (t) is the cooling power of the compressor at time t, is the instantaneous rate of change of motor temperature, is the instantaneous rate of change of battery temperature.

[0023] According to the electric vehicle intelligent energy management method provided by the present invention, the target cost function is:

[0024]

[0025]

[0026]

[0027] Among them, J means the instantaneous cost function, x t It means the system state variable at time t, u t It means the system control variable at time t, L bat (t) is the instantaneous battery life decay cost at time t, L ele (t) is the instantaneous power consumption cost at time t, A h (t) is the cumulative battery consumption, A h_20% (t) is the maximum battery power consumption when the power battery capacity decays to 20%, σ bat is the total cost of the power battery pack, σ ele The cost of electricity consumption.

[0028] According to the electric vehicle intelligent energy management method provided by the present invention, the cost function is:

[0029] J(Z K )=J y (Z K )+J U (Z K )+J V (ZK )

[0030] Among them, J(Z K ) is the error between the motor and battery temperature and the target temperature in the predicted time domain, J y (Z K ) is the error caused by the calculation process of MPC optimization controller, J U (Z K ) is the error caused by the system control quantity, J V (Z K ) is the error caused by the system disturbance.

[0031] According to the electric vehicle intelligent energy management method provided by the present invention, the method for obtaining the driving energy consumption of the whole vehicle and the energy recovery efficiency of the vehicle by calculation is:

[0032]

[0033]

[0034] Among them, E car is the power system energy consumption, T1 is the front electric drive torque, T2 is the rear electric drive torque, n1 is the front electric drive speed, n2 is the rear electric drive speed, μ1 is the front electric drive efficiency, μ2 is the rear electric drive efficiency, is the energy recovery efficiency of the electric vehicle, Q is the cumulative discharge of the battery pack, and Q1 is the cumulative power feed to the drive motor.

[0035] Technical effect of the invention: The present invention discloses an intelligent energy management method for electric vehicles. First, the user's navigation planning path information is obtained, and online map data is used to obtain real-time location information and real-time road condition information. The intelligent network technology is used to obtain various constraint information such as environmental information and altitude information in the driving path. Based on the identification of travel characteristic parameters, the global driving condition of the vehicle travel is accurately constructed. Then, through a hierarchical energy management strategy, the target control temperature of the motor and the battery is first solved from the perspective of global energy minimization through a dynamic programming algorithm, and then the random model predictive control is used to achieve the optimal allocation of compressor power, refrigerant flow and energy flow path. The information fed back to the upper control strategy by the lower control strategy is used to achieve real-time online adjustment of the target temperature of the motor and the battery, and complete the real-time energy optimization management of the entire electric vehicle. The present invention ultimately achieves the optimal allocation of the energy flow path of the entire vehicle, reduces the energy consumption of the entire vehicle, and extends the vehicle's cruising range. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0037] Figure 1 A schematic diagram of a flow chart of an intelligent energy management method for an electric vehicle according to an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of a process for obtaining navigation information and predicting driving conditions according to an embodiment of the present invention;

[0039] Figure 3 A schematic diagram of a process for formulating an upper-layer control strategy according to an embodiment of the present invention;

[0040] Figure 4 A schematic diagram of the process of formulating the lower-level control strategy according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0042] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0043] like Figure 1 As shown, this embodiment provides an electric vehicle intelligent energy management method, including:

[0044] Obtain the user's navigation information, use online map data to obtain real-time vehicle information and real-time road condition information to predict driving conditions, use intelligent Internet of Things technology to obtain constraint information in the driving path, identify travel characteristic parameters, construct global driving conditions, and obtain speed time curves;

[0045] Formulate upper-layer control strategies and lower-layer control strategies based on hierarchical energy management strategies;

[0046] Input navigation information and speed-time curve into the upper-level control strategy, and use dynamic programming algorithm to optimize the target temperature of the motor and battery from the perspective of global energy minimization to determine the target control temperature of the motor and battery;

[0047] The target control temperature of the motor and battery is input into the lower-level control strategy, and the energy distribution is optimized and controlled using a random model prediction algorithm. The information fed back to the upper-level control strategy by the lower-level control strategy is used to achieve real-time online adjustment of the target control temperature of the motor and battery, thus completing the intelligent energy management of electric vehicles.

[0048] Vehicle information includes vehicle location information and vehicle speed information; road condition information includes average traffic flow speed, road conditions, traffic light status information and altitude information.

[0049] Furthermore, before obtaining the user's navigation information, it is necessary to build a driving condition fragment library, including:

[0050] Based on the historical driving data of electric vehicles, new travel characteristic parameters are constructed through principal component analysis;

[0051] The fuzzy C-means clustering algorithm is used to construct the initial driving condition segment library. The kinematic segments containing a preset number of categories in the driving condition segment library are obtained through silhouette coefficient evaluation to complete the construction of the driving condition segment library.

[0052] Specifically include: Figure 2 As shown in the figure, based on the historical driving data of electric vehicles, including information such as vehicle speed, longitude and latitude, the principal component analysis method is introduced into the three characteristic parameters of vehicle speed, acceleration and time to construct a set of new variables as travel characteristic parameters, which can not only reduce the time of identifying the parameters representing the travel segment, but also achieve the effect of reducing the dimension of the characteristic parameters.

[0053] The principal component analysis method needs to calculate the contribution of the i-th influencing factor and the total contribution of the first j influencing factors respectively. When the total contribution of the first j influencing factors is greater than 85%, the first j influencing factors can replace all the original parameters.

[0054] Among them, the contribution of the i-th impact factor is The total contribution of the first j impact factors is α i is the influence of each parameter of the ith order, α j is the total influence of the first j parameters, ε i is the non-negative characteristic root of the i-th influencing factor.

[0055] The initial cluster centers of the fuzzy C-means clustering algorithm need to be manually selected in advance, which is easy to fall into the local optimal solution. Therefore, based on the division of kinematic segments, the optimized fuzzy C-means clustering algorithm (OFCM) is used to classify the kinematic segments.

[0056] The objective function of iterative update of cluster centers in OFCM clustering algorithm is:

[0057]

[0058] Among them, the sample point x in the OFCM clustering algorithm i The degree of membership to the k-th cluster category is:

[0059]

[0060] Among them J OFCM is the result of iterative update of cluster centers, that is, the distance measure of each individual. is the sample point x i The membership degree of the k-th cluster category, and the sample point x i The sum of the membership of each cluster category is 1, n is the total number of sample points, that is, the maximum number of iterations of the algorithm, q is the total number of cluster centers, B is the weighting coefficient, and 1≤B≤∞, D ik is the sample point x i The Euclidean distance to the kth cluster center, d ij is the Euclidean distance between the i-th cluster center and the j-th cluster center.

[0061] The silhouette coefficient (SC) of the IFCM clustering algorithm was selected as the evaluation index of the clustering results, and the number of clusters when the SC was the highest was taken as the number of major categories of the kinematic segments.

[0062] The calculation formula of silhouette coefficient is:

[0063]

[0064] Where a(i) is the average distance between the i-th sample and other samples in the same cluster category. The smaller a(i) is, the more reasonable it is. b(i) is the minimum distance between the i-th sample and each sample in other cluster categories.

[0065] According to the historical travel data, the silhouette coefficient of the OFCM clustering algorithm is the highest when the number of clusters is 4, so the kinematic fragments based on the historical travel data are divided into four categories, namely, low-speed driving conditions with most vehicle speeds below 25km / h; medium-speed driving conditions with more idling and a speed of 25-60km / h, but a lower average speed; medium-speed driving conditions with less idling and a speed of 25-60km / h, but the average speed is higher than the previous category; high-speed driving conditions with most vehicle speeds greater than 60km / h, thus completing the construction of the driving condition fragment library.

[0066] Route planning is performed based on the travel information input by the driver into the navigation system, real-time location information and real-time road condition information are obtained using online map data, and various constraint information such as environmental information and altitude information in the driving route are obtained using intelligent Internet of Things technology. The construction of the global driving condition is completed based on the identification of travel characteristic parameters.

[0067] Furthermore, the target temperature of the motor and the battery is optimized from the perspective of global energy minimization using a dynamic programming algorithm. The method for determining the target control temperature of the motor and the battery includes:

[0068] According to the current vehicle information and road condition information, the system state variables and system decision variables are selected; the state displacement equation for the whole vehicle is constructed based on the system state variables and system decision variables; the target cost function is established from the perspective of global energy minimization; the constraints in the dynamic programming solution process are established; the optimal objective function and optimal decision variables in each stage are obtained through reverse optimization solution; based on the optimal objective function and optimal decision variables, the global optimal state variable solution is obtained through forward optimization solution to determine the target control temperature of the motor and battery.

[0069] Specifically include: Figure 3 As shown, in the upper-level control strategy in the hierarchical energy management strategy formulation stage, after the driving route is determined based on the travel information input into the navigation system, the target control temperature of the motor and battery is determined by optimizing from the perspective of global energy minimization through a dynamic programming algorithm based on information such as the current ambient temperature, mileage, expected maximum speed and average speed, and current battery temperature and battery charge status.

[0070] Select the motor temperature T mot and battery temperature T bat As the system state variable x t , select SOC and battery circuit cooling capacity Q bc As the system decision variable u t .

[0071] Establish the state transfer equation for the whole vehicle:

[0072]

[0073] Where P mot (t) is the motor power required at time t, P bat (t) is the battery power demand at time t, P AC (t) is the cooling power of the compressor at time t, is the instantaneous rate of change of motor temperature, is the instantaneous rate of change of battery temperature.

[0074] From the perspective of global energy minimization, the target cost function is established as the instantaneous battery life decay cost L bat (t) and instantaneous power consumption cost L ele The sum of (t), that is, the target cost function is:

[0075]

[0076] in

[0077] Among them, J means the instantaneous cost function, the sum of the two types of costs, x tIt means the system state variables (motor temperature and battery temperature) at time t, u t It means the system control variables (battery SOC and battery circuit cooling capacity) at time t, L bat (t) is the instantaneous battery life decay cost at time t, L ele (t) is the instantaneous power consumption cost at time t, A h (t) is the cumulative battery consumption, A h_20% (t) is the maximum battery power consumption when the power battery capacity decays to 20%, σ bat is the total cost of the power battery pack, σ ele is the electricity cost, in RMB per kilometer, where σ bat , σ ele are all constants.

[0078] Due to the limitations of the characteristics of each component of the system and the constraints of the vehicle power system, it is necessary to limit the value range of each variable, that is, the following constraints need to be met in the entire dynamic programming solution process:

[0079]

[0080] Reverse optimization, calculate the instantaneous cost of each state in the tth stage, and the optimal cost function value from each state in the tth stage to the terminal, and finally obtain the optimal decision amount for each state in the tth stage until t = 1, completing the reverse optimization process

[0081] The instantaneous cost of each state in stage t is is the i-th state variable in stage t, is the i-th decision variable in stage t, is the state x of stage t i Take u i The instantaneous cost incurred, The state of the t+1 stage The optimal cost function to the terminal has the recursive relationship: U t Take values ​​for the state variables in the entire space;

[0082] According to the optimal cost function Solve the corresponding optimal decision

[0083] The above process is repeated until the initial stage of the global optimization problem is reached, that is, stage t=1, completing the reverse optimization process of the dynamic programming algorithm.

[0084] Forward optimization: According to the optimal decision quantity at each moment output by the reverse optimization process, the optimal state quantity of the system at each stage is forward solved through the vehicle system state equation until t=t N, and finally output the global optimal state variable solution to complete the forward optimization process, that is, determine the target control temperature of the motor and battery, and realize the optimization process of the target control temperature.

[0085] Furthermore, the method of optimizing the energy distribution based on the target control temperature of the motor and the battery by using a stochastic model prediction algorithm and realizing the real-time online adjustment of the target control temperature of the motor and the battery by feeding back information from the lower control strategy to the upper control strategy includes:

[0086] Based on the target control temperature of the motor and battery, the system state quantity, system control quantity and system disturbance quantity are selected; a cost function is constructed with the goal of minimizing the error between the motor and battery temperature in the prediction time domain and the target control temperature; constraints are set according to the system state quantity, system control quantity and system disturbance quantity; a dynamic programming algorithm is combined with multi-step instant differential learning to complete rolling optimization of the target control temperature of the motor and battery; the control strategy is adjusted in real time through feedback correction, the driving energy consumption of the whole vehicle and the vehicle's energy recovery efficiency are obtained by calculation and periodically fed back to the upper-level control strategy, so as to realize real-time online adjustment of the target control temperature of the motor and battery.

[0087] Specifically include: Figure 4 As shown, in the lower-level control strategy in the hierarchical energy management strategy formulation stage, the random model predictive control follows the target temperature of the motor and battery solved by the upper-level controller, and uses this to optimally allocate the compressor power, refrigerant flow and energy flow path to achieve the purpose of energy optimization management.

[0088] A basic model predictive controller is designed, and the motor and battery target temperatures optimized by the upper controller are used as the follow-up targets of the MPC controller to optimally allocate the compressor power, refrigerant flow rate, and energy flow path. Therefore, the motor and battery temperatures are used as system state quantities, and the battery circuit cooling capacity Q is used as the cooling capacity of the battery circuit. bc As the system control quantity, the above is used as the input value of the MPC controller. The output value of the MPC controller is the optimal allocation of compressor power, refrigerant flow and energy flow path. At the same time, it outputs the motor and battery temperature trajectories in the predicted time domain to achieve energy optimization management.

[0089] The path information obtained from the navigation system, the real-time road conditions and environmental information obtained from the online map data, and the short-term vehicle driving condition information obtained from the working condition prediction are used as variables affecting the system, and the representative vehicle speed V is extracted. car and ambient temperature T env As the system disturbance, it is also input into the MPC controller to design a stochastic MPC controller (SMPC controller) to improve the robustness of the algorithm and the accuracy of the control system.

[0090] The system state quantity expression is X = [T mot_opt , T bat_opt ], the system control quantity expression is U=[Q bc ], the system disturbance expression is V = [V car , T env ], the system output value expression is Y = [A res ];

[0091] Among them, T mot_opt represents the motor target temperature obtained based on the upper control strategy mentioned above, T bat_opt represents the battery target temperature obtained based on the upper control strategy mentioned above, Q bc Indicates the cooling capacity of the battery circuit, V car represents the vehicle speed, T env Indicates the ambient temperature, A res It represents the optimal allocation result of SMPC controller for compressor power, refrigerant flow and energy flow path, that is, A res =[P com , F cool , R ene ], where P com Indicates compressor power, F cool Represents the refrigerant flow rate, R ene Represents the energy flow path.

[0092] Construct an optimization problem. The optimization goal of MPC is to minimize the error between the motor and battery temperature in the prediction time domain and the motor and battery target temperature, and complete the optimal allocation of compressor power, refrigerant flow, and energy flow path, so as to construct a cost function.

[0093] The cost function expression is: J(Z K )=J y (Z K )+J U (Z K )+J V (Z K ), that is, the optimization objective expression is: min{J(Z K )}=min{J y (Z K )+J U (Z K )+J V (Z K )}, where J(Z K ) is the error between the motor and battery temperature and the target temperature in the predicted time domain, J y (Z K ) is the error caused by the calculation process of MPC optimization controller, J U (Z K) is the error caused by the system control quantity, J V (Z K ) is the error caused by the system disturbance.

[0094] Reasonable constraints are set for the motor and battery temperature, motor and battery power requirements, compressor cooling power, battery SOC, and battery circuit cooling capacity at each moment in the solution process:

[0095]

[0096] Rolling optimization uses a combination of dynamic programming algorithm and multi-step instant differential learning to solve control actions, reducing the amount of algorithm calculations while efficiently completing the optimal allocation of energy.

[0097] The steps of multi-step instant differential learning are:

[0098] Create a Q value table to record the Q value corresponding to each state and action, initialized to 0 or a random number;

[0099] Setting parameters include learning rate (alpha), discount factor (gamma), exploration rate (epsilon)

[0100] The learning rate determines the degree to which the Q value is updated in each iteration and is set to 0.1. The discount factor is used to represent the impact of delayed rewards on the current Q value and is set to 0.9. The exploration rate indicates that during the training process, random actions are selected with a certain probability to explore unknown states and actions. It is initially set to 0.9, gradually decreases in the later stage, and finally approaches 0.

[0101] Use the epsilon-greedy strategy to select random actions with a probability of exploration rate epsilon, and select the action with the largest current Q value from the Q value table based on the current state;

[0102] Perform the chosen action and observe the new state and immediate reward;

[0103] Use the time difference formula (TD formula) to update the Q value.

[0104] Q(S, A)=(1-alpha)*Q(S, A)+alpha*(R+gamma*maxQ(S', a))

[0105] Where Q(S, A) represents the Q value of taking action A in state S, R represents the immediate reward, S' represents the new state, a represents the action in the new state, and maxQ(S', a) represents the maximum Q value of the optional action in the new state;

[0106] Repeat the above steps until the maximum number of steps or training time is reached.

[0107] Feedback correction, first use the system feedback information to adjust the control strategy in real time to reduce the prediction error caused by the prediction model. At the same time, the driving energy consumption of the whole vehicle model and the energy recovery efficiency of the vehicle are calculated using the formula and periodically fed back to the upper control system to achieve real-time online adjustment of the target temperature.

[0108] The formula for calculating the driving energy consumption of the whole vehicle is: E car =∫(T1n1 / 9549 / μ1+T2n2 / 9549 / μ2)dt,E car represents the energy consumption of the power system in kWh, T1 represents the front electric drive torque in Nm, T2 represents the rear electric drive torque in Nm, n1 represents the front electric drive speed in rpm, n2 represents the rear electric drive speed in rpm, μ1 represents the front electric drive efficiency, and μ2 represents the rear electric drive efficiency;

[0109] The power system constraints are:

[0110]

[0111] The energy recovery efficiency calculation formula of the whole vehicle is: represents the energy recovery efficiency of electric vehicles, Q represents the cumulative discharge of the battery pack, and Q1 represents the cumulative power supply of the drive motor;

[0112] in Q t represents the instantaneous discharge of the battery pack, t1 represents the start time of the electric vehicle, and t2 represents the parking time of the electric vehicle. It indicates the instantaneous power supply amount of the driving motor, t3 indicates the start time of the driving motor power supply, and t4 indicates the end time of the driving motor power supply.

[0113] Based on the target temperature, the SMPC controller performs real-time optimization control on the energy distribution of the thermal management system during vehicle driving.

[0114] The present invention discloses an intelligent energy management method for electric vehicles. First, navigation planning path information of a user is obtained, online map data is used to obtain real-time location information and real-time road condition information, and intelligent network technology is used to obtain various constraint information such as environmental information and altitude information in the driving path. Based on the identification of travel characteristic parameters, the global driving condition of the vehicle is accurately constructed. Then, through a hierarchical energy management strategy, the target control temperature of the motor and the battery is first solved from the perspective of global energy minimization through a dynamic programming algorithm, and then the random model predictive control is used to achieve the optimal allocation of compressor power, refrigerant flow and energy flow path. The information fed back to the upper control strategy by the lower control strategy is used to achieve real-time online adjustment of the target temperature of the motor and the battery, and the real-time energy optimization management of the whole electric vehicle is completed. The present invention finally achieves the optimal allocation of the energy flow path of the whole vehicle, reduces the energy consumption of the whole vehicle, and extends the vehicle's cruising range.

[0115] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent energy management method for electric vehicles, characterized in that: include: Obtain the user's navigation information, use online map data to obtain real-time vehicle information and real-time road condition information to predict driving conditions, use intelligent Internet of Things technology to obtain constraint information in the driving path, identify travel characteristic parameters, construct global driving conditions, and obtain speed time curves; Formulate upper-layer control strategies and lower-layer control strategies based on hierarchical energy management strategies; Inputting the navigation information and the speed-time curve into the upper-level control strategy, optimizing the target temperature of the motor and the battery from the perspective of global energy minimization using a dynamic programming algorithm, and determining the target control temperature of the motor and the battery; The target control temperature of the motor and the battery is input into the lower control strategy, and the energy distribution is optimized and controlled by using a random model prediction algorithm. The information fed back to the upper control strategy by the lower control strategy is used to realize real-time online adjustment of the target control temperature of the motor and the battery, thereby completing the intelligent energy management of the electric vehicle; The target temperature of the motor and the battery is optimized from the perspective of global energy minimization using a dynamic programming algorithm. The method for determining the target control temperature of the motor and the battery includes: According to the current vehicle information and the road condition information, select system state variables and system decision variables; construct a state displacement equation for the whole vehicle according to the system state variables and the system decision variables; establish a target cost function based on global energy minimization; establish constraints in the dynamic programming solution process; obtain the optimal objective function and optimal decision variables in each stage through reverse optimization; according to the optimal objective function and the optimal decision variables, obtain the global optimal state variable solution through forward optimization, and determine the target control temperature of the motor and the battery; The method to construct the state displacement equation for the whole vehicle is: in, is the motor temperature, is the battery temperature, is the motor power required at time t, is the battery power demand at time t, is the cooling power of the compressor at time t, is the instantaneous rate of change of motor temperature, is the instantaneous rate of change of battery temperature; The target cost function is: in, The meaning is the instantaneous cost function, It means the system state variable at time t. It means the system control variable at time t. is the instantaneous battery life decay cost at time t, is the instantaneous power consumption cost at time t, To accumulate battery power consumption, It is the maximum battery power consumption when the power battery capacity decays to 20%. is the total cost of the power battery pack. The cost of electricity consumption.

2. The electric vehicle intelligent energy management method according to claim 1, characterized in that: Before obtaining the user's navigation information, you need to build a driving condition fragment library, including: Based on the historical driving data of electric vehicles, new travel characteristic parameters are constructed through principal component analysis; The fuzzy C-means clustering algorithm is used to construct the initial driving condition segment library. The kinematic segments containing a preset number of categories in the driving condition segment library are obtained through silhouette coefficient evaluation to complete the construction of the driving condition segment library.

3. The electric vehicle intelligent energy management method according to claim 1, characterized in that: The vehicle information includes vehicle position information and vehicle speed information; the road condition information includes average traffic flow speed, road conditions, traffic light status information and altitude information.

4. The electric vehicle intelligent energy management method according to claim 1, characterized in that: The method of optimizing the energy distribution based on the target control temperature of the motor and the battery by using a random model prediction algorithm, and implementing real-time online adjustment of the target control temperature of the motor and the battery through information fed back to the upper control strategy by the lower control strategy includes: Based on the target control temperature of the motor and the battery, the system state quantity, the system control quantity and the system disturbance quantity are selected; a cost function is constructed with the goal of minimizing the error between the motor and battery temperature and the target control temperature in the prediction time domain; constraints are set according to the system state quantity, the system control quantity and the system disturbance quantity; a dynamic programming algorithm is combined with multi-step instant differential learning to complete rolling optimization of the target control temperature of the motor and the battery; the control strategy is adjusted in real time through feedback correction, the driving energy consumption of the whole vehicle and the energy recovery efficiency of the vehicle are obtained by calculation and periodically fed back to the upper-level control strategy, so as to realize real-time online adjustment of the target control temperature of the motor and the battery.

5. The electric vehicle intelligent energy management method according to claim 2, characterized in that: The method to obtain the silhouette coefficient is: in, is the average distance between the i-th sample and other samples in the same cluster category, is the minimum distance between the i-th sample and each sample in other cluster categories.

6. The electric vehicle intelligent energy management method according to claim 4, characterized in that: The cost function is: in, To predict the error between the motor and battery temperature and the target temperature in the time domain, The error caused by the calculation process of the MPC optimization controller, The error caused by the system control quantity, The error caused by the system disturbance.

7. The electric vehicle intelligent energy management method according to claim 4, characterized in that: The method for calculating the driving energy consumption of the whole vehicle and the energy recovery efficiency of the vehicle is: in, is the power system energy consumption, is the front electric drive torque, is the rear electric drive torque, is the front electric drive speed, is the rear electric drive speed, For the front electric drive efficiency, For the rear electric drive efficiency, For the energy recovery efficiency of electric vehicles, is the cumulative discharge amount of the battery pack, The accumulated power supply to drive the motor.