Hierarchical energy management method for plug-in hybrid electric vehicle based on historical data and real-time speed prediction

Abstract

The invention relates to a plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction, and belongs to the technical field of new energy automobiles, and the method comprises the following steps: S1, obtaining and preprocessing historical commuting data of a single vehicle, establishing a historical commuting data set, and predicting a future global driving condition; s2, constructing a vehicle part simplified model, and solving an optimal reference SOC track by using a dynamic programming algorithm; s3, constructing a real-time speed predictor by using a neural network to realize real-time speed prediction; and S4, optimizing a real-time speed prediction result in combination with the optimal reference SOC trajectory, and realizing energy management by using a PMP-MPC algorithm.

Description

Technical Field

[0001] The invention belongs to the technical field of new energy vehicles and relates to a hierarchical energy management method for plug-in hybrid electric vehicles based on historical data and real-time speed prediction. Background Art

[0002] As the number of cars increases, energy shortages and high carbon emissions have become more severe. Plug-in hybrid vehicles are one of the most promising technologies for energy conservation and emission reduction. However, it is still a challenge to reduce vehicle energy consumption in terms of energy management strategies that combine historical vehicle speed predictions.

[0003] At present, the SOC trajectory in hybrid vehicles must be adjusted and optimized according to the mileage. In a long journey, the battery will be exhausted prematurely, and the mileage is too short, resulting in the battery not being completely exhausted. In the case of shorter mileage, the energy management strategy widely uses the CDCS strategy, which uses pure electric endurance, which is indeed the best choice, but if the mileage is exceeded, it will cause excessive fuel consumption. Without considering the driving mileage, most of the existing energy management strategies use linear programming reference SOC to optimize fuel economy, but this method has poor applicability to different roads, and when the mileage is uncertain, the fuel saving effect may not be as good as the CDCS strategy. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction includes the following steps:

[0007] S1: Obtain and preprocess the historical commuting data of a single vehicle, establish a historical commuting data set, and predict future global driving conditions;

[0008] S2: Build a simplified model of vehicle components and use a dynamic programming algorithm to solve the optimal reference SOC trajectory;

[0009] S3: Use neural network to build a real-time speed predictor to achieve real-time prediction of vehicle speed;

[0010] S4: Optimize the real-time speed prediction results in combination with the optimal reference SOC trajectory and use the PMP-MPC algorithm to achieve energy management.

[0011] Further, the step S1 specifically includes the following steps:

[0012] S11: Select a test vehicle and obtain historical commuting data of the vehicle;

[0013] S12: preprocessing the historical commuting data, including differentiation and smoothing operations;

[0014] S13: establishing a historical commuting data set according to the historical commuting data obtained through preprocessing, wherein the historical commuting data set includes speed information in the spatial domain and the temporal domain;

[0015] S14: extracting spatial domain speed information from the historical commuting dataset;

[0016] S15: Construct a future space domain global velocity information prediction model based on the GRU neural network. The calculation process of the GRU neural network includes:

[0017] r t =σ(W r ·[h t-1 ,X t ]+b r )

[0018] z t =σ(W z ·[h t-1 ,X t ]+b Z )

[0019]

[0020] Among them, σ and tanh represent gated activation functions; σ refers to the sigmoid function, expressed as σ = 1 / (1 + exp(-x)), and tanh represents the hyperbolic tangent function, expressed as tanh = (exp(x) - exp(-x)) / (exp(x) + exp(-x)). represents the linear activation function; W r , W z , W0 represents the weight matrix of reset gate, update gate, short-term output and output layer respectively; b r 、b Z , b0 represents the corresponding bias matrix; r t and z t Refers to the output of the reset gate and the update gate; and h t represents candidate hidden states and hidden states; y t Represented as output;

[0021] S16: inputting the spatial domain speed information in the historical commuting data set into the trained future spatial domain global speed information prediction model to predict the future spatiotemporal domain global speed information;

[0022] S17: Extract the predicted global speed prediction information and construct a global driving condition.

[0023] Further, the step S2 specifically includes the following steps:

[0024] S21: simplifying the selected model parameters of the parallel hybrid electric vehicle components to obtain a simplified model of the parallel hybrid electric vehicle components;

[0025] S22: Establish the longitudinal dynamics model of the vehicle;

[0026] S23: Combined with the simplified model, a dynamic programming algorithm is used to solve the optimal reference SOC trajectory and obtain the optimal SOC reference curve.

[0027] Further, the objective function for solving the optimal reference SOC trajectory is:

[0028]

[0029] Among them, m fuel is the fuel consumption rate, c e is the equivalent coefficient of converting electrical energy into fuel consumption, m e is the power consumption, x0 and x f They represent the starting point and end point of the journey respectively.

[0030] Further, the S3 specifically includes the following steps:

[0031] S31: extracting time domain speed information from historical commuting dataset;

[0032] S32: According to the prediction requirements, use LSTM neural network, multi-layer perceptron, random forest or support vector machine to establish speed prediction models with different prediction time domain lengths;

[0033] S33: training the speed prediction model using the extracted time domain speed information;

[0034] S34: Using the trained speed prediction model, setting the prediction time length, and making a real-time prediction of the actual vehicle speed.

[0035] Further, the S4 specifically includes the following steps:

[0036] S41: extracting the predicted SOC global optimal reference curve;

[0037] S42: using the real-time speed predictor to predict the short-term speed sequence of the next 5 seconds, and passing it to the optimal energy allocation module;

[0038] S43: Combine the global SOC optimal curve and the short-term speed sequence to calculate the upper and lower bounds of the SOC trajectory in the prediction domain. The Pontryagin algorithm PMP is used to solve the optimal control problem in the prediction domain. There are two key points in the PMP algorithm calculation. The first is the selection of the initial value of the co-state, which is generally selected by the shooting method, and the second is the x of each segment. target Selection of

[0039] S44: Use the MPC algorithm to solve the energy management problem in real time, calculate the optimal power distribution while ensuring the validity of the constraints, and distribute the optimal control amount to the controller to control the hybrid vehicle.

[0040] Further, in step S44, the cost function is:

[0041]

[0042] Among them, m fuel is the fuel consumption rate, T is the prediction domain length of MPC control, SOC(t) is the position of the vehicle at time t, SOC ref (t) is the SOC value of the terminal in the predicted domain at time t, is the weight coefficient for penalizing SOC tracking.

[0043] The beneficial effects of the present invention are:

[0044] 1) The present invention uses historical data to quickly predict the global operating conditions and uses a simplified model to quickly plan the optimal reference SOC, thereby improving the energy consumption performance of the vehicle and optimizing the battery discharge strategy.

[0045] 2) The real-time speed prediction model trained by the present invention using the historical vehicle speed data will have a better prediction effect.

[0046] 3) The present invention utilizes the optimal reference SOC to perform reference constraints on the real-time speed prediction domain terminal, thereby reducing the number of engine starts and stops and improving fuel economy.

[0047] 4) The lower-level real-time energy management algorithm of the present invention adopts the PMP-MPC algorithm, and the obtained result is more economical.

[0048] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0050] Figure 1 It is a logical structure diagram of the energy management strategy of the present invention;

[0051] Figure 2 It is a schematic diagram of the target shooting method structure;

[0052] Figure 3 Schematic diagram of energy management strategy based on PMP-MPC. DETAILED DESCRIPTION

[0053] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0054] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0055] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0056] See also Figure 1 ,The hierarchical energy management method for plug-in hybrid vehicles based on historical data and real-time speed prediction can be divided into the following steps:

[0057] S1: Obtain the historical commuting data of a single vehicle, preprocess the data, establish a historical commuting dataset, and predict future global driving conditions;

[0058] S2: Obtain a simplified model of vehicle components and use a dynamic programming algorithm to solve the optimal reference SOC trajectory;

[0059] S3: Combine the neural network with the speed data set to build a real-time speed predictor, and use the real-time speed predictor to predict the vehicle speed in real time;

[0060] S4: Combined with the optimal reference SOC trajectory, the PMP-MPC controller is used to achieve energy management, such as Figure 3 The present invention utilizes historical commuting data to plan a global reference SOC trajectory in advance, thereby improving the economy of the hybrid vehicle.

[0061] As an optional embodiment, the above S1 specifically includes S11-S16:

[0062] S11: Select a test vehicle and obtain personal historical commuting data of the vehicle.

[0063] S12: preprocessing the data, including differentiation, smoothing and other operations, to reduce errors caused by problems such as traffic lights, overtaking, and GPS signal loss;

[0064] S13: Establishing a data set based on the historical commuting data obtained through preprocessing, wherein the data set includes speed information in the spatial domain and the temporal domain.

[0065] S14: extracting spatial domain speed information from the historical commuting dataset;

[0066] S15: Construct a future space domain global velocity information prediction model based on the GRU neural network. The calculation process of the GRU neural network includes:

[0067] r t =σ(W r ·[h t-1 ,X t ]+b r )

[0068] z t =σ(W r ·[h t-1 ,X t ]+b Z )

[0069]

[0070] Among them, σ and tanh represent gated activation functions. σ refers to the sigmoid function, expressed as σ = 1 / (1 + exp(-x)), and tanh refers to the hyperbolic tangent function, expressed as tanh = (exp(x) - exp(-x)) / (exp(x) + exp(-x)). represents the linear activation function. W r , W z , W0 represents the weight matrix of the reset gate, update gate, short-term output and output layer respectively. r , b Z , b0 represents the corresponding bias matrix. t and z t Refers to the output of the reset gate and the update gate. and h t Denotes candidate hidden states and hidden states. t Represented as output.

[0071] S16: inputting the spatial domain speed information in the historical commuting data set into the trained future spatial domain global speed information prediction model to predict the future spatiotemporal domain global speed information;

[0072] S17: Extract the predicted global speed prediction information and construct a global driving condition.

[0073] As an optional embodiment, the above S2 specifically includes S21-S24:

[0074] S21: simplifying the selected model parameters of the parallel hybrid electric vehicle components to obtain a simplified model of the parallel hybrid electric vehicle components;

[0075] S22: Establish the longitudinal dynamics model of the vehicle;

[0076] S23: Combined with the simplified model, a dynamic programming algorithm is used to solve the optimal reference SOC trajectory and obtain the optimal SOC reference curve.

[0077] Further, the objective function for solving the optimal reference SOC trajectory is:

[0078]

[0079] Among them, m fuel is the fuel consumption rate, c e is the equivalent coefficient of converting electrical energy into fuel consumption, m e The amount of power consumed.

[0080] As an optional embodiment, the above S3 specifically includes S31-S34:

[0081] S31: extracting time domain speed information from historical commuting dataset;

[0082] S32: According to the prediction requirements, use LSTM neural network, multi-layer perceptron, random forest or support vector machine to establish speed prediction models with different prediction time domain lengths;

[0083] S33: training the speed prediction model using the extracted time domain speed information;

[0084] S34: Using the trained speed prediction model, setting the prediction time length, and making a real-time prediction of the actual vehicle speed.

[0085] Optionally, the speed prediction model in S33 selects an LSTM algorithm, but is not limited to this machine learning method, and may also be an algorithm such as a multi-layer perceptron, a random forest, or a support vector machine.

[0086] As an optional embodiment, the above S3 specifically includes S41-S43:

[0087] S41: extracting the predicted SOC global optimal reference curve;

[0088] S42: using the real-time speed predictor to predict the short-term speed sequence of the next 5 seconds, and passing it to the optimal energy allocation module;

[0089] S43: Combine the global SOC optimal curve and the short-term speed sequence to calculate the upper and lower bounds of the SOC trajectory in the prediction domain, and use the PMP algorithm to solve the optimal control problem in the prediction domain. There are two key points in the PMP algorithm calculation. The first is the selection of the initial value of the co-state, which is generally selected by the shooting method, and the second is the x of each segment. target The selection of the target shooting method is as follows Figure 2 shown.

[0090] S44: Use the MPC algorithm to solve the energy management problem in real time, calculate the optimal power distribution while ensuring the validity of the constraints, and distribute the optimal control amount to the controller to control the hybrid vehicle.

[0091] Further, in step S44, the cost function is:

[0092]

[0093] Among them, m fuel is the fuel consumption rate, T is the prediction domain length of MPC control, SOC(t) is the position of the vehicle at time t, SOC ref (t) is the SOC value of the terminal in the predicted domain at time t, is the weight coefficient for penalizing SOC tracking.

[0094] In the above embodiments, the description's reference to "this embodiment" indicates that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least some embodiments, but not necessarily all embodiments. Multiple occurrences of "this embodiment" do not necessarily all refer to the same embodiment.

[0095] In the above-described embodiments, although the invention has been described in conjunction with specific embodiments of the invention, many substitutions, modifications, and variations of these embodiments will be apparent to those of ordinary skill in the art based on the foregoing description. For example, other storage structures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed. Embodiments of the invention are intended to encompass all such substitutions, modifications, and variations that fall within the broad scope of the appended claims.

[0096] This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, any one of the methods in this embodiment is implemented.

[0097] This embodiment also provides an electronic terminal, including: a processor and a memory;

[0098] The memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the terminal executes any one of the methods in this embodiment.

[0099] The computer-readable storage medium in this embodiment can be understood by ordinary technicians in this field: all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to the computer program. The aforementioned computer program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

[0100] The electronic terminal provided in this embodiment includes a processor, a memory, a transceiver and a communication interface. The memory and the communication interface are connected to the processor and the transceiver and complete communication with each other. The memory is used to store computer programs, the communication interface is used to communicate, and the processor and the transceiver are used to run computer programs so that the electronic terminal executes each step of the above method.

[0101] In this embodiment, the memory may include a random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0102] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0103] The present invention can be used in many general or special computing system environments or configurations, such as personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and the like.

[0104] The present invention may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction, characterized in that: The following steps are involved: S1: Obtain and preprocess the historical commuting data of a single vehicle, establish a historical commuting data set, and predict future global driving conditions; S2: Build a simplified model of vehicle components and use a dynamic programming algorithm to solve the optimal reference SOC trajectory; S3: Use neural network to build a real-time speed predictor to achieve real-time prediction of vehicle speed; S4: Optimize the real-time speed prediction results in combination with the optimal reference SOC trajectory and use the PMP-MPC algorithm to achieve energy management.

2. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: The step S1 specifically includes the following steps: S11: Select a test vehicle and obtain historical commuting data of the vehicle; S12: preprocessing the historical commuting data, including differentiation and smoothing operations; S13: establishing a historical commuting data set according to the historical commuting data obtained through preprocessing, wherein the historical commuting data set includes speed information in the spatial domain and the temporal domain; S14: extracting spatial domain speed information from the historical commuting dataset; S15: Construct a future space domain global velocity information prediction model based on the GRU neural network. The calculation process of the GRU neural network includes: r t =σ(W r ·[h t-1 ,X t ]+b r ) z t =σ(W z ·[h t-1 ,X t ]+b Z ) Among them, σ and tanh represent gated activation functions; σ refers to the sigmoid function, expressed as σ = 1 / (1 + exp(-x)), and tanh represents the hyperbolic tangent function, expressed as tanh = (exp(x) - exp(-x)) / (exp(x) + exp(-x)). represents the linear activation function; W r , W z , W0 represents the weight matrix of reset gate, update gate, short-term output and output layer respectively; b r , b Z , b0 represents the corresponding bias matrix; r t and z t Refers to the output of the reset gate and the update gate; and h t represents candidate hidden states and hidden states; y t Represented as output; S16: inputting the spatial domain speed information in the historical commuting data set into the trained future spatial domain global speed information prediction model to predict the future spatiotemporal domain global speed information; S17: Extract the predicted global speed prediction information and construct a global driving condition.

3. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: The step S2 specifically includes the following steps: S21: simplifying the selected model parameters of the parallel hybrid electric vehicle components to obtain a simplified model of the parallel hybrid electric vehicle components; S22: Establish the longitudinal dynamics model of the vehicle; S23: Combined with the simplified model, a dynamic programming algorithm is used to solve the optimal reference SOC trajectory and obtain the optimal SOC reference curve.

4. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: The objective function for solving the optimal reference SOC trajectory is: Among them, m fue l is the fuel consumption rate, c is e is the equivalent coefficient of converting electrical energy into fuel consumption, m e is the power consumption, x0 and x f They represent the starting point and end point of the journey respectively.

5. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: The S3 specifically includes the following steps: S31: extracting time domain speed information from historical commuting dataset; S32: According to the prediction requirements, use LSTM neural network, multi-layer perceptron, random forest or support vector machine to establish speed prediction models with different prediction time domain lengths; S33: training the speed prediction model using the extracted time domain speed information; S34: Using the trained speed prediction model, setting the prediction time length, and making a real-time prediction of the actual vehicle speed.

6. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: The S4 specifically comprises the following steps: S41: extracting the predicted SOC global optimal reference curve; S42: using the real-time speed predictor to predict the short-term speed sequence of the next 5 seconds, and passing it to the optimal energy allocation module; S43: Combine the global SOC optimal curve and the short-term speed sequence to calculate the upper and lower bounds of the SOC trajectory in the prediction domain, and use the Pontryagin algorithm PMP to solve the optimal control problem in the prediction time domain; S44: Use the MPC algorithm to solve the energy management problem in real time, calculate the optimal power distribution while ensuring the validity of the constraints, and distribute the optimal control amount to the controller to control the hybrid vehicle.

7. The plug-in hybrid electric vehicle hierarchical energy management method based on historical data and real-time speed prediction according to claim 1, characterized in that: In step S44, the cost function is: Among them, m fuel is the fuel consumption rate, T is the prediction domain length of MPC control, SOC(t) is the position of the vehicle at time t, SOC ref (t) is the SOC value of the terminal in the predicted domain at time t, is the weight coefficient for penalizing SOC tracking.

Images