Vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicle

Abstract

The invention discloses a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid electric vehicle. The method comprises the following steps: acquiring working condition information data and vehicle system data; according to the working condition information data and the vehicle system data, an SOC optimal track and an SOC optimal domain under the corresponding working condition and the vehicle power system are obtained; constructing an energy map drawing model according to the SOC optimal domain, and obtaining an energy map and an SOC track by training and testing the energy map drawing model; and constructing an optimal SOC trajectory generation model according to the energy map and the SOC trajectory, correcting the SOC trajectory by using the optimal SOC trajectory generation model, and obtaining an SOC reference trajectory. According to the method provided by the invention, the obtained SOC reference trajectory is similar to the optimal trajectory obtained by a dynamic planning algorithm, the oil-saving capability is remarkable, the calculation load of the algorithm is greatly reduced, and a foundation is laid for real-time application of energy management.

Description

Technical Field

[0001] The present invention relates to the technical field of hybrid vehicle energy management, and in particular to a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid vehicle. Background Art

[0002] This section merely provides background information related to the present disclosure and is not necessarily prior art.

[0003] Currently, rule-based energy management approaches are widely used in real-world applications. These methods determine threshold parameters based on engineering experience and theoretical knowledge, and select operating modes based on predefined strategies. These methods are relatively mature and offer good real-time performance, making them the most widely used in real-world applications. While simple and easy to implement, these methods rely heavily on expert experience, resulting in fuel savings only at specific times and limited utilization of global operating conditions.

[0004] To maximize the fuel-saving potential of energy management, research has focused on optimization-based strategies. Based on the system's control objective function and constraints, optimization algorithms are used to identify and apply the optimal control strategy. Strategies based on dynamic programming have significantly improved fuel economy over the entire driving cycle and serve as the foundation and criterion for evaluating other algorithms. However, these algorithms suffer from high computational load, low efficiency, and high computer power requirements, making online applications difficult.

[0005] Therefore, for energy management algorithms, fuel-saving ability and real-time performance are two important indicators. How to reduce the computational load of the optimization algorithm to achieve real-time application and how to obtain a control sequence with fuel-saving ability similar to that of the dynamic programming algorithm are key issues that need to be solved. Summary of the Invention

[0006] The purpose of the present invention is to solve the problem of how to reduce the computational load of the optimization algorithm to realize real-time application. This purpose is achieved by the following technical solutions:

[0007] The present invention proposes a global optimization energy management method for a hybrid electric vehicle based on an energy map real-time algorithm. The global optimization energy management method for a hybrid electric vehicle based on an energy map real-time algorithm comprises:

[0008] Obtain working condition information data and vehicle system data;

[0009] Obtaining an optimal SOC trajectory and an optimal SOC range under corresponding working conditions and vehicle power systems according to the working condition information data and the vehicle system data;

[0010] Constructing an energy map drawing model according to the SOC optimal domain, and obtaining an energy map and an SOC trajectory by training and testing the energy map drawing model;

[0011] An optimal SOC trajectory generation model is constructed according to the energy map and the SOC trajectory, the SOC trajectory is corrected using the optimal SOC trajectory generation model, and an SOC reference trajectory is obtained.

[0012] In some embodiments of the present invention, the operating condition information data includes a plurality of vehicle speed operating condition curves;

[0013] The vehicle system data includes but is not limited to the vehicle's required power model data, engine model data, motor model data, and battery model data.

[0014] In some embodiments of the present invention, the step of obtaining the required power model data includes:

[0015] The power demand model is established according to the following formula:

[0016]

[0017] Among them, f is the rolling resistance coefficient, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance coefficient, A is the frontal area, δ is the rotation mass conversion coefficient, u a is the vehicle speed, η T is the mechanical efficiency of the transmission system.

[0018] In some embodiments of the present invention, the step of obtaining the engine model data includes:

[0019] According to the formula be=f1(n e ,T e ) Establish an engine model;

[0020] Calculation formula based on engine fuel consumption per unit time Calculate fuel consumption;

[0021] n e is the speed, T e is the torque, b e (g / kWh) is the fuel consumption rate, B is the fuel consumption per unit time (g / s), P e is the engine power (kW).

[0022] In some embodiments of the present invention, the step of obtaining the motor model data includes:

[0023] According to the formula η m =f2(n m ,T m ) Establish a motor model;

[0024] In some embodiments of the present invention, the step of obtaining the battery model data includes:

[0025] According to the formula Build a battery model;

[0026] Among them, U oc is the battery open circuit voltage, R int is the equivalent internal resistance of the battery, P bat is the battery power, and C is the battery capacity.

[0027] In some embodiments of the present invention, the step of obtaining the SOC optimal trajectory and SOC optimal range under the corresponding working condition and vehicle power system according to the working condition information data and the vehicle system data includes:

[0028] Determine a feasible region of a state variable SOC according to the battery model;

[0029] According to the feasible domain of the SOC, a dynamic programming algorithm is adopted to obtain the optimal SOC trajectory and the optimal SOC domain under the corresponding working conditions and vehicle power system with the goal of minimizing fuel consumption.

[0030] In some embodiments of the present invention, the step of constructing an energy mapping model according to the SOC optimal domain and obtaining an energy map and an SOC trajectory by training and testing the energy mapping model includes:

[0031] Assigning 1 to the optimal state point in the SOC optimal domain and 0 to the remaining points, thereby obtaining an energy map in the form of a SOC optimal domain 01 matrix;

[0032] Create multiple input features;

[0033] Building an energy map drawing model based on the network structure of the spatial perception network, substituting the plurality of input features into the input end of the energy map drawing model, and substituting the corresponding energy map into the output end of the energy map drawing model;

[0034] The energy map and the SOC trajectory are obtained through data training and testing.

[0035] In some embodiments of the present invention, the plurality of input features include X=(v,a,t,Preq p ,Preq n ,SOC);

[0036] Where v represents the vehicle speed at time t, a represents the acceleration, and Preq p represents the sum of the cumulative positive power demand at time t, Preq nIt represents the sum of the accumulated negative power demand at time t, where t represents time. SOC is the abbreviation of State of Charge, which usually represents the state of charge, and can also be called the state of remaining power. SOC is used to describe the percentage of the remaining power of the battery in electric vehicles, hybrid vehicles or other energy storage systems to the total capacity. In the formula X = (v, a, t, Preq p ,Preq n , SOC), SOC represents the feasible region of state of charge.

[0037] In some embodiments of the present invention, the steps of constructing an optimal SOC trajectory generation model based on the energy map and the SOC trajectory, correcting the SOC trajectory using the optimal SOC trajectory generation model, and obtaining an SOC reference trajectory include:

[0038] According to the energy map The probability matrix of

[0039] According to the Obtain the maximum value of each column of the probability matrix, and obtain the SOC trajectory in the time dimension based on the maximum value;

[0040] Create multiple input features;

[0041] An optimal SOC trajectory generation model is built based on a time series network architecture, and multiple input features are substituted into the input end of the optimal SOC trajectory generation model. The output of the optimal SOC trajectory generation model is the SOC reference trajectory.

[0042] In some embodiments of the present invention, the plurality of input features X include X=(v,a,t,Preq p ,Preq n ,SOC peak );

[0043] Where v represents the vehicle speed at time t, a represents the acceleration, and Preq p represents the sum of the cumulative positive power demand at time t, Preq n It represents the sum of the cumulative negative power demand at time t, where t represents time and SOC peak Represents the SOC trajectory output by the energy map model.

[0044] The vehicle global optimization energy management method for hybrid vehicles provided by the present invention, based on a real-time energy map algorithm, obtains an SOC reference trajectory that is similar to the optimal trajectory obtained by a dynamic programming algorithm, has significant fuel-saving capabilities, and greatly reduces the computational load of the algorithm, laying the foundation for the real-time application of energy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference numerals are used throughout the accompanying drawings to denote the same components. In the accompanying drawings:

[0046] Figure 1 This is a flowchart of the steps of a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid vehicle according to an embodiment of the present invention;

[0047] Figure 2 This is a flowchart of step 2 of a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid vehicle according to an embodiment of the present invention;

[0048] Figure 3 This is a flowchart of step 3 of a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid vehicle according to an embodiment of the present invention;

[0049] Figure 4 This is a flow chart of the method for implementing real-time vehicle energy management based on the energy map algorithm described in the present invention.

[0050] Figure 5 Schematic diagram of the feasible region of the state variable SOC described in the present invention.

[0051] Figure 6 This is a schematic diagram of the energy map described in the present invention.

[0052] Figure 7 This is the network structure of the energy map drawing model described in the present invention.

[0053] Figure 8 Schematic diagram of the optimal SOC trajectory generation model described in the present invention. DETAILED DESCRIPTION

[0054] Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0055] It should be understood that the terms used herein are for the purpose of describing specific example embodiments only and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "comprise", "include", "contain" and "have" are inclusive and therefore specify the presence of stated features, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be performed in the specific order described or illustrated, unless the order of execution is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0056] Although the terms first, second, third, etc. can be used in the text to describe multiple elements, components, regions, layers and / or sections, these elements, components, regions, layers and / or sections should not be limited by these terms. These terms can only be used to distinguish an element, component, region, layer or section from another region, layer or section. Unless the context clearly indicates otherwise, terms such as "first", "second" and other numerical terms do not imply order or sequence when used in the text. Therefore, the first element, component, region, layer or section discussed below can be referred to as the second element, component, region, layer or section without departing from the teaching of the example embodiments.

[0057] For ease of description, spatially relative terms may be used herein to describe the relationship of one element or feature relative to another element or feature as shown in the figures, such as "inside," "outside," "inside," "outside," "below," "beneath," "above," and the like. Such spatially relative terms are intended to include different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if the device in the figures is flipped, an element described as "below" or "below" another element or feature would then be oriented as "above" or "above" another element or feature. Thus, the example term "below" can include both above and below orientations. The device may be otherwise oriented (rotated 90 degrees or in other orientations) and the spatially relative descriptors used herein are interpreted accordingly.

[0058] like Figure 1 and Figure 4 As shown, according to an embodiment of the present invention, a vehicle global optimization energy management method based on an energy map real-time algorithm for a hybrid vehicle is proposed, comprising the following steps:

[0059] Step 1: Obtain working condition information data and vehicle system data;

[0060] Step 2: Obtain the optimal SOC trajectory and optimal SOC range under the corresponding working condition and vehicle power system based on the working condition information data and vehicle system data;

[0061] Step 3: Construct an energy mapping model based on the SOC optimal domain, and obtain the energy map and SOC trajectory by training and testing the energy mapping model;

[0062] Step 4: Construct an optimal SOC trajectory generation model based on the energy map and SOC trajectory, use the optimal SOC trajectory generation model to correct the SOC trajectory, and obtain the SOC reference trajectory.

[0063] Steps 1 and 2 are the data preparation phase. Specifically, the vehicle energy management acquires operating condition information and, through the utilization of the vehicle's powertrain components, controls the power of these components, thereby ensuring the proper allocation of energy from different energy sources. Therefore, data preparation begins with three perspectives: operating condition information, the vehicle's powertrain, and energy allocation strategies.

[0064] In step 1, the operating condition information data refers to vehicle speed operating condition curves. Because the energy map algorithm involves training and testing, multiple vehicle speed operating condition curves are prepared as training and testing samples. The multiple vehicle speed operating condition curves refer to greater than or equal to 100 vehicle speed operating condition curves. For example, in some embodiments, the number referred to by "multiple" can be 100, 150, 200, 300, 400, 500, 600, etc. In this embodiment, the multiple is preferably 500.

[0065] For the vehicle power system, after obtaining the vehicle speed operating condition curve, the backward modeling method can be used to start from the wheel-side demand and transfer energy to the power system through the transmission system.

[0066] In detail, data preparation for vehicle systems includes the following aspects:

[0067] The demand power model is as follows:

[0068]

[0069] Among them, f is the rolling resistance coefficient, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance coefficient, A is the frontal area, δ is the rotation mass conversion coefficient, u a is the vehicle speed, η T is the mechanical efficiency of the transmission system.

[0070] In order to calculate the fuel consumption and electricity consumption at each moment, it is also necessary to model the main components of the power system.

[0071] (1) Engine model: Based on the quasi-static assumption, the engine fuel consumption model uses the nonlinear fuel consumption map measured by the engine steady-state experiment. e and torque T e Fuel consumption rate b e (g / kWh) is obtained by interpolation from f1(·).

[0072] be=f1(n e ,T e )

[0073] The fuel consumption per unit time of the engine is:

[0074]

[0075] Among them, B is the fuel consumption per unit time (g / s), P e is the engine power (kW).

[0076] (2) Motor model: Similarly, the experimental modeling method is also used to build the motor model, and the efficiency characteristics are represented by nonlinear 3-D MAP. m and torque T m Motor efficiency η under m Obtained by interpolation of f2(·).

[0077] η m =f2(n m ,T m )

[0078] (3) Battery model: The battery model is simplified to an equivalent internal resistance model. The calculation method of the battery state of charge (SOC) at each moment is:

[0079]

[0080] Among them, U oc is the battery open circuit voltage, R int is the equivalent internal resistance of the battery, P bat is the battery power, C is the battery capacity, and SOC is the remaining battery capacity, also known as the state of charge.

[0081] Step 2 belongs to the data preparation of the allocation strategy. Specifically, in step 2, the data preparation of the allocation strategy refers to using a dynamic programming algorithm to obtain the optimal SOC trajectory and optimal domain under the corresponding working conditions and vehicle power system with the goal of minimizing fuel consumption.

[0082] like Figure 2 As shown, step 2 includes the following steps:

[0083] Step 2.1: Determine the feasible region of the state variable SOC based on the battery model;

[0084] Step 2.2: Based on the feasible domain of SOC, a dynamic programming algorithm is used to obtain the optimal SOC trajectory and optimal SOC domain under the corresponding working conditions and vehicle power system with the goal of minimizing fuel consumption.

[0085] Specifically, in step 2.1, the feasible region of the state variable SOC is determined based on the battery model, such as Figure 5 As shown, including the maximum and minimum boundaries SOC max and SOC min , and the initial and final conditions SOC start and SOC terminal , and the state feasible domain is discretized according to the discrete interval ΔSOC.

[0086] In step 2.2, based on the given vehicle speed conditions and the energy consumption modeling of the entire vehicle and its components, as well as the discrete SOC state feasible domain, the fuel consumption between any two state points at adjacent moments can be calculated and stored in the fuel consumption matrix, represented by fuel(m,n,k) (representing the fuel consumption from the mth state point at the kth moment to the nth state point at the k+1th moment); next, a dynamic programming algorithm is used to find the optimal path from each state point to the starting point, record all the optimal state points that each state point has passed through in the previous moment, and finally backtrack to obtain the global optimal domain. The position of the optimal state point is assigned a value of 1 to represent the optimal domain of the energy map, and the rest are assigned a value of 0 to represent the background. The obtained global optimal domain (01 matrix) is the energy map corresponding to the operating condition information and the vehicle state, such as Figure 6 shown.

[0087] like Figure 3 As shown, step 3 includes the following steps:

[0088] Step 3.1: Assign the optimal state point in the SOC optimal domain to 1 and the rest to 0, and obtain the energy map in the form of the SOC optimal domain 01 matrix;

[0089] Step 3.2: Create multiple input features;

[0090] Step 3.3: Build an energy map drawing model based on the network structure of the spatial perception network, substitute multiple input features into the input end of the energy map drawing model, and substitute the corresponding energy map into the output end of the energy map drawing model;

[0091] Step 3.4: Obtain energy map and SOC trajectory through data training and testing.

[0092] Specifically, the energy map is drawn by analyzing a series of input parameter matrices, extracting their high-dimensional features, and then classifying each point of the output matrix into 01. Then, an association between a label or category and each point of the output matrix is ​​established, and each pixel in the energy map is classified according to its semantic category, thereby realizing pixel-level energy map drawing.

[0093] In step 3.1, the energy map refers to the process of obtaining the global optimal domain through data preparation. The optimal state point is assigned a value of 1, representing the optimal domain of the energy map, and the remaining values ​​are assigned a value of 0, representing the background. The obtained global optimal domain, that is, the 01 matrix, is the energy map.

[0094] In steps 3.2 to 3.4, the network structure based on spatial perception network is used to implement the above functions. The output of the model That is the energy map mapped to the 01 matrix, where I1=(SOC max -SOC min ) / ΔSOC+1, I2 is the working time; the input of the model Where I3 represents the number of input features, including but not limited to the following six features and derived features constructed by combining multiple features: vehicle speed v at time t, acceleration a, and the sum of the cumulative positive power demand at time t Preq p , the sum of the cumulative negative power demand at time t Preq n , time t, and state feasible region SOC.

[0095] X=(v,a,t,Preq p ,Preq n ,SOC)

[0096] in, middle I indicates that the input feature X is a real matrix of I1×I2×I3; middle I indicates that the output matrix Y is an I1×I2 real matrix.

[0097] The energy map drawing model structure based on space perception network is as follows Figure 7 As shown, it consists of an encoder, a decoder and a jump connection part.

[0098] The encoder aims to extract deep features from the input tensor. Each layer includes an activation function and a normalization layer. The activation function introduces nonlinear characteristics to each layer, making the network more capable of solving complex problems. By performing nonlinear transformations, the network can better capture complex patterns and features in the data. Common activation functions include but are not limited to the Tanh function, the ReLU function, and the Leaky ReLU function. The normalization layer ensures that the inputs of each layer maintain the same distribution, accelerating network convergence and preventing overfitting.

[0099] The encoder is responsible for gradually extracting features from the input matrix and reducing the spatial resolution through the step size. In order to complete the conversion of extracted features to energy maps, the decoder is required to restore the extracted feature matrix to the size of the original input matrix and perform image-level classification. The final output is a result matrix with the same size as the original input matrix and segmentation labels. The decoder is used to implement feature upsampling in order to better restore the spatial resolution and achieve end-to-end model tasks. Each layer includes an activation function and a normalization layer. Its activation function and normalization layer are the same as those of the encoder, and are symmetrical in terms of network structure. It is worth noting that since the final energy map is actually composed of pixel-level segmentation labels, the Sigmoid function is used as the activation function in the final classification.

[0100] When downsampling is performed in the encoder, the feature information of the original matrix will be lost. This process is not reversible. Therefore, a jump connection is used to connect the feature matrix of the encoder stage with the feature matrix recovered by the decoder stage, providing the decoder with multi-scale and multi-level information, thereby obtaining a more refined segmentation effect.

[0101] Since the energy map model finally outputs a matrix with 01 labels, the binary classification problem uses the cross entropy loss function. Since the 0 samples and 1 samples in the target are unbalanced, Focalloss is used instead of the cross entropy loss function, as shown below.

[0102]

[0103] Where y pred is the model’s predicted probability for the sample, y true is the true value of the sample.

[0104] The output y after the above encoder and decoder is a probability matrix. In order to obtain the final binary classification energy map, a threshold θ needs to be set. The formula is as follows.

[0105] Y=Heaviside(y-θ)

[0106] This completes the drawing of the energy map.

[0107] Specifically, in the process of generating the energy map in step 2, the Sigmoid function will output a The probability matrix of each column is obtained by taking the maximum value of each column to obtain the SOC trajectory in the time dimension. This trajectory is the SOC trajectory output by the energy map model based on the spatial perception network architecture. peak , thus obtaining the SOC trajectory.

[0108] Step 4 includes the following steps:

[0109] Step 4.1: Obtain from energy map The probability matrix of

[0110] Step 4.2: According to The probability matrix obtains the maximum value of each column and obtains the SOC trajectory in the time dimension based on the maximum value;

[0111] Step 4.3: Create multiple input features;

[0112] Step 4.4: Build an optimal SOC trajectory generation model based on the time series network architecture, substitute multiple input features into the input of the optimal SOC trajectory generation model, and the output of the optimal SOC trajectory generation model is the SOC reference trajectory.

[0113] In detail, in order to realize the online application of the algorithm, a reference SOC trajectory needs to be obtained. In the process of generating the energy map in step 2, the Sigmoid function will output a The probability matrix of each column is obtained by taking the maximum value of each column to obtain the SOC trajectory in the time dimension. This trajectory is the SOC trajectory output by the energy map model based on the spatial perception network architecture. peak , in order to obtain a more accurate reference SOC trajectory SOC ref , including but not limited to the following six features and the derived features constructed by combining multiple features as input vectors: vehicle speed v at time t, acceleration a, the sum of the cumulative positive demand power Preq at time t p , the sum of the cumulative negative power demand at time t Preq n , time t, and SOC peak , the output is SOC ref .

[0114] X=(v,a,t,Preq p ,Preq n ,SOC peak )

[0115] Predicting the optimal SOC trajectory for the entire vehicle speed profile is a long-term prediction problem, and time series network structures are well suited to processing this type of long-term input sequence information. Therefore, using a time series network-based architecture to predict the optimal SOC trajectory for the entire vehicle speed profile is reasonable. It is worth noting that the SOC trajectory for minimum fuel consumption obtained using a dynamic programming algorithm is global, as input information at each moment influences each other. In other words, the SOC at each moment is affected by both upstream and downstream factors.

[0116] The network structure is as follows Figure 8 The optimal SOC trajectory generation model based on the time series network consists of two layers with the same number of neurons, the forward network layer and the reverse network layer. For the input information, two separate hidden states representing the forward and reverse sequences are calculated respectively. and Forward encoding of the forward network layer Mainly capture from x0 to x t Past information, reverse encoding of the reverse network layer Main capture from x t to x n The output of the forward network layer and the reverse network layer are superimposed and fused according to certain weights to obtain the final output result h t .

[0117] The reference SOC trajectory is thus obtained. The global optimized energy management method for hybrid vehicles, based on a real-time energy map algorithm, provided by the present invention, obtains an SOC reference trajectory similar to the optimal trajectory obtained by the dynamic programming algorithm, significantly saving fuel and significantly reducing the algorithm's computational load, laying the foundation for real-time application of energy management.

[0118] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A vehicle global optimization energy management method for a hybrid electric vehicle based on an energy map real-time algorithm, characterized in that: The method comprises: Obtain working condition information data and vehicle system data; Obtaining an optimal SOC trajectory and an optimal SOC range under corresponding working conditions and vehicle power systems according to the working condition information data and the vehicle system data; Constructing an energy map drawing model according to the SOC optimal domain, and obtaining an energy map and an SOC trajectory by training and testing the energy map drawing model; An optimal SOC trajectory generation model is constructed according to the energy map and the SOC trajectory, the SOC trajectory is corrected using the optimal SOC trajectory generation model, and an SOC reference trajectory is obtained.

2. The vehicle global optimization energy management method based on the energy map real-time algorithm of the hybrid electric vehicle according to claim 1 is characterized in that: The operating condition information data includes a plurality of vehicle speed operating condition curves; The vehicle system data includes but is not limited to the vehicle's required power model data, engine model data, motor model data, and battery model data.

3. The vehicle global optimization energy management method based on the energy map real-time algorithm of the hybrid electric vehicle according to claim 2 is characterized in that: The step of obtaining the required power model data includes: The power demand model is established according to the following formula: Among them, f is the rolling resistance coefficient, m is the mass of the car, g is the acceleration of gravity, α is the slope angle, C D is the air resistance coefficient, A is the frontal area, δ is the rotation mass conversion coefficient, u a is the vehicle speed, η T is the mechanical efficiency of the transmission system.

4. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 2, characterized in that: The step of obtaining the engine model data includes: According to the formula be=f1(n e ,T e ) Establish an engine model; Calculation formula based on engine fuel consumption per unit time Calculate fuel consumption; n e is the speed, T e is the torque, b e (g / kWh) is the fuel consumption rate, B is the fuel consumption per unit time (g / s), P e is the engine power (kW).

5. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 2, characterized in that: The step of obtaining the motor model data includes: According to the formula η m =f2(n m ,T m ) Establish a motor model; Among them, n m is the speed, T m is the torque, η m is the motor efficiency.

6. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 2, characterized in that: The steps of obtaining the battery model data include: According to the formula Build a battery model; Among them, U oc is the battery open circuit voltage, R int is the equivalent internal resistance of the battery, P bat is the battery power, and C is the battery capacity.

7. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 6, characterized in that: The step of obtaining the SOC optimal trajectory and SOC optimal range under the corresponding working condition and vehicle power system according to the working condition information data and the vehicle system data includes: Determine a feasible region of a state variable SOC according to the battery model; According to the feasible domain of the SOC, a dynamic programming algorithm is adopted to obtain the optimal SOC trajectory and the optimal SOC domain under the corresponding working conditions and vehicle power system with the goal of minimizing fuel consumption.

8. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 1, characterized in that: The steps of constructing an energy map drawing model according to the SOC optimal domain and obtaining an energy map and an SOC trajectory by training and testing the energy map drawing model include: Assigning 1 to the optimal state point in the SOC optimal domain and 0 to the remaining points, thereby obtaining an energy map in the form of a SOC optimal domain 01 matrix; Create multiple input features; Building an energy map drawing model based on the network structure of the spatial perception network, substituting the plurality of input features into the input end of the energy map drawing model, and substituting the corresponding energy map into the output end of the energy map drawing model; The energy map and the SOC trajectory are obtained through data training and testing.

9. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 8, characterized in that: The multiple input features include X=(v, a, t, Preq p ,Preq n ,SOC); Where v represents the vehicle speed at time t, a represents the acceleration, and Preq p represents the sum of the cumulative positive power demand at time t, Preq n It represents the sum of the cumulative negative power demand at time t, t represents time, and SOC represents the feasible region of state of charge.

10. The vehicle global optimization energy management method based on energy map real-time algorithm for hybrid electric vehicles according to claim 1, characterized in that: The steps of constructing an optimal SOC trajectory generation model according to the energy map and the SOC trajectory, correcting the SOC trajectory using the optimal SOC trajectory generation model, and obtaining an SOC reference trajectory include: According to the energy map The probability matrix of According to the Obtain the maximum value of each column of the probability matrix, and obtain the SOC trajectory in the time dimension based on the maximum value; Create multiple input features; An optimal SOC trajectory generation model is built based on a time series network architecture, and multiple input features are substituted into the input end of the optimal SOC trajectory generation model. The output of the optimal SOC trajectory generation model is the SOC reference trajectory. The multiple input features X include X=(v,a,t,Preq p ,Preq n ,SOC peak ); Where v represents the vehicle speed at time t, a represents the acceleration, and Preq p represents the sum of the cumulative positive power demand at time t, Preq n It represents the sum of the cumulative negative power demand at time t, where t represents time and SOC peak Represents the SOC trajectory output by the energy map model.

Images