Automobile driving control method and device based on dynamic planning, controller and readable storage medium

Abstract

The invention relates to an automobile driving control method and device based on dynamic planning, a controller and a computer readable storage medium. The method comprises the following steps: acquiring statistical information of battery energy change in the driving process of the extended-range automobile; gaussian iterative transformation processing is carried out based on the distribution information, and Gaussian distribution information corresponding to the battery energy change of the extended-range automobile is obtained; determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval; according to the target energy change unit and a preset correlation model, determining the discrete grid density of the battery state of charge of the extended-range automobile; and performing energy control on the extended-range automobile based on a preset dynamic planning algorithm and the discrete grid density. By adopting the method, the calculation efficiency of a dynamic programming algorithm can be improved, so that better range-extended automobile energy control is realized.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a vehicle driving control method, device, controller, and computer-readable storage medium based on dynamic programming. Background Art

[0002] Energy control for extended-range electric vehicles (EREVs) aims to optimize the vehicle's powertrain efficiency or fuel economy over a specific timeframe by controlling the powertrain components. Energy control strategies based on dynamic programming (DP) can theoretically achieve the globally optimal fuel economy, serving as a benchmark for measuring the energy-saving effectiveness of other control strategies. However, EREVs, due to their dual motors and engine, have a large number of state and control variables, resulting in poor computational efficiency in dynamic programming algorithms, which in turn impacts EREV energy control. Summary of the Invention

[0003] Based on this, it is necessary to provide a vehicle driving control method, device, controller and computer-readable storage medium based on dynamic programming to address the above technical problems, so as to improve the computational efficiency of the dynamic programming algorithm and thus achieve better energy control of extended-range vehicles.

[0004] In a first aspect, the present application provides a vehicle driving control method based on dynamic programming, comprising:

[0005] Obtain statistical information on battery state of charge changes during driving of extended-range vehicles;

[0006] Based on statistical information and a preset correlation model, determine the distribution information of the frequency of battery energy changes of the extended-range vehicle as the battery state of charge changes; wherein the correlation model is used to characterize the correlation between the battery energy changes and the battery state of charge changes of the extended-range vehicle;

[0007] Performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval;

[0008] Determine the discrete grid density of the battery state of charge of the range-extended vehicle based on the target energy change unit;

[0009] Energy control of range-extended vehicles based on a preset dynamic programming algorithm and discrete grid density;

[0010] The expression corresponding to the association model is:

[0011]

[0012] in, Indicates the change in battery energy within the same time interval;

[0013] Indicates the next moment and the current moment SOC changes between

[0014] is the reference voltage;

[0015] is the reference capacity of the battery;

[0016] Indicates the total number of single cells connected in series in the battery pack;

[0017] Indicates the maximum power of the motor;

[0018] Indicates battery power efficiency.

[0019] In one embodiment, performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information includes:

[0020] Perform Gaussian iterative transformation on the distribution information;

[0021] After each iteration, the skewness index of the current transformation is determined;

[0022] When the skewness index does not meet the tolerance condition, the transformation coefficient in the Gaussian iterative transformation is adjusted according to the skewness index;

[0023] The next iteration is performed based on the adjusted transformation coefficient until the skewness index satisfies the tolerance condition, and Gaussian distribution information corresponding to the distribution information is obtained.

[0024] In one embodiment, the skewness index is calculated as:

[0025] , ,

[0026] in, represents the skewness index;

[0027] represents the transform coefficient;

[0028] represents the average value of the battery energy change distribution in the statistical information;

[0029] Indicates the distribution of battery energy changes in the statistical information The actual value of the samples;

[0030] is the number of values ​​of battery energy change distribution in the statistical information;

[0031] The initial value is set to =1.1, With each iteration increase by 0.1;

[0032] is a fixed value of 0.1, is the number of iterations;

[0033] According to the skewness index, the transformation coefficients in the Gaussian iterative transformation are adjusted, including:

[0034] The changing trend of the skewness index is determined, and the transformation coefficients in the Gaussian iterative transformation are adjusted based on the changing trend.

[0035] In one embodiment, when the skewness index does not satisfy the tolerance condition, adjusting the transformation coefficients in the Gaussian iterative transformation according to the skewness index includes:

[0036] When the skewness index satisfies the tolerance inequality, it is determined that the skewness index does not meet the tolerance condition; wherein the tolerance inequality is:

[0037] ;in, represents the Gaussian distribution standardized variable, i.e. the tolerance level;

[0038] When the skewness index does not satisfy the tolerance condition, the transformation coefficients in the Gaussian iterative transformation are adjusted according to the skewness index.

[0039] In one embodiment, determining the change confidence interval from Gaussian distribution information according to a preset tolerance condition includes:

[0040] Determine the mean and standard deviation of the distribution values ​​in the Gaussian distribution information;

[0041] Determine the confidence interval of the Gaussian distribution information based on the preset tolerance conditions, mean value and standard deviation.

[0042] In one embodiment, determining a confidence interval of Gaussian distribution information based on a preset tolerance condition, a mean value, and a standard deviation includes:

[0043] Determine the confidence interval for the Gaussian distribution information based on the following expression:

[0044] ,in,

[0045] represents the confidence interval of the change;

[0046] represents the average value;

[0047] Indicates the preset tolerance condition;

[0048] Represents standard deviation.

[0049] In a second aspect, the present application further provides a vehicle driving control device based on dynamic programming, comprising:

[0050] An acquisition module is used to obtain statistical information about the change in battery state of charge of the extended-range vehicle during driving;

[0051] a determination module for determining, based on statistical information and a preset correlation model, distribution information of the frequency of battery energy changes as the battery state of charge of the extended-range vehicle changes; wherein the correlation model is used to characterize the correlation between the battery energy change and the battery state of charge change of the extended-range vehicle; performing a Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval; and determining a discrete grid density of the battery state of charge of the extended-range vehicle based on the target energy change unit;

[0052] Among them, the expression corresponding to the association model is:

[0053]

[0054] in, Indicates the change in battery energy within the same time interval;

[0055] Indicates the next moment and the current moment SOC changes between

[0056] is the reference voltage;

[0057] is the reference capacity of the battery;

[0058] Indicates the total number of single cells connected in series in the battery pack;

[0059] Indicates the maximum power of the motor;

[0060] represents the battery power efficiency;

[0061] The control module is used to perform energy control on the range-extended vehicle based on a preset dynamic programming algorithm and discrete grid density.

[0062] In a third aspect, the present application further provides a controller comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method in the first aspect when executing the computer program.

[0063] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method in the first aspect when executed by a processor.

[0064] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, which implements the steps of the method in the first aspect when executed by a processor.

[0065] The above-described dynamic programming-based vehicle driving control method, device, controller, and computer-readable storage medium determine a target energy change unit based on statistical information about battery state of charge changes during driving of an extended-range vehicle. Because the target energy change unit is determined based on a change confidence interval, which is determined based on Gaussian distribution information corresponding to the distribution information of the battery energy change frequency as the battery state of charge changes, the target energy change unit can reflect the actual battery energy changes of the extended-range vehicle from the perspective of a Gaussian distribution. Based on the target energy change unit and the associated model, the discrete grid density of the battery state of charge is determined. This ensures that the discrete grid density follows the laws of the Gaussian distribution, thereby avoiding the increase in computational complexity and decrease in computational efficiency caused by an excessively low discrete grid density for the battery state of charge, and also avoiding the poor computational accuracy caused by an excessively high battery state of charge. Energy control of the extended-range vehicle is performed according to a preset dynamic programming algorithm and discrete grid density, which improves the computational efficiency of the dynamic programming algorithm and achieves better energy control of the extended-range vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0067] Figure 1 A schematic diagram of a possible transmission structure of a range-extended vehicle in one embodiment;

[0068] Figure 2 1 is a flow chart of a vehicle driving control method based on dynamic programming in one embodiment;

[0069] Figure 3 is another flow chart of a vehicle driving control method based on dynamic programming in one embodiment;

[0070] Figure 4 : is an energy distribution diagram after Gaussian iterative transformation under WLTP conditions in one embodiment;

[0071] Figure 5 A partial schematic diagram of error statistics of a vehicle driving control method based on dynamic programming in one embodiment;

[0072] Figure 6 is a structural block diagram of a vehicle driving control device based on dynamic programming in one embodiment;

[0073] Figure 7 FIG. 4 is a diagram showing the internal structure of a controller in one embodiment. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0075] An extended-range electric vehicle (EV) is equipped with an onboard auxiliary power generation system (also known as a range extender). This system consists of an engine, generator, and controller. When the onboard rechargeable energy storage system cannot meet the vehicle's driving requirements, the range extender provides power to the vehicle's powertrain. A range extender generally refers to an EV component that provides additional power, thereby increasing the EV's range. Traditionally, a range extender refers to a combination of an engine and a generator.

[0076] Extended-range electric vehicles (EREVs), combining the advantages of both pure electric vehicles and hybrid vehicles, are gaining increasing market recognition. Energy management addresses the issue of optimizing the vehicle's transmission system's operating efficiency or fuel economy over a specific timeframe through the control of powertrain components. The energy transfer process in EREVs involves both mechanical and electrical systems, and coordinating these two systems for efficient and stable operation is a current research hotspot. Currently, energy control strategies based on dynamic programming (DP) algorithms can theoretically achieve globally optimal fuel economy, serving as a benchmark for measuring the energy-saving effectiveness of other control strategies. However, EREVs, due to their dual motors and engine, have a large number of state and control variables and a high control dimensionality. For traditional DP algorithms seeking global optimality, the computational load and time increase exponentially with the number of dimensionality, making them difficult to meet computational efficiency requirements. During parameter optimization, as the number of candidate optimization parameters (transmission and component parameters) increases, the design space also explodes. Therefore, the EREV parameter optimization and matching process places high demands on the algorithm's computational efficiency. Developing an efficient, adaptive dynamic programming-based energy management method for range-extended vehicles (REVs) is a key technical challenge currently in need of breakthroughs. To rapidly evaluate fuel economy under a wide range of parameters and fuel consumption under various operating conditions, as well as to quickly assess and calibrate the range extender (REV) match, a key technical challenge. However, the numerous variables and complex control methods inherent in REV energy management methods can lead to a slow solution process, making it difficult to achieve both optimal optimization efficiency and optimal results. Therefore, designing an energy management method that balances computational efficiency and optimization performance is of great scientific and engineering value.

[0077] Based on the above analysis, the present application provides a vehicle driving control method based on dynamic programming to improve the computational efficiency of the dynamic programming algorithm, thereby achieving better energy control of extended-range vehicles.

[0078] The technical solutions provided by this application are described below by way of examples:

[0079] In one embodiment, a vehicle driving control method based on dynamic programming is provided. This embodiment takes the application of this method to an extended-range electric vehicle as an example for explanation. Figure 1 As shown in the figure, a possible schematic diagram of the structure of an extended-range electric vehicle is provided. In the figure, 1 represents the battery, 2 represents the inverter, 3 represents the second motor, 4 represents the gear, 5 represents the gear, 6 represents the wheel, 7 represents the differential, 8 represents the engine, 9 represents the gear, and 10 represents the first motor. An extended-range electric vehicle can have two modes: pure electric drive and extended-range drive. Figure 2 As shown, the range-extended vehicle energy control method may include steps S201 to S205:

[0080] Step S201: Obtain statistical information on changes in the battery state of charge of the extended-range vehicle during driving.

[0081] The battery state of charge change statistical information may be information generated by collecting statistics on related situations such as a decrease or increase in the battery state of charge in the extended-range vehicle.

[0082] In some embodiments, the statistical information may be updated in real time.

[0083] In some embodiments, the extended-range vehicle may have at least two driving modes, such as a pure electric driving mode and an extended-range driving mode.

[0084] In some embodiments, the driving process can be broadly understood to include various operating conditions of the extended-range vehicle, such as normal driving, temporary parking, parking standby, etc. In some embodiments, whether the extended-range vehicle is in the driving process can be determined based on whether the electronic control unit in the vehicle is powered on.

[0085] In some embodiments, the range-extended vehicle is equipped with an engine to charge the battery. When the battery is fully charged, the range-extended vehicle is powered by electricity and does not consume fuel. When the battery charge drops below a certain level, the engine starts to charge the battery, thereby consuming fuel.

[0086] Step S202: Based on the statistical information and a preset correlation model, determine the distribution information of the battery energy change frequency of the extended-range vehicle as the battery state of charge changes; wherein the correlation model is used to characterize the correlation relationship between the battery energy change and the battery state of charge change of the extended-range vehicle.

[0087] The Gaussian iterative transformation process may be a nonlinear transformation method based on a Gaussian function. Specifically, the Gaussian iterative transformation process may refer to a process of gradually adjusting the distribution of statistical information by repeatedly applying nonlinear transformations related to the Gaussian function, ultimately causing it to approach or approximate a Gaussian distribution (normal distribution).

[0088] In some embodiments, the Gaussian iterative transformation process can be used to smooth statistical information, emphasize or suppress the probability density of certain regions, improve sampling efficiency, or accelerate convergence.

[0089] Step S203: performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval.

[0090] The tolerance condition can be a choice that represents the confidence level, reflecting the reliability or confidence in the desired estimated interval. In other words, the tolerance condition actually refers to the acceptable range of error or the tolerance for uncertainty in the estimated result. This is usually expressed by setting a confidence level. The confidence level is the degree of confidence that the confidence interval contains the true value, such as 95%.

[0091] A confidence interval (CI) is a range used in statistics to estimate a population parameter. It represents the range within which a population parameter is likely to fall, given a certain confidence level. For example, a confidence interval can be a range, such as [45, 55]. Therefore, there is a 95% confidence that this interval contains the true population parameter (e.g., the population mean).

[0092] In some embodiments, based on the premise that the distribution information of the known battery energy change frequency as the battery state of charge changes follows a Gaussian distribution (i.e., Gaussian distribution information), an interval can be calculated according to a specific confidence level (e.g., 95%) to estimate the possible value range of the actual energy change, that is, to determine the change confidence interval.

[0093] The target energy change unit may be an energy change unit representing a specific energy change value determined from a change confidence interval.

[0094] In some embodiments, the target energy change unit may be determined based on the median, minimum, mode, or mean of the change confidence interval.

[0095] Step S204: Determine the discrete grid density of the battery state of charge of the range-extended vehicle according to the target energy change unit and the preset correlation model.

[0096] A battery's state of charge (SOC) refers to the ratio of a battery's remaining capacity after a period of use or long-term disuse to its fully charged capacity. It's often expressed as a percentage. For example, the SOC value can range from 0 to 1, with SOC = 0 indicating a fully discharged battery and SOC = 1 indicating a fully charged battery.

[0097] In some embodiments, the SOC can be discretized. When performing complex optimization calculations, directly processing continuous variables increases computational difficulty and resource consumption. By dividing the SOC interval into several discrete points (i.e., a discrete grid), the problem can be simplified, making it more suitable for numerical solutions.

[0098] The discrete grid density refers to the size of the interval between discrete points or the number of discrete points contained in a unit length when a continuous variable (such as the battery state of charge) is divided into several discrete points.

[0099] In some embodiments, a higher grid density means finer division, that is, more discrete points and smaller intervals, and correspondingly more computational effort; a lower grid density means the opposite.

[0100] In some embodiments, the correlation between the change in battery energy and the change in battery state of charge of the range-extended vehicle may be a linear relationship or a nonlinear relationship.

[0101] Step S205: Based on a preset dynamic programming algorithm and discrete grid density, energy control is performed on the range-extended vehicle.

[0102] Among them, the dynamic programming (DP) algorithm can be an algorithm that solves complex problems by decomposing the original problem into relatively simple sub-problems.

[0103] In some embodiments, a dynamic programming algorithm may be called based on the discrete grid density of the battery state of charge determined in the aforementioned steps to achieve energy control of the range-extended vehicle.

[0104] The above technical solution determines a target energy change unit based on statistical information about battery state of charge (SOC) changes during driving. Because the target energy change unit is determined based on a change confidence interval, which is determined based on Gaussian distribution information corresponding to the distribution of the frequency of battery energy changes as the battery SOC changes, the target energy change unit can reflect the actual battery energy changes of the extended-range vehicle from the perspective of a Gaussian distribution. Based on the target energy change unit and the associated model, the discrete grid density of the battery SOC is determined. This ensures that the discrete grid density follows the laws of the Gaussian distribution, thereby avoiding the increase in computational complexity and decrease in computational efficiency caused by an excessively low discrete grid density for the battery SOC, and also avoiding the poor computational accuracy caused by an excessively high SOC. Energy control of the extended-range vehicle is performed based on a preset dynamic programming algorithm and discrete grid density, which improves the computational efficiency of the dynamic programming algorithm and achieves better energy control for the extended-range vehicle.

[0105] In one embodiment, the aforementioned "performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information" may include: performing Gaussian iterative transformation on the distribution information; determining the skewness index after the current transformation after each iteration; when the skewness index does not meet the tolerance condition, adjusting the transformation coefficient in the Gaussian iterative transformation according to the skewness index; performing the next iteration based on the adjusted transformation coefficient until the skewness index meets the tolerance condition, thereby obtaining the Gaussian distribution information corresponding to the distribution information.

[0106] The skewness index may be an index used to describe skewness, and skewness may be a statistic that describes the degree of asymmetry of data distribution.

[0107] In some embodiments, skewness = 0: indicates that the distribution is symmetrical (such as a standard normal distribution); skewness > 0: right skewed (long tail on the right); skewness < 0: left skewed (long tail on the left).

[0108] In some embodiments, after each Gaussian transformation iteration, the skewness index of the current data distribution is calculated to assess whether the transformation has made the distribution closer to a Gaussian distribution (i.e., whether the skewness approaches 0). This is because the Gaussian distribution is symmetrical and has a skewness of 0. Therefore, monitoring changes in the skewness during the iteration process can help determine whether the current transformation is effective, whether it is sufficiently close to a Gaussian distribution, and whether the transformation parameters (e.g., the transformation coefficient m) need to be adjusted.

[0109] The above technical solution performs Gaussian iterative transformation on the distribution information and determines the skewness index after the current transformation after each iteration; if the skewness index does not meet the tolerance condition, the transformation coefficient in the Gaussian iterative transformation is adjusted and the iteration is continued until the skewness index meets the tolerance condition, thereby obtaining more accurate Gaussian distribution information.

[0110] In one embodiment, the calculation formula of the aforementioned skewness index is:

[0111] , ,

[0112] in, represents the skewness index;

[0113] represents the transformation coefficient;

[0114] Indicates the average value of battery energy change distribution in statistical information;

[0115] Indicates the distribution of battery energy changes in the statistical information. The actual value of the samples;

[0116] is the number of values ​​of battery energy change distribution in the statistical information;

[0117] The initial value is set to =1.1, With each iteration increase by 0.1;

[0118] is a fixed value of 0.1, is the number of iterations;

[0119] According to the skewness index, the transformation coefficients in the Gaussian iterative transformation are adjusted, including:

[0120] The changing trend of the skewness index is determined, and the transformation coefficients in the Gaussian iterative transformation are adjusted based on the changing trend.

[0121] In one embodiment, the aforementioned “adjusting the transformation coefficients in the Gaussian iterative transformation according to the skewness index when the skewness index does not satisfy the tolerance condition” may include: determining that the skewness index does not satisfy the tolerance condition when the skewness index satisfies a tolerance inequality; wherein the tolerance inequality is:

[0122] ;in, represents the Gaussian distribution standardized variable, i.e. the tolerance level;

[0123] When the skewness index does not satisfy the tolerance condition, the transformation coefficients in the Gaussian iterative transformation are adjusted according to the skewness index.

[0124] In one embodiment, the expression corresponding to the aforementioned association model is:

[0125] in, Indicates the change in battery energy within the same time interval;

[0126] Indicates the next moment and the current moment SOC changes between

[0127] is the reference voltage;

[0128] is the reference capacity of the battery;

[0129] Indicates the total number of single cells connected in series in the battery pack;

[0130] Indicates the maximum power of the motor;

[0131] Indicates battery power efficiency.

[0132] In one embodiment, the aforementioned "determining the change confidence interval from the Gaussian distribution information according to the preset tolerance condition" may include: determining the mean value and standard deviation corresponding to the distribution values ​​in the Gaussian distribution information; determining the change confidence interval of the Gaussian distribution information based on the preset tolerance condition, mean value and standard deviation.

[0133] By determining the mean value and standard deviation corresponding to the distribution values ​​in the Gaussian distribution information, and according to the preset tolerance conditions, mean value and standard deviation, a more accurate confidence interval of the change of the Gaussian distribution information can be determined.

[0134] In one embodiment, the determining of the change confidence interval of the Gaussian distribution information based on a preset tolerance condition, the mean value, and the standard deviation includes: determining the change confidence interval of the Gaussian distribution information based on the following expression:

[0135] ,in, represents the confidence interval of the change; represents the average value; Indicates the preset tolerance condition; represents the standard deviation.

[0136] In an exemplary embodiment, a vehicle driving control method based on dynamic programming is provided. The method estimates the battery energy change at each time step (corresponding to step S303 below), performs statistical analysis on the estimated battery energy change, and determines the Gaussian distribution of the battery energy change frequency as the battery state of charge changes, thereby determining the optimal SOC discrete grid density, and further achieving energy control for the extended-range vehicle. Figure 3 As shown, the range-extended vehicle energy control method may include steps S301 to S306:

[0137] Step S301: establishing a longitudinal dynamics model of the vehicle and a dynamics model of the range extender according to the parameters of the vehicle.

[0138] According to the classical longitudinal dynamics formula, the driving force is equal to the sum of friction resistance, slope resistance, air resistance and acceleration resistance. On this basis, the longitudinal dynamics model of the car is established as follows:

[0139] , ,

[0140] in, Indicates the driving force of the vehicle (N=kg·m / s 2 ), Represents the vehicle's curb weight (kg), is the gravity coefficient, represents the rolling resistance coefficient, Represents the road slope during driving. represents the air resistance coefficient, Represents the frontal area of ​​the car (m 2 ), is the air density ( ), Represents vehicle speed (m / s). is the vehicle's rotational mass conversion factor, the above calculation The equation for is dimensionless. is the moment of inertia of the vehicle wheel end, is the moment of inertia of the flywheel, is the vehicle wheel radius, is the transmission efficiency, and are the transmission ratios of the transmission and the final drive, respectively.

[0141] By establishing the above model, a model foundation is laid for subsequent simulations, which serves the development of subsequent steps.

[0142] Step S302: Analyze the simulation accuracy sensitivity and calculation time of the dynamic programming algorithm under different SOC grid densities to determine the simplification requirements of the state variable space and the control variable space.

[0143] In some embodiments, different parameter grid sizes can be set, and the calculation time under different parameter grids can be calculated to determine the sensitivity; by finding the SOC grid density that balances the calculation time and simulation accuracy, it is used as a variable simplification requirement, so that the various variables involved can be simplified according to the variable simplification requirement to reduce the number of variables.

[0144] Step S303: Based on the statistical analysis of battery energy change frequency, establish energy change Changes with SOC The energy change distribution is calculated based on the correlation model.

[0145] Based on the statistical analysis of battery energy change frequency, establish energy change Changes with SOC The energy change distribution is calculated based on the correlation model of

[0146] ,

[0147] in, Indicates the change in battery energy within the same time interval;

[0148] Indicates the next moment and the current moment SOC changes between

[0149] is the reference voltage;

[0150] is the reference capacity of the battery;

[0151] Indicates the total number of single cells connected in series in the battery pack;

[0152] Indicates the maximum power of the motor;

[0153] Indicates battery power efficiency.

[0154] In some embodiments, the association model can be used to calculate Indicates the change in battery energy within the same time interval.

[0155] Step S304: Perform Gaussian iterative transformation on the statistical results of energy change distribution, and dynamically adjust the transformation coefficients The update rule is such that the skewness index Meet the preset tolerance conditions .

[0156] Perform Gaussian iterative transformation on the statistical results of energy change distribution and dynamically adjust the transformation coefficient The update rules include:

[0157] 1) When the skewness after transformation (i.e. skewness index) tends to decrease, increases value;

[0158] 2) When skewness When it tends to increase, decrease value to avoid divergence;

[0159] 3) The value increases or decreases according to Value and The distance is dynamically adjusted. The farther the distance, the greater the increase or decrease; the closer the distance, the smaller the increase or decrease.

[0160] Among them, skewness It is an indicator of distribution asymmetry. The Gaussian shape (complete symmetry) has a zero skewness value. After generating the energy transformation frequency distribution graph, the skewness index is calculated by the following formula :

[0161] ,

[0162] in, It is the coefficient that needs to be adjusted during the iterative process.

[0163] in, is the mean of the distribution, It is The actual value of the samples, is the number of distribution values.

[0164] In some embodiments, a negative skewness index represents negative asymmetry, while a positive skewness index results from the presence of positive asymmetry.

[0165] Check whether the following conditions are met:

[0166] ,

[0167] in, is an indicator of the overall skewness of the data. represents a Gaussian distribution standardized variable, which refers to the tolerance level. For example, the tolerance level a is set to 95% ( =1.96)、90%( =1.65) and 68% ( =1). By dynamically adjusting the transformation coefficient The update rule is such that the skewness index Meet the preset tolerance conditions .

[0168] Depending on the skewness value, if negative or positive asymmetry is found, the distribution is transformed as follows:

[0169] , ,

[0170] in, It is the coefficient that needs to be adjusted in the iterative process, and the initial value is set to =1.1, is a fixed value of 0.1, is the number of iterations. With each iteration it increases by 0.1.

[0171] like Figure 4 The figure shows a possible distribution diagram of the battery energy change frequency as a function of the battery state of charge after Gaussian iterative transformation under the WLTP (Worldwide Harmonised Light Vehicles Test Procedure) operating conditions. The vertical axis represents the battery energy change frequency, and the horizontal axis represents the battery state of charge change.

[0172] Step S305: Determine the optimal density of the SOC discrete grid based on the transformed Gaussian distribution confidence interval, so that the DP algorithm can reduce the computational complexity while ensuring the simulation accuracy.

[0173] In some embodiments, the optimal density of the SOC discrete grid is determined by the following parameters:

[0174] 1) Calculate the mean of the transformed statistical distribution and standard deviation ;

[0175] 2) According to the tolerance level a and the corresponding The confidence interval of the change is determined by the following formula :

[0176]

[0177] 3) The lower limit of the transformed confidence interval is chosen to be the minimum energy change that is statistically consistent in the batch of data and is identified in the original domain due to the inverse transformation of the distribution. For example, the minimum energy change value corresponding to the lower limit of the confidence interval is selected based on the values ​​corresponding to the confidence interval, such as 95% and 75%. As a benchmark for SOC grid discretization.

[0178] Select the minimum energy change corresponding to the lower limit of the confidence interval As a benchmark for SOC grid discretization, the benchmark is used for subsequent SOC discretization levels Calculation of the quantity.

[0179] 4) SOC discretization level for DP The number is finally calculated by the following formula:

[0180] ,

[0181] in, is the value of the energy step of the SOC discretization grid, that is, the benchmark determined above, is the portion of battery energy related to the usable SOC range (from arrive , that is, the upper and lower limits of the battery SOC fluctuation at the current moment. For extended-range vehicles, it is usually 20% to 80%).

[0182] The forward and reverse phases of the dynamic programming algorithm use the same SOC grid, and the rationality of the state grid selection is verified through interpolation error statistics. In some embodiments, the simulation accuracy and computational efficiency of the DP algorithm under different SOC grids can be analyzed by setting different SOC grid densities. If the SOC grid in the DP calculation is not within the specified SOC grid value, such as 0.512, or is not between 0.510 and 0.515 (the grid is selected at 0.05), interpolation is required to approximate it closer to 0.510.

[0183] The method is applied to extended-range electric vehicles, including but not limited to plug-in or conventional series hybrid architectures, and solves the "dimensionality explosion" problem of traditional dynamic programming algorithms by optimizing the SOC grid discretization.

[0184] The algorithm is applied to various driving cycle requirements, verifies the effectiveness of the operating point through dynamic simulation, and generates an optimal control strategy for the engine-motor collaboration. These multiple operating conditions include: WLTP, CLTC (China Light-Duty Vehicles Test Cycle), a light-duty vehicle driving cycle standard developed by the China Automotive Research Center under the commission of the Ministry of Industry and Information Technology of China and adapted to China's road traffic conditions and driving habits; NEDC (New European Driving Cycle), a European endurance standard test cycle; Japan 1015, a comprehensive Japanese test cycle; FTP72 (Federal Test Procedure), a standard issued by the U.S. Department of Energy for testing the economy and emissions of passenger vehicles in urban conditions; and HWEET (Highway Fuel Economy Test), a U.S. highway test cycle.

[0185] Step S306: constraining the working states of the components of the transmission system, and calculating the optimal power distribution while ensuring that the constraints are effective.

[0186] The specific constraints on the working status of each component during the operation of the transmission system are:

[0187] Where, and , Represent the speed and torque of the engine and motor respectively, Indicates the charge and discharge power of the battery. Indicates minimum, Indicates the maximum, Indicates the current moment. To ensure the normal operation of the range extender system and the life of the battery, the battery power level must always fluctuate within an appropriate range during operation. It should not be too low to be depleted, nor too full.

[0188] In some embodiments, such as Figure 5 As shown in the figure, a possible error statistics local diagram corresponding to the above-mentioned extended range vehicle energy control method is given. This diagram can be a discrete diagram of the SOC grid calculation, which is an approximation of the grid division value. In the figure, the horizontal axis is the time axis, which is used to represent the discrete time step. For the current moment, is the next moment. The vertical axis is the SOC grid node, Indicates the current moment time step, The state of charge value of each SOC grid node. Indicates that at the current moment, The following is taken A control strategy ( =1, 2, ....). error represents the error statistic, Represents a strategy-node combination, i.e. The grid nodes use The error when using the same control strategy; min dts represents the minimum discretization step.

[0189] The above technical solution has the following beneficial effects: First, an adaptive DP algorithm that can simplify the state space is proposed. This method adaptively discretizes the DP SOC state grid based on the statistical method of battery energy change frequency, achieving the purpose of reducing data dimensions to simplify calculations. While ensuring calculation accuracy, this algorithm significantly improves the calculation efficiency of fuel economy in configuration evaluation; second, it solves the problem of multiple variables and complex control in the energy management method of extended-range electric vehicles, which leads to a slow solution process, and can simultaneously ensure that the optimization efficiency and optimization results are close to optimal; third, the proposed adaptive DP offline control strategy can significantly shorten the simulation time required for parameter optimization and correction of extended-range electric vehicles while ensuring high accuracy.

[0190] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily performed in sequence in the order indicated by the arrows. Unless clearly stated herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of the steps or stages in other steps or other steps. It is understandable that the various steps in different embodiments can be freely combined as needed, and the various non-contradictory schemes formed by the combination all fall within the scope of protection of this application.

[0191] Based on the same inventive concept, embodiments of the present application also provide a vehicle driving control device based on dynamic programming for implementing the above-mentioned extended-range vehicle energy control method. The solution provided by this device is similar to the solution described in the above-mentioned method. Therefore, the specific limitations of one or more embodiments of the vehicle driving control device based on dynamic programming provided below can be found in the above-mentioned limitations of the extended-range vehicle energy control method and will not be repeated here.

[0192] In an exemplary embodiment, Figure 6 As shown, a vehicle driving control device 600 based on dynamic programming is provided, comprising:

[0193] An acquisition module 601 is used to obtain statistical information about the state of charge of the battery of the extended-range vehicle during driving;

[0194] Determination module 602 is configured to determine, based on statistical information and a preset correlation model, distribution information of the frequency of battery energy changes as a function of battery state of charge for the extended-range vehicle; wherein the correlation model is configured to characterize the correlation between battery energy changes and battery state of charge changes for the extended-range vehicle; perform a Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; determine a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determine a target energy change unit from the change confidence interval; and determine a discrete grid density for the battery state of charge of the extended-range vehicle based on the target energy change unit.

[0195] Among them, the expression corresponding to the association model is:

[0196] in, Indicates the change in battery energy within the same time interval;

[0197] Indicates the next moment and the current moment SOC changes between

[0198] is the reference voltage;

[0199] is the reference capacity of the battery;

[0200] Indicates the total number of single cells connected in series in the battery pack;

[0201] Indicates the maximum power of the motor;

[0202] represents the battery power efficiency;

[0203] The control module 603 is used to perform energy control on the range-extended vehicle based on a preset dynamic programming algorithm and discrete grid density.

[0204] In one embodiment, the determination module 602 is also used to perform Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information, including: performing Gaussian iterative transformation on the distribution information; determining the skewness index after the current transformation after each iteration; when the skewness index does not meet the tolerance condition, adjusting the transformation coefficient in the Gaussian iterative transformation according to the skewness index; performing the next iteration based on the adjusted transformation coefficient until the skewness index meets the tolerance condition, thereby obtaining the Gaussian distribution information corresponding to the distribution information.

[0205] In one embodiment, the skewness index is calculated as:

[0206] , ,

[0207] in, represents the skewness index;

[0208] represents the transform coefficient;

[0209] represents the average value of the battery energy change distribution in the statistical information;

[0210] Indicates the distribution of battery energy changes in the statistical information The actual value of the samples;

[0211] is the number of values ​​of battery energy change distribution in the statistical information;

[0212] The initial value is set to =1.1, With each iteration increase by 0.1;

[0213] is a fixed value of 0.1, is the number of iterations;

[0214] According to the skewness index, the transformation coefficients in the Gaussian iterative transformation are adjusted, including:

[0215] The changing trend of the skewness index is determined, and the transformation coefficients in the Gaussian iterative transformation are adjusted based on the changing trend.

[0216] In one embodiment, the determining module 602 is further configured to adjust the transformation coefficients in the Gaussian iterative transformation according to the skewness index when the skewness index does not satisfy the tolerance condition, including: determining that the skewness index does not satisfy the tolerance condition when the skewness index satisfies the tolerance inequality; wherein the tolerance inequality is:

[0217] ;in, represents the Gaussian distribution standardized variable, i.e. the tolerance level;

[0218] When the skewness index does not satisfy the tolerance condition, the transformation coefficients in the Gaussian iterative transformation are adjusted according to the skewness index.

[0219] In one embodiment, the determination module 602 is further used to determine a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, including: determining the mean value and standard deviation corresponding to the distribution values ​​in the Gaussian distribution information; and determining the change confidence interval of the Gaussian distribution information based on the preset tolerance condition, mean value, and standard deviation.

[0220] In one embodiment, the determination module 602 is further configured to determine the confidence interval of the Gaussian distribution information based on a preset tolerance condition, a mean value, and a standard deviation, including determining the confidence interval of the Gaussian distribution information based on the following expression: ,in, represents the confidence interval of the change; represents the average value; Indicates the preset tolerance condition; Represents standard deviation.

[0221] Each module in the aforementioned dynamic programming-based vehicle driving control device can be implemented in whole or in part through software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor within a controller in hardware form, or stored in a memory within the controller in software form, allowing the processor to call and execute the corresponding operations of each module.

[0222] In an exemplary embodiment, a controller is provided. The controller may be a server, and its internal structure diagram may be as shown in FIG. Figure 7 As shown. The controller includes a processor, a memory, an input / output interface (I / O) and a communication interface. The processor, memory and input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the controller is used to provide computing and control capabilities. The memory of the controller includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the controller is used to store data required for executing the extended-range vehicle energy control method, such as statistical information on battery energy changes. The input / output interface of the controller is used to exchange information between the processor and external devices. The communication interface of the controller is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a vehicle driving control method based on dynamic programming is implemented.

[0223] Those skilled in the art will understand that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the controller to which the solution of the present application is applied. The specific controller may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0224] In an exemplary embodiment, a controller is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0225] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0226] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0227] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0228] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0229] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A vehicle driving control method based on dynamic programming, characterized in that: The method comprises: Obtain statistical information on battery state of charge changes during driving of extended-range vehicles; Based on the statistical information and a preset correlation model, determining distribution information of the frequency of battery energy changes of the extended-range vehicle as the battery state of charge changes; wherein the correlation model is used to characterize the correlation between the battery energy changes and the battery state of charge changes of the extended-range vehicle; Performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; Determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval; determining a discrete grid density of a battery state of charge of the range-extended vehicle according to the target energy change unit; Performing energy control on the range-extended vehicle based on a preset dynamic programming algorithm and the discrete grid density; The expression corresponding to the association model is: in, Indicates the change in battery energy within the same time interval; Indicates the next moment and the current moment SOC changes between is the reference voltage; is the reference capacity of the battery; Indicates the total number of single cells connected in series in the battery pack; Indicates the maximum power of the motor; Indicates battery power efficiency.

2. The method according to claim 1, characterized in that Performing Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information includes: Performing Gaussian iterative transformation on the distribution information; After each iteration, the skewness index of the current transformation is determined; When the skewness index does not satisfy the tolerance condition, adjusting the transformation coefficients in the Gaussian iterative transformation according to the skewness index; The next iteration is performed based on the adjusted transformation coefficient until the skewness index satisfies the tolerance condition, thereby obtaining Gaussian distribution information corresponding to the distribution information.

3. The method according to claim 2, characterized in that The calculation formula of the skewness index is: ， ， in, represents the skewness index; represents the transform coefficient; represents the average value of the battery energy change distribution in the statistical information; Indicates the distribution of battery energy changes in the statistical information The actual value of the samples; is the number of values ​​of battery energy change distribution in the statistical information; The initial value is set to =1.1, With each iteration increase by 0.1; is a fixed value of 0.1, is the number of iterations; Adjusting the transformation coefficients in the Gaussian iterative transformation according to the skewness index includes: A change trend of the skewness index is determined, and a transformation coefficient in the Gaussian iterative transformation is adjusted based on the change trend.

4. The method according to claim 3, characterized in that When the skewness index does not satisfy the tolerance condition, adjusting the transformation coefficient in the Gaussian iterative transformation according to the skewness index includes: When the skewness index satisfies the tolerance inequality, it is determined that the skewness index does not satisfy the tolerance condition; wherein the tolerance inequality is: ； in, represents the Gaussian distribution standardized variable, i.e. the tolerance level; When the skewness index does not satisfy the tolerance condition, the transformation coefficients in the Gaussian iterative transformation are adjusted according to the skewness index.

5. The method according to claim 1, wherein Determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition includes: Determining the mean and standard deviation of the distribution values ​​in the Gaussian distribution information; A confidence interval of a change in the Gaussian distribution information is determined based on a preset tolerance condition, the mean value, and the standard deviation.

6. The method according to claim 5, characterized in that Determining a change confidence interval of the Gaussian distribution information according to a preset tolerance condition, the mean value, and the standard deviation includes: The confidence interval of the Gaussian distribution information is determined based on the following expression: ,in, represents the confidence interval of the change; represents the average value; Indicates the preset tolerance condition; represents the standard deviation.

7. A vehicle driving control device based on dynamic programming, characterized in that: The device comprises: An acquisition module is used to obtain statistical information about the change in battery state of charge of the extended-range vehicle during driving; a determination module for determining, based on the statistical information and a preset correlation model, distribution information of the frequency of battery energy changes of the extended-range vehicle as the battery state of charge changes; wherein the correlation model is used to characterize the correlation between the battery energy change and the battery state of charge change of the extended-range vehicle; performing a Gaussian iterative transformation on the distribution information to obtain Gaussian distribution information corresponding to the distribution information; determining a change confidence interval from the Gaussian distribution information according to a preset tolerance condition, and determining a target energy change unit from the change confidence interval; and determining a discrete grid density of the battery state of charge of the extended-range vehicle based on the target energy change unit; The expression corresponding to the association model is: in, Indicates the change in battery energy within the same time interval; Indicates the next moment and the current moment SOC changes between is the reference voltage; is the reference capacity of the battery; Indicates the total number of single cells connected in series in the battery pack; Indicates the maximum power of the motor; represents the battery power efficiency; A control module is used to perform energy control on the range-extended vehicle based on a preset dynamic programming algorithm and the discrete grid density.

8. A controller comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images