Whole vehicle energy management method for pure electric commercial vehicle

Abstract

The invention discloses a whole vehicle energy management method for a pure electric commercial vehicle. Aiming at the defects in the prior art, the invention provides a new energy management strategy which integrates the advantages of a rule-based strategy and an optimal control-based strategy. By introducing a hybrid strategy combining rolling horizon optimization and rule control, the real-time performance can be ensured, and meanwhile, the optimality of energy distribution and the adaptive capacity of the system can be improved. Especially when the energy of the battery is low, the system economy can be effectively improved through the optimal control strategy, and the limitation that a traditional rule-based strategy is poor in adaptability and insufficient in economy under complex working conditions is overcome. In addition, by introducing the man-machine interaction module, dynamic cooperation of vehicle energy management and driver demands is achieved, and the driving experience and the energy efficiency management effect are further improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy management of new energy vehicles, and specifically relates to a whole vehicle energy management method for a pure electric commercial vehicle. Background Art

[0002] Pure electric vehicles have significant advantages in environmental friendliness and energy efficiency, but their widespread application still faces multiple technical challenges, especially in energy management. Energy management refers to improving the economy and efficiency of the entire system through the optimized management of energy acquisition, storage, distribution and recovery. In electric vehicles, reasonable energy management can not only optimize the battery charging and discharging process and extend the driving range, but also significantly improve the overall energy efficiency and reduce energy consumption and costs.

[0003] At present, the energy management strategy of electric vehicles is mainly based on information such as the real-time charging and discharging capacity of the battery, the use status of vehicle accessories, the driving conditions of the vehicle, and the driver's operating intentions, and dynamically allocates the energy use of multiple on-board power modules such as the motor, battery thermal management system, and low-voltage power system. Its goal is to achieve the vehicle's energy-saving goals, improve endurance, and further improve driving economy.

[0004] The existing pure electric vehicle energy management strategies can generally be divided into two categories: Rule-Based Energy Management Strategy (RB-EMS) and Optimal Control-based Energy Management Strategy. Existing technology 1: rule-based energy management strategy Technical Solution

[0005] The rule-based energy management strategy is to manage energy by defining energy allocation rules by technicians or experts based on the vehicle's power performance requirements, the working characteristics and high-efficiency range of each component, the characteristics of driving conditions, and rich engineering experience. This method usually adjusts the power allocation of each module in real time according to a series of preset rules to meet the energy needs of the vehicle under different working conditions. This strategy is simple to implement and has low requirements on the computing power of the controller. It is one of the more mature control methods currently used.

[0006] 2. Disadvantages Although rule-based energy management strategies are widely used, they also have some significant disadvantages: The energy allocation scheme lacks optimality: Regularized strategies are usually designed based on experience and specific working conditions. They cannot provide the optimal energy allocation scheme under all working conditions and often fail to fully tap the maximum potential of the system. Difficulty in parameter calibration: This strategy requires the calibration of a large number of parameters according to different driving environments, working conditions and actual vehicle status, which increases the complexity and maintenance cost in practical applications; Insufficient adaptability: For complex and ever-changing driving environments and complex power system architectures, rule-based strategies cannot be adjusted in real time, have poor adaptability, and are unable to cope with dynamically changing energy demands.

[0007] Existing technology 2: Energy management strategy based on optimal control Technical Solution

[0008] The energy management strategy based on optimal control establishes a mathematical model of the vehicle power system and solves a constrained optimal control problem to achieve the optimal energy economy and power distribution rationality. This method models factors such as vehicle driving conditions and equipment status in the vehicle system, sets optimization goals and introduces constraints, and uses direct or indirect methods to iteratively approximate the optimal solution to obtain the best energy distribution strategy. This type of method can more accurately meet the predetermined economy and power distribution goals.

[0009] 2. Disadvantages Although the energy management strategy based on optimal control can provide a more accurate energy management solution, there are also the following major problems in its application: High computational complexity: With the increase of constraints and state dimensions, the solution process of the optimal control problem involves a large number of complex calculations, which leads to a significant increase in the amount of calculation, thus affecting the system's operational efficiency and real-time performance, making it difficult to meet the requirements of real-time allocation and regulation of the vehicle energy system; Easy to fall into local optimality: The optimal solution obtained based on the optimal control algorithm may be a local optimal solution rather than a global optimal solution, and thus cannot adapt well to all driving conditions and dynamically changing energy requirements. In summary, the energy management strategies for pure electric vehicles currently proposed mainly focus on optimizing the power conversion efficiency or the real-time optimization of energy loss of a few important power-consuming modules. Although this strategy design method can improve the economic efficiency of important power-consuming modules, it does not coordinate the energy consumption relationship of other vehicle power-consuming modules such as the battery thermal management system, cabin thermal management system, and low-voltage power system. Summary of the invention

[0010] The present invention aims to overcome the shortcomings of the prior art and provide a pure electric commercial vehicle whole vehicle energy management method that integrates rolling time domain optimization and rule control, which can dynamically adjust power distribution according to different driving conditions, improve the adaptability of the system under complex conditions, and ensure that the system stability can be maintained when the optimal control strategy fails.

[0011] In order to solve the above technical problems, the technical method adopted by the present invention is as follows: The present invention discloses a vehicle energy management method for a pure electric commercial vehicle, comprising the following steps: S1. Obtain the driver's required power information through the driving demand collection module, including: The vehicle speed in the next 1-3 seconds is predicted based on the KNN algorithm, and the GADPC graph theory-density peak clustering algorithm is used to identify the driver's acceleration intention and calculate the driving power demand; Identify the braking intensity based on fuzzy reasoning rules and calculate the braking power demand in combination with the braking force distribution strategy; Read the real-time power requirements of the comfort power module and the non-comfort power module; S2. Evaluate the maximum charge and discharge capacity of the battery and the current recovered power through the battery status detection module; S3. Dynamically select control strategies through energy management optimization allocation module: When the battery SOC is ≥ 30% or the optimal control module fails, a rule-based power allocation strategy is adopted; When the battery SOC is less than 30%, the driving power response ω is established. 1 、Braking power responsivenessω 2 、Power responsiveness of power consumption module ω 3 To optimize the rolling horizon optimization model of variables, solve the optimal power allocation scheme; S4. Provide the driver with suggestions for adjusting the comfortable power consumption module through the human-computer interaction module, and dynamically adjust the power distribution based on the driver's feedback.

[0012] Further, the driver acceleration intention recognition includes clustering analysis of an offline data set by using a GADPC graph theory-density peak clustering algorithm, wherein the data set includes an accelerator pedal opening, an accelerator pedal opening change rate, a vehicle speed, and a corresponding acceleration; By calculating the distance between the test sample and the cluster center point, combined with the adjustment coefficient and distance correction coefficient, the real-time acceleration demand is predicted.

[0013] Further, the GADPC graph theory-density peak clustering algorithm includes the following steps: SA1. Select state quantity , control amount To serve as sample data, good data that can better describe normal urban road driving conditions are collected and screened offline as training data sets, including is the accelerator pedal opening, is the rate of change of the accelerator pedal opening, v is the vehicle speed, a is the corresponding acceleration, and the local density of the data points is calculated. and distance ; In the formula, the cutoff distance dc is the manually set neighborhood cutoff distance of data point i, and takes a value that makes the number of neighborhood points of each data point account for 1%-2% of the total number of data points on average. ;X ik , X jk is the state element of the data point, and n is the number of states; SA2. Add endpoints based on coordinates and edge lines based on cutoff distances, build a sparse graph and filter suspicious cluster center points; SA3. Determine the final cluster center point through cut point and bridge detection; SA4. The percentage of edge points and outliers should not exceed 2% as the data set screening criteria.

[0014] Furthermore, in step S1, the step of identifying the braking intensity by using the fuzzy inference rule includes: Define the membership function of the brake pedal output voltage, the output voltage change rate and the fuzzy linguistic variables of the braking intensity; Reasoning is performed according to the braking intensity recognition inference rules, and the braking intensity z is output through defuzzification using the center of gravity method.

[0015] Furthermore, the objective function of the rolling horizon optimization model is: ; where t i , The time interval is t c The discrete moments have the following constraints: ; Among them, ω 1 ,ω 2 ,ω 3 is the weight coefficient, is the driving mode's influence coefficient on driving power, is the influence coefficient of driving mode on braking power, It is the influence coefficient of driving mode on the power of vehicle power module.

[0016] Furthermore, the driving modes include: Sports mode, satisfying , so that the motor drive power demand can be met first; Comfort mode, satisfying , so that the demand for comfortable power modules can be met first; Economic model, satisfying , so that the energy recovery power demand can be met first.

[0017] Furthermore, in step S3, the rule-based power allocation strategy includes: SB1. Power distribution during braking or coasting: When the vehicle is in braking or coasting state, the motor enters the electric feedback braking state, and the following power distribution rules are adopted: , where k gen P is the current power adjustment parameter of the motor. 2_r is the power required for motor braking, P 3_r is the power required by other power-consuming modules, P S For safety demand power, , where P STS and P BS are the rated power of the steer-by-wire system and the brake-by-wire system, P safe Reserve power for battery safety; when When satisfying , , , Under the constraints, 3_r The power consumption modules in the system meet the requirements of each module in turn according to the priority of BMS, BTMS, PTMS, and DCS. BMS_min , P BTMS_min , P PTMS_min , P DCS_min Both are the minimum power required to maintain the operation of the power module; when In the case of 2 Dynamically adjust to meet constraints; SB2. Power distribution in driving state: When the vehicle is in driving state, the motor is in working state, and the power distribution rules are as follows: ; When the required power meets the constraint condition, power allocation is performed according to the required power. If it cannot be met, the following adjustments are made: when In this case, let the power allocated to the drive motor be ,in The driving power adjustment factor affected by the driving mode is set to , comfort mode is , the economic model is In satisfying , , , Where P BMS_min , P BTMS_min , P PTMS_min , P DCS_min Under the condition that both are the minimum power constraints required to maintain the operation of the power module, 3_rThe power consumption module requirements in the system are met in order of priority of BMS, BTMS, PTMS, and DCS until Re-established.

[0018] Furthermore, in step S2, the battery status detection module calculates the SOC using the Thevenin equivalent circuit model combined with the ampere-hour integration method, and uses the formula: Determine the maximum power of battery charging and discharging; Where n s Indicates the number of battery cells connected in series in the battery pack, n p Indicates the number of parallel battery cells in the battery pack, U t,max and U t,min They are the voltage thresholds of the battery during charging and discharging respectively; future Maximum allowable charge and discharge current after time , is the maximum battery charging current, is the maximum discharge current of the battery, S max is the maximum SOC value, S min is the minimum SOC value; The maximum charge and discharge current of the entire battery pack .

[0019] Furthermore, the human-computer interaction module provides suggestions based on the following conditions: When SOC>30% and there is insufficient power in the power module, it is prompted to reduce the comfortable power demand; When SOC<30% and ω 1 When <1.0, the prompt switches to economic mode; When SOC<30% and ω 3 When <1.0, it is recommended to turn off PTMS or DCS.

[0020] Furthermore, the comfort power module includes a cabin thermal management module and a low-voltage power system, and the non-comfort power module includes a battery management system and a battery thermal management system.

[0021] Beneficial effects: The present invention combines rolling time domain optimization with rule control, accurately identifies the driver's intention, designs a human-computer interaction module, and dynamically selects a control strategy based on the battery status, thereby comprehensively improving the optimality, adaptability, real-time performance, user experience, and fault tolerance of the energy management strategy of pure electric commercial vehicles.

[0022] Specifically, the advantages include: 1. Improve the optimality of energy distribution: This invention introduces the rolling horizon optimization (RHC) algorithm, comprehensively considers the impact of the driving conditions in the next 1-3 seconds on the energy flow of the vehicle, and dynamically adjusts the power distribution strategy. When the battery SOC is high, rule control is used to ensure real-time performance; when the SOC is low, it switches to optimal control to improve economy. This strategy can adjust power distribution in real time according to the predicted driving conditions to ensure the optimal energy utilization efficiency under different driving conditions.

[0023] 2. Enhance system adaptability and robustness: This invention introduces a graph-based density peak clustering algorithm (GADPC) and a fuzzy reasoning braking intention recognition algorithm to accurately identify the driver's acceleration and braking intentions. At the same time, a human-computer interaction module is designed to feed back the power adjustment suggestions of the comfort power module to the driver, forming a closed loop of "optimization suggestions-manual decision-dynamic adjustment". By accurately identifying the driver's intentions and real-time feedback adjustments, the system can adapt to complex and changing driving environments, improve the driving experience, and ensure the stability and effectiveness of the energy management strategy.

[0024] 3. Improve computational efficiency and real-time performance: The present invention reduces computational complexity by adopting a rule-based control strategy when the SOC is high, and switches to a rolling time domain optimal control strategy when the SOC is low to improve economic performance. By dynamically selecting a control strategy based on the battery status, computational efficiency and energy economy are balanced, real-time requirements are met, and stable operation of the system under different working conditions is ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a framework diagram of the vehicle energy management strategy in Example 1 of the present invention; Figure 2 This is a schematic diagram of the vehicle power system framework, energy flow and control signal transmission in Example 1 of the present invention; Figure 3 This is a flow chart of the KNN vehicle speed prediction algorithm in Example 1 of the present invention; Figure 4 This is a flow chart of the GADPC algorithm in Example 1 of the present invention; Figure 5 is the membership function image of the brake pedal output voltage in Embodiment 1 of the present invention; Figure 6 is a membership function image of the brake pedal output voltage change rate in embodiment 1 of the present invention; Figure 7 is the braking intensity membership function image in Example 1 of the present invention; Figure 8 Schematic diagram of the Thevenin battery model in Example 1 of the present invention. DETAILED DESCRIPTION

[0026] The present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0027] Example 1 A method for energy management of a pure electric commercial vehicle comprises the following steps: S1. Obtain the driver's required power information through the driving demand collection module, including: The vehicle speed in the next 1-3 seconds is predicted based on the KNN algorithm, and the GADPC graph theory-density peak clustering algorithm is used to identify the driver's acceleration intention and calculate the driving power demand; Identify the braking intensity based on fuzzy reasoning rules and calculate the braking power demand in combination with the braking force distribution strategy; Read the real-time power requirements of the comfort power module and the non-comfort power module; S2. Evaluate the maximum charge and discharge capacity of the battery and the current recovered power through the battery status detection module; S3. Dynamically select control strategies through energy management optimization allocation module: When the battery SOC is ≥ 30% or the optimal control module fails, a rule-based power allocation strategy is adopted; When the battery SOC is less than 30%, the driving power response ω is established. 1 、Braking power responsivenessω 2 、Power responsiveness of power consumption module ω 3 To optimize the rolling horizon optimization model of variables, solve the optimal power allocation scheme; S4. Provide the driver with suggestions for adjusting the comfortable power consumption module through the human-computer interaction module, and dynamically adjust the power distribution based on the driver's feedback.

[0028] Among them, Figure 1 As shown in the figure, the vehicle energy management system mainly includes four core modules: driving demand acquisition module, battery status detection module, energy management optimization allocation module and human-computer interaction module.

[0029] The driving demand acquisition module analyzes the driver's driving needs based on past and future driving conditions, the driver's operation of the acceleration and deceleration pedal control and the human-computer interaction module, mainly including the demand for vehicle driving power, braking power and comfort power module power. The vehicle driving power demand is calculated based on past driving conditions and the driver's operation of the accelerator pedal; the vehicle braking power demand is calculated based on the driver's operation of the brake pedal and the vehicle braking force distribution strategy; the comfort power module power is calculated based on the driver's operation instructions on the human-computer interaction interface. Finally, this module outputs the driving demand power information to the energy management optimization allocation module.

[0030] The battery status detection module evaluates the maximum charge and discharge capacity of the battery based on the battery model and the current battery status; at the same time, it calculates and evaluates the current recovered power of the battery based on the current vehicle driving information. Finally, this module transmits the evaluated battery information to the energy management optimization allocation module.

[0031] The energy management optimization allocation module receives the information transmitted by the driving demand acquisition module and the battery status detection module, and transmits it to the rule-based power allocation module and the optimal control-based power allocation module for processing. Then, the adjustment results of the motor and battery management system obtained by the two parts are transmitted to the motor controller and the battery management system for power control, and the adjustment suggestion information of the comfort power module is transmitted to the human-machine interaction module.

[0032] The human-computer interaction module reminds the driver and makes operational suggestions based on the adjustment results and adjustment suggestion information obtained by the energy management optimization allocation module, and further adjusts the power distribution of the cabin thermal management system and the low-voltage power system according to the decision made by the driver.

[0033] The vehicle electrical system mainly includes electric motors, vehicle control unit VCU (Vehicle Control Unit), brake system BS (Brake System), steering system STS (Steering System), battery management module BMS (Battery Management System), battery thermal management module BTMS (Battery Thermal Management System), passenger cabin thermal management module PTMS (Passenger Thermal Management Syestem) and DC / DC low-voltage power system DCS (Direct Current System) and other power modules. The motor power is controlled by the motor controller, the BS power is controlled by the electric brake control unit, the STS power is controlled by the steering system main controller, the BMS and BTMS power are controlled by the BMS, the PTMS power is controlled by the air conditioning automatic control system, and the DCS power is controlled by the DC / DC control management module. BMS, PTMS, BTMS and DCS participate in power optimization and distribution, and PTMS and DCS are comfortable power components. The accelerator pedal information and brake pedal information are collected by the A / D acquisition module, and the gearbox information is collected by the gearbox controller. Figure 2 Shown is the framework of the vehicle's electrical system and a schematic diagram of the energy flow and control signal transmission.

[0034] Preferably, the driving demand collection module is divided into three main parts: driving power demand collection, braking power demand collection and on-board electrical component demand collection.

[0035] 1. Drive power demand collection The driving power demand collection part is divided into three parts: vehicle speed prediction, acceleration demand analysis and power calculation.

[0036] Vehicle speed prediction: The KNN algorithm is used to predict the vehicle speed in the next 1-3 seconds. The vehicle speed is predicted by fusing GIS and GPS information with vehicle sensors and calculating driving conditions (such as slope, adhesion coefficient, and temperature) in the short term in the future.

[0037] Acceleration demand calculation: The GADPC graph theory-density peak clustering algorithm is used to perform offline clustering analysis on data such as the accelerator pedal opening change rate and vehicle speed to identify the driver's acceleration intention.

[0038] Power calculation: The required driving force is calculated based on the longitudinal vehicle dynamics model, and the required motor drive power is further derived.

[0039] 2. Braking power demand collection Braking intensity requirement calculation: The fuzzy control strategy is used to process the brake pedal output voltage and its rate of change to calculate the required braking intensity.

[0040] Calculation of the maximum braking torque of the motor: Calculate the maximum braking torque of the motor based on the maximum electric braking force limit of the motor and the maximum charging current limit of the battery.

[0041] Braking force distribution strategy: When the battery SOC is lower than 90%, the motor feedback braking and mechanical braking force are distributed according to the set ratio.

[0042] Motor feedback braking recovery power calculation: Calculate the motor's recovery power based on the braking intensity and motor braking torque.

[0043] 3. Demand collection of other vehicle-mounted power modules Non-comfort power module: includes the power required by the battery management system (BMS) and the battery thermal management system (BTMS).

[0044] Comfort power module: includes the power requirements of the cabin thermal management module (PTMS) and the low-voltage power system (DCS).

[0045] More preferably, it also includes the following contents: (1) Vehicle speed prediction: The vehicle speed is predicted using the KNN (K-nearest neighbor) algorithm. The specific process is as follows: Collect the state quantity driving road slope in each state offline in advance , road adhesion coefficient , Ambient temperature T , target variable vehicle speed v The historical information is used as the training data set In actual driving, the GIS and GPS information are combined with the vehicle's multi-sensor data to predict the road slope, road adhesion coefficient and ambient temperature in the short term (usually 1-3 seconds) in the future. T As a test sample, calculate its Euclidean distance with each training sample, that is, ,in is the training sample point, is the test sample point. Then the training samples are sorted in ascending order according to the distance, the nearest k training samples are selected, and the average speed of these k training samples is calculated. , the result is the predicted vehicle speed. Figure 3 This is the flow chart of the KNN vehicle speed prediction algorithm.

[0046] Driver acceleration demand calculation: Offline classification of driving status: Selecting status quantity , the control quantity is a As sample data, good data that can better describe normal urban road driving conditions are collected and screened offline as training data sets, where is the accelerator pedal opening, is the accelerator pedal opening rate, v is the vehicle speed, a is the corresponding acceleration. The GADPC graph theory-density peak clustering algorithm is used to perform cluster analysis on the data set. First, calculate each data point i With other data points j The local density , where the cutoff distance d c Manually set data points i The neighborhood cutoff distance is usually set to a value that makes the number of neighborhood points of each data point account for 1%-2% of the total number of data points. Then calculate the distance between points , where x ik 、 x jk is the state element of the data point, n is the number of states. Then add endpoints based on the coordinates and add edge lines based on the cutoff distance, and construct a sparse graph based on the endpoints and edges. Next, calculate the gamma value of each point , and displayed in descending order on the gamma graph, according to Calculate the turning angle of each point on the gamma graph And thus construct the angle map, where a, b, c are the gamma map , , The length of each side of a triangle connected by three points, selected from the angle diagram The point before the transition to 180° (with the increase of point sequence in the turning angle diagram, the turning angle will eventually increase to 180°) is used as the center point of the suspicious cluster. The final cluster center point is determined by screening based on whether the connection path between the points in the cluster to which the suspicious center point belongs includes cut points and bridges. Then loop through each point and aggregate it to nearby points with higher density and stronger connectivity. Finally, edge points are detected with low centrality and outliers are detected with low centrality and long distance values. The number of edge points and outliers does not exceed 2% of the total number as the evaluation standard for a good data set. The number of clusters is obtained. N and the corresponding cluster center value set The overall algorithm flow chart is as follows: Figure 4 .

[0047] (2) Online acceleration prediction: During real-time vehicle operation, the VCU obtains test data points by collecting the current driver's operation of the accelerator pedal and the vehicle speed. , classify the state and calculate the distance between the test sample and the center point of the cluster, and finally obtain the required acceleration under the driving state. The calculation formula is as follows:

[0048] in , , , , are the required acceleration, accelerator pedal opening, accelerator pedal opening rate, and vehicle speed of the test sample. , , , , are the acceleration at the cluster center, the accelerator pedal opening, the accelerator pedal opening rate of change, and the vehicle speed. It is the adjustment coefficient formulated according to the actual situation, and its value range is , is the distance correction factor, , is the Euclidean distance to the test data point The acceleration of the most recent training sample point.

[0049] (3) Power calculation: The power demand of the vehicle can be calculated by the longitudinal vehicle dynamics model. The driving force F t The calculation formula is as follows: ; Where F t 、F W 、F i 、F jThey are rolling resistance, air resistance, slope resistance and acceleration resistance, among which F f The calculation formula is as follows: ; Where M is the mass of the vehicle, g is the acceleration of gravity, and f is the rolling resistance coefficient. The slope of the road in front of the vehicle is measured by the fusion information of GPS, GIS and multiple sensors; F W The calculation formula is as follows: ; Where C d is the air resistance coefficient, A is the frontal area of ​​the vehicle, is the air density, v is the predicted vehicle speed obtained in the vehicle speed prediction step; F i The calculation formula is as follows: ; F j The calculation formula is as follows: ; In the formula is the rotational mass conversion factor.

[0050] The required torque can be obtained from the driving force: Where T tq is the motor torque, r is the wheel radius, i g is the gearbox ratio, i 0 is the main reduction gearbox transmission ratio, For transmission system efficiency.

[0051] In summary, the motor predicted drive demand power is: ; Where n is the motor speed.

[0052] More preferably, (1) The calculation of the braking intensity requirement includes the following steps: The vehicle controller VCU uses a fuzzy control strategy to process the input brake pedal output voltage U and brake pedal output voltage change rate dU, and outputs the required braking intensity Z. The specific method of the fuzzy control strategy is as follows: Taking the brake pressure sensor with a rated voltage of 5V as an example, the brake pedal output voltage is [0,5], the unit is volt (V), and the fuzzy input language is U{small (US), smaller (UNS), medium (UM), larger (UNB), large (UB)}. The membership function uses a triangular function, and the function graph is as follows Figure 5 As shown. The function expression is as follows: ; ; ; ; ; The brake pedal output voltage change rate dU, domain is [0,70], unit is volt per second (V / s), fuzzy input language is dU{small (DUS), smaller (DUNS), medium (DUM), larger (DUNB), large (DUB)}. The membership function is a triangular function, the function graph is as follows Figure 6 As shown. The function expression is as follows: ; ; ; ; ; Braking intensity Z, domain is [0,1], fuzzy input language is {small (ZS), smaller (ZNS), medium (ZM), larger (ZNB), large (ZB)}. Membership functions are all triangular functions, the function graph is as follows Figure 7 As shown. The function expression is as follows: ; ; ; ; ; The fuzzy reasoning rules are shown in Table 1. Table 1 Reasoning rules for braking intensity identification

[0053] After obtaining the fuzzy expression of the braking intensity, defuzzification is performed. Here, the defuzzification method adopts the centroid method, and its calculation method is as follows: Assume that the set of output variables z is , ; Where z is the output variable of the control quantity, and p is the variable output number of Z (p=1,2,…,5); is the fuzzy membership function.

[0054] (2) Maximum braking torque of the motor Maximum electric braking force of the motor F reg_max , and its calculation formula is: ; Where i g is the gearbox ratio, i 0 is the reduction ratio of the main reducer, is the transmission efficiency from the motor to the wheel, r is the wheel radius, F reg_max is the maximum motor braking torque. Limited by the motor peak torque and the maximum battery charging current, the maximum braking torque that the motor can provide is the smaller of the two, that is: ; ; ; Where n is the motor speed; n 0 The minimum motor speed for braking energy recovery determined by the motor efficiency; T m_mot is the maximum braking torque that the motor can provide when the motor speed is n; T m_max is the peak braking torque of the motor; P m_max is the peak braking power of the motor; T m_bat is the maximum braking torque of the motor determined by the maximum charging power of the battery when the motor speed is n; The maximum allowable charging power of the battery calculated by the battery status assessment module; For battery charging efficiency.

[0055] (3) Braking force distribution strategy When the battery SOC is lower than 90%, the motor feedback braking is allowed to intervene in the braking system. This pure electric commercial vehicle adopts a parallel structure of the braking energy recovery system, in which the mechanical braking force and the electric braking force exist at the same time, and they are distributed according to a certain ratio. When the driver steps on the brake pedal to a certain opening, the pedal pressure sensor converts the pedal signal into an electrical signal and inputs it into the fuzzy controller. The fuzzy controller outputs the required braking intensity and sends a command to the motor to generate the corresponding braking force. The total braking force of the front axle is calculated as: ; Where G is the vehicle weight, b is the distance between the rear axle and the vehicle center of mass, h g is the height of the center of mass and L is the wheelbase.

[0056] The total braking force of the rear axle is calculated as: ;Where a is the distance between the front axle and the vehicle's center of mass.

[0057] The calculation formula for the motor force required for the corresponding axle is: ; where k is the rate of change of total braking force with braking intensity z, z is braking intensity, F reg_max The maximum electric braking force.

[0058] The corresponding motor braking torque is: ; Where T reg is the motor braking torque.

[0059] (4) Calculation of motor feedback braking power The estimated electric feedback braking recovery power from the above three links is: ;In the formula is the power generation efficiency of the motor, n N is the rated speed of the motor.

[0060] Therefore, the motor predicted braking power requirement is P 2_r =P reg .

[0061] Further preferably, the non-comfortable power modules are the power required by BMS and BTMS, and the VCU reads the current power required by the two, which are P BMS_r With P BTMS_r .

[0062] The comfort power modules are the power required by PTMS and DCS respectively. VCU can read the driver's operation of the comfort power module from the human-machine interface, and obtain the power P required by each through the air conditioning automatic control system and DC / DC low-voltage power management module. PTMS_r With P DCS_r .

[0063] Total demand for other on-board power modules P 3_r The calculation formula is: .

[0064] Further preferably, the battery status evaluation module is divided into two parts: battery peak charge and discharge capacity and current battery recovered power.

[0065] (1) Battery peak charge and discharge capability The battery peak charge and discharge capability is divided into three parts: battery SOC state calculation, battery maximum allowable current calculation and peak discharge charging power calculation.

[0066] Battery SOC calculation The equivalent circuit models commonly used in electric vehicle simulation include the Rint model, PNGV model, Thevenin model, and second-order RC model. Considering that the Rint model is too simple, and the PNGV model and second-order RC model are too complex to be used in practical engineering applications, this method uses the Thevenin model to model and identify the parameters of the battery module of the pure electric commercial vehicle under study. The schematic diagram of the Thevenin battery model is shown in Figure 8 shown.

[0067] The mathematical model of the Thevenin battery model is: ; Where U is the output voltage, I is the load current, and U rc is the dynamic branch voltage, R 0 The above model parameters can be identified by HPPC (compound pulse power test).

[0068] Combined with the ampere-hour integration method The SOC after time is calculated: ; In the formula is the Coulomb efficiency parameter of the current, C t is the rated capacity of the battery, i k is the current of the battery cell.

[0069] Calculation of maximum allowable battery current: Future The maximum allowable charge and discharge current after the time can be expressed as: ; In the formula is the maximum battery charging current, is the maximum discharge current of the battery, S max is the maximum SOC value, S min is the minimum SOC value.

[0070] The maximum charge and discharge current of the entire battery pack is: ; Peak discharge charging power calculation: Through the above calculation, we can get the future After a certain time, the maximum charge and discharge power of the battery pack is: ; Where n s Indicates the number of battery cells connected in series in the battery pack, n p Indicates the number of parallel battery cells in the battery pack, U t,max and U t,min They are the voltage thresholds of the battery during charging and discharging respectively.

[0071] (2) The current battery has recovered power Vehicle controller recording time The vehicle speed before and collect the current vehicle speed through sensors , and then calculate the vehicle acceleration information for each period of time: ; When the measured When it is less than zero, the slope of the road where the passing vehicle is traveling is ignored, and the braking force F t The calculation formula is as follows: ; Where F f 、F w 、F i 、F j They are rolling resistance, air resistance and acceleration resistance, among which F f The calculation formula is as follows: F f =Mgf; Where M is the mass of the vehicle, g is the acceleration of gravity, and f is the rolling resistance coefficient; F w The calculation formula is as follows: ; Where C d is the air resistance coefficient, A is the frontal area of ​​the vehicle, is the air density, v is the vehicle speed at time t; F w The calculation formula is as follows: ;In the formula is the rotational mass conversion factor.

[0072] The motor torque required for braking can be obtained from the driving force: ; Where T is the motor torque required for braking, r is the wheel radius, i g is the gearbox ratio, i 0 is the main reduction gearbox transmission ratio, is the transmission system efficiency, and k is the ratio of the motor braking force to the total braking force set by the vehicle controller.

[0073] From the motor model, we can know that the torque of motor regenerative braking is: ; Where T d is the actual torque of the motor, n N is the rated speed of the motor, n is the motor speed, and the calculation formula is: , where R is the wheel radius.

[0074] In summary, according to the models of permanent magnet brushless DC motors and permanent magnet synchronous motors commonly used in electric vehicles on the market, it can be seen that the motor drive and braking characteristics have a certain consistency. The difference lies in the difference between power generation efficiency and electric output efficiency. Therefore, the current recovered power of the motor is: ;In the formula The power generation efficiency of the motor.

[0075] Further optimized, the energy optimization and regulation module is divided into two parts: a rule-based power allocation module and an optimal control-based power allocation module. When the battery SOC state is higher than 30% or the optimal control module function fails for some reason, the rule-based power allocation module will perform power allocation, and when the battery SOC state is lower than 30%, the optimal control-based power allocation module will perform power allocation.

[0076] 1. Power allocation rules: The power distribution rules are divided into two types: braking and driving.

[0077] (1) Braking or coasting When the vehicle is in braking or coasting state, the motor is in electric feedback braking state, P 1 =0, so the following strategy is adopted: ; Where k gen P is the current power adjustment parameter of the motor. 2_r is the power required for motor braking as mentioned above, P3_r is the power required by other power-consuming modules as mentioned above, P s The safety power requirement is calculated as follows: .

[0078] Where P STS and P BS are the calibrated rated powers of the wire-controlled steering system and the wire-controlled brake system, respectively. Since the power consumed by both accounts for a relatively small proportion of the vehicle's total power and both have a significant impact on vehicle driving safety, their power changes and distribution are not considered; P safe Power is reserved for battery safety and safety rules are formulated for the battery based on the material properties, packaging and manufacturing process, and usage properties including temperature, current, voltage, external pressure, service life, and health status of the battery used in this vehicle.

[0079] When the required power meets the above constraints, power allocation is performed according to the required power; if the constraints cannot be met, two cases are discussed.

[0080] against In the case of satisfying the following constraints, 3_r The power consumption modules in the system meet the requirements of each module in turn according to the priority of BMS, BTMS, PTMS and DCS: ; ; ; ; Where P BMS_min , P BTMS_min , P PTMS_min , P DCS_min Both are the minimum power required to maintain the operation of the power-consuming modules.

[0081] against In the case of 2 Dynamically adjust to meet constraints.

[0082] (2) Drive status When the vehicle is in driving state, the motor is in working state, P 2 =0, so the following strategy is adopted: ; When the required power meets the constraint condition, power allocation is performed according to the required power; if the constraint condition cannot be met, it is divided into two cases for discussion.

[0083] against In this case, let the power allocated to the drive motor be ,in is the driving power adjustment factor affected by the driving mode. , comfort mode is , the economic model is Under the following constraints, 3_r The power consumption module requirements in the system are met in order of priority of BMS, BTMS, PTMS, and DCS until Re-established.

[0084] ; ; ; ; Where P BMS_min , P BTMS_min , P PTMS_min , P DCS_min Both are the minimum power required to maintain the operation of the power-consuming modules.

[0085] against Considering that the power consumed by the drive motor and the power module should be much greater than the current recovered power of the motor after the coefficient is adjusted, this situation is not considered.

[0086] 2. Power distribution optimization control part: The drive power response , Braking power responsiveness 、Power responsiveness of other on-board power modules As the decision variable to be optimized, the following rolling horizon optimization problem is established:

[0087] Where t i , The time interval is t c The following constraints apply: ; In the formula is the driving mode's influence coefficient on driving power, is the influence coefficient of driving mode on braking power, The driving mode affects the power of the vehicle power module. Solve for the current time t 1 The optimal power response control sequence in the prediction time domain is obtained by solving the rolling time domain optimization problem. 1 Responsiveness at all times , , As the power responsiveness control strategy at the current moment, the control amount at other moments is discarded.

[0088] After t c After time, VCU solves t 2 Repeat the above operation for the rolling time domain optimization problem at each moment.

[0089] When 1 <1, send a command to the motor controller to allocate the motor drive power ; When 2 <1, the vehicle controller adjusts the electric braking torque proportional coefficient k and sends a command to the motor controller to adjust the electric braking torque required by the braking system so that the allocated motor braking power ; When 3 <1, give the following allocation scheme information to the human-computer interaction module so that the power of other on-board power modules can be allocated , the power requirements of each module are met in turn according to the priority of BMS, BTMS, PTMS, and DCS under the following constraints: ; ; ; ; ; The impact of different driving modes on parameters is as follows: When in sports mode, , so that the motor drive power demand is met first, please refer to the setting , , ; When in comfort mode, , so that the demand for comfortable power modules is met first, which can be set as , , ; When in economy mode, , so that the energy recovery power demand is met first, which can be set as , , .

[0090] The parameter settings of the above three modes should all meet the following parameter conditions: ; Preferably, a human-machine interface (HMI) receives the power adjustment information transmitted by the energy optimization and regulation module through a vehicle controller and feeds it back to the driver for decision making.

[0091] When the battery SOC is greater than 30%, if the power distribution rules show that the power consumption components cannot meet the power demand, the HMI performs the following operations: when in braking or coasting state, the driver is reminded to reduce the power demand for the comfort power module according to the comfort power module adjustment strategy provided by the energy optimization regulation module; when in acceleration state, the driver is informed that the motor drive power has been limited and is reminded to reduce the power demand.

[0092] When the battery SOC is less than 30%, if ω 1 <1, the HMI prompts the driver that the motor drive power has been limited, and recommends the driver to reduce the power demand and select the economic mode; if 2 <1, HMI recommends the driver to take appropriate gliding or slow braking operations to reduce the intensity of regenerated energy; if ω 3 <1, the HMI recommends that the driver reduce the demand for the comfort power module or turn off the comfort power module according to the adjustment plan of the energy optimization regulation module.

[0093] Example 2: A pure electric commercial vehicle energy management system A pure electric commercial vehicle energy management system, comprising: Driving demand collection module, used to obtain driving power demand, braking power demand and comfort power module power demand; Battery status detection module, used to evaluate the maximum charge and discharge capacity of the battery and the current recovered power; Energy management optimization allocation module, configured to dynamically switch rule control and rolling time domain optimization control according to battery SOC; A human-machine interaction module, which is used to provide power adjustment suggestions to the driver and receive feedback instructions; The system realizes data interaction and control signal transmission between modules through the vehicle controller (VCU).

[0094] Preferably, the driving demand collection module includes: The vehicle speed prediction unit uses the KNN algorithm to predict the vehicle speed in the next 1-3 seconds; The acceleration intention recognition unit uses the GADPC graph theory-density peak clustering algorithm to classify driver needs; The braking intention recognition unit uses a fuzzy controller to output the braking intensity z.

[0095] Preferably, the energy management optimization allocation module includes: A rule control unit predefines the power allocation priority in braking / driving states; Rolling time domain optimization unit, with power responsiveness , , are decision variables, and solve the multi-objective optimization problem.

[0096] Preferably, the human-computer interaction module includes: A display unit for displaying power allocation suggestions and SOC status; The input unit is used to receive the driver's adjustment instructions for the comfort power module.

[0097] Preferably, the battery status detection module includes: Thevenin equivalent circuit model unit, used to calculate battery terminal voltage and SOC in real time; Peak power calculation unit, based on the battery maximum charge and discharge current and voltage threshold output , where n s Indicates the number of battery cells connected in series in the battery pack, n p Indicates the number of parallel battery cells in the battery pack, U t,max and U t,min They are the voltage thresholds of the battery during charging and discharging respectively.

[0098] In view of the deficiencies in the prior art, the present invention proposes a new energy management strategy that combines the advantages of rule-based energy management methods and optimal control-based strategies. By introducing a hybrid strategy that combines rolling time domain optimization with rule control, the present invention can improve the optimality of energy distribution and the adaptability of the system while ensuring real-time performance. In particular, when the battery SOC is low, the system economy can be improved through the optimal control strategy, avoiding the deficiencies of existing rule-based strategies under complex working conditions. In addition, the present invention also realizes the dynamic coordination of vehicle energy management and driver needs by introducing a human-computer interaction module, further improving the driving experience and energy efficiency management effects.

[0099] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A vehicle energy management method for a pure electric commercial vehicle, characterized in that: The following steps are involved: S1. Obtain the driver's required power information through the driving demand acquisition module, including The vehicle speed in the next 1-3 seconds is predicted based on the KNN algorithm, and the GADPC graph theory-density peak clustering algorithm is used to identify the driver's acceleration intention and calculate the driving power demand; Identify the braking intensity based on fuzzy reasoning rules and calculate the braking power demand in combination with the braking force distribution strategy; Read the real-time power requirements of the comfort power module and the non-comfort power module; S2. Evaluate the maximum charge and discharge capacity of the battery and the current recovered power through the battery status detection module; S3. Dynamically select control strategies through energy management optimization allocation module: When the battery SOC is ≥ 30% or the optimal control module fails, a rule-based power allocation strategy is adopted; When the battery SOC is less than 30%, a rolling time domain optimization model is established with the driving power response ω1, the braking power response ω2, and the power response ω3 of the power module as optimization variables to solve the optimal power allocation solution; S4. Provide the driver with suggestions for adjusting the comfortable power consumption module through the human-computer interaction module, and dynamically adjust the power distribution based on the driver's feedback.

2. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: The driver acceleration intention recognition includes clustering analysis of an offline data set using a GADPC graph theory-density peak clustering algorithm, wherein the data set includes an accelerator pedal opening, an accelerator pedal opening change rate, a vehicle speed, and a corresponding acceleration; By calculating the distance between the test sample and the cluster center point, combined with the adjustment coefficient and distance correction coefficient, the real-time acceleration demand is predicted.

3. The method for energy management of a pure electric commercial vehicle according to claim 2, characterized in that: The GADPC graph theory-density peak clustering algorithm comprises the following steps: SA1. Select state quantity , control amount To serve as sample data, good data that can better describe normal urban road driving conditions are collected and screened offline as training data sets, including is the accelerator pedal opening, is the rate of change of the accelerator pedal opening, v is the vehicle speed, a is the corresponding acceleration, and the local density of the data points is calculated. and distance ; In the formula, the cutoff distance d c is the manually set neighborhood cutoff distance of data point i, and takes a value that makes the number of neighborhood points of each data point account for 1%-2% of the total number of data points on average. ;X ik , X jk is the state element of the data point, and n is the number of states; SA2. Add endpoints based on coordinates and edge lines based on cutoff distances, build a sparse graph and filter suspicious cluster center points; SA3. Determine the final cluster center point through cut point and bridge detection; SA4. The percentage of edge points and outliers should not exceed 2% as the data set screening criteria.

4. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: In step S1, the step of identifying the braking intensity by using fuzzy inference rules includes: Define the membership function of the brake pedal output voltage, the output voltage change rate and the fuzzy linguistic variables of the braking intensity; Reasoning is performed according to the braking intensity recognition inference rules, and the braking intensity z is output through defuzzification using the center of gravity method.

5. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: The objective function of the rolling horizon optimization model is: ; where t i , The time interval is t c The discrete moments have the following constraints: ; Among them, ω1, ω2, ω3 are weight coefficients, is the driving mode's influence coefficient on driving power, is the influence coefficient of driving mode on braking power, It is the influence coefficient of driving mode on the power of vehicle power module.

6. The method for energy management of a pure electric commercial vehicle according to claim 5, characterized in that: The driving modes include: Sports mode, satisfying , so that the motor drive power demand can be met first; Comfort mode, satisfying , so that the demand for comfortable power modules can be met first; Economic model, satisfying , so that the energy recovery power demand can be met first.

7. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: In step S3, the rule-based power allocation strategy includes: SB1. Power distribution during braking or coasting: When the vehicle is in braking or coasting state, the motor enters the electric feedback braking state, and the following power distribution rules are adopted: , where k gen P is the current power adjustment parameter of the motor. 2_r is the power required for motor braking, P 3_r is the power required by other power-consuming modules, P S For safety demand power, , where P STS and P BS are the rated power of the steer-by-wire system and the brake-by-wire system, P safe Reserve power for battery safety; when When satisfying , , , Under the constraints, 3_r The power consumption modules in the system meet the requirements of each module in turn according to the priority of BMS, BTMS, PTMS, and DCS. BMS_min , P BTMS_min , P PTMS_min , P DCS_min Both are the minimum power required to maintain the operation of the power module; when In the case of , the vehicle controller adjusts the electric braking torque proportional coefficient k to dynamically adjust the braking power P2 to meet the constraint conditions; SB2. Power distribution in driving state: When the vehicle is in driving state, the motor is in working state, and the power distribution rules are as follows: ; When the required power meets the constraint condition, power allocation is performed according to the required power. If it cannot be met, the following adjustments are made: when In this case, let the power allocated to the drive motor be ,in The driving power adjustment factor affected by the driving mode is set to , comfort mode is , the economic model is In satisfying , , , Where P BMS_min , P BTMS_min , P PTMS_min , P DCS_min Under the condition that both are the minimum power constraints required to maintain the operation of the power module, 3_r The power consumption module requirements in the system are met in order of priority of BMS, BTMS, PTMS, and DCS until Re-established.

8. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: In step S2, the battery status detection module calculates the SOC using the Thevenin equivalent circuit model combined with the ampere-hour integration method, and uses the formula: Determine the maximum power of battery charging and discharging; Where n s Indicates the number of battery cells connected in series in the battery pack, n p Indicates the number of parallel battery cells in the battery pack, U t,max and U t,min They are the voltage thresholds of the battery during charging and discharging respectively; future Maximum allowable charge and discharge current after time , is the maximum battery charging current, is the maximum discharge current of the battery, S max is the maximum SOC value, S min is the minimum SOC value; The maximum charge and discharge current of the entire battery pack .

9. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: The human-computer interaction module provides suggestions based on the following conditions: When SOC>30% and there is insufficient power in the power module, it is prompted to reduce the comfortable power demand; When SOC<30% and ω1<1.0, it prompts to switch to economic mode; When SOC<30% and ω3<1.0, it is recommended to shut down PTMS or DCS.

10. The method for energy management of a pure electric commercial vehicle according to claim 1, characterized in that: The comfort power module includes a cabin thermal management module and a low-voltage power system, and the non-comfort power module includes a battery management system and a battery thermal management system.

Images