Battery charging and discharging strategy optimization method for new energy commercial vehicle

Abstract

The invention discloses a battery charging and discharging strategy optimization method for a new energy commercial vehicle. The method comprises the following steps: obtaining charging and discharging habit data of the new energy commercial vehicle in each region; evaluating the health level of the batteries after the new energy commercial vehicle is operated for one period, and screening the batteries with the optimal health level and the corresponding charging and discharging habits of the optimal level; searching regions with geographic features similar to those of the region where the target new energy commercial vehicle is located, and extracting optimal-grade charging and discharging habits of the regions; setting a threshold range conforming to scoring qualification by using the habits, calibrating a weight coefficient of a first charge-discharge score according to the overlapping degree, and performing weighted summation to obtain the first charge-discharge score; taking the charging and discharging habit of the target new energy commercial vehicle as a scoring standard to obtain a second charging and discharging score; and comparing the two scores, selecting a scoring standard corresponding to the lower score as an optimization basis, and generating a battery charging and discharging optimization strategy of the target new energy commercial vehicle, thereby improving the performance of the battery and prolonging the service life of the battery.

Description

Technical Field

[0001] The present invention belongs to the technical field of new energy vehicles, and in particular relates to a method for optimizing battery charging and discharging strategies of new energy commercial vehicles. Background Art

[0002] With the increasing popularity of new energy vehicles, new energy commercial vehicles have occupied an important position in the field of transportation. They not only bring convenience to passengers, but also help protect the environment, which makes new energy commercial vehicles have great development potential. New energy commercial vehicles are vehicles that use electricity as a power source. They mainly rely on batteries to store energy. Charging and discharging behavior habits have an important impact on battery life and performance. Because the bad charging and discharging habits of new energy commercial vehicles damage the battery, reduce the battery life, reduce the battery capacity and performance, which is not conducive to ensuring the sustainable performance and efficiency of new energy commercial vehicles. It is precisely because of this adverse effect that a battery charging and discharging strategy optimization method is needed, which can not only learn the good charging and discharging habits of new energy commercial vehicles in similar areas, but also evaluate its own charging and discharging behavior habits through scoring, recognize its own bad charging and discharging behavior habits, adjust the charging and discharging strategy, and avoid damage to the battery during operation and maintenance.

[0003] Application number "202011553224.7" discloses a battery charging and discharging control method, device, server and storage medium to dynamically manage battery charging and discharging to improve battery charging and discharging safety protection and increase battery life and usage cycle.

[0004] Application number "202211240659.5" discloses a new energy vehicle lithium battery charging and discharging management system, which detects the discharge process of the new energy vehicle lithium battery and issues an alarm when an abnormality is detected, reminding the driver that it needs to be inspected and maintained.

[0005] Application number "202310608962.4" discloses a charging reminder method, device and equipment, which obtains the historical charging data of the target vehicle and reminds the target user to charge the target vehicle when at least one charging reminder condition is met.

[0006] Judging from the technologies disclosed in existing patent documents, most of the data on charging and discharging behavior of new energy vehicles are aimed at private cars, without considering that commercial vehicles such as electric buses have longer daily mileage and complex road conditions, and their charging and discharging behavior habits have a more important impact on the battery. The increasingly serious problem is how to avoid charging and discharging behavior habits that are not conducive to the battery.

[0007] Although the application number "202110347085.0" can avoid some risks by obtaining the charging and discharging behavior habits of new energy vehicles, it refers to its own charging and discharging habits for reference and learning, rather than the charging and discharging behavior habits of other new energy vehicles. Can a strategy be formulated to obtain good charging and discharging habits that have a good impact on the battery from the good charging and discharging behavior habits of other new energy operating vehicles in similar regions, so as to optimize the charging and discharging strategy of new energy operating vehicles themselves, improve battery performance, and extend battery life? Summary of the Invention

[0008] The object of the present invention is to provide a method for optimizing the battery charging and discharging strategy of new energy operating vehicles. This method selects excellent charging and discharging behavior habits as the scoring criteria from other regions with geographical characteristics similar to those of the target new energy operating vehicle, obtains the first score of the charging and discharging habits, takes the charging and discharging behavior habits of the target new energy operating vehicle itself as the scoring criteria, obtains the second score of the charging and discharging habits, and then optimizes the battery charging and discharging strategy according to the scoring.

[0009] To solve the above technical problems, the present invention adopts the following technical solutions: A method for optimizing the battery charging and discharging strategy of new energy operating vehicles, including the following steps:

[0010] S1. Monitor charging and discharging habits: Use a driving recorder, a battery management system BMS, and a variety of sensors to fully and comprehensively monitor the charging and discharging behavior habits of new energy operating vehicles, obtain charging and discharging behavior habit data information, and transmit it to the cloud server;

[0011] S2. Evaluate the battery health level after one operating cycle: By monitoring the charging and discharging behavior habits of a new energy operating vehicle in a certain area and recording it with one year of operation as a cycle, establish a K-means evaluation model according to the battery parameter indicators to evaluate the battery health level of new energy operating vehicles, which is divided into three grades: m1, m2, and m3. Among them, m3 is the optimal grade, m2 is the medium grade, and m1 is the worst grade. Since the level of battery health reflects the quality of charging and discharging behavior habits, only the optimal grade M3 charging and discharging behavior habit data recorded by the batteries with the optimal grade m3 of health level is extracted;

[0012] S3. Screen the optimal grade charging and discharging habits of regions similar to the target region: Determine the region where the target new energy operating vehicle whose charging and discharging habits need to be optimized is located, and use it as the target region. Use the optimal grade M3 charging and discharging behavior habit data as the database, preset the threshold ranges of temperature, latitude, longitude, and altitude of the target region, and use the SVM model algorithm for classification. Through this classification process, screen out the regions that meet the conditions, and extract the optimal grade M3 charging and discharging behavior habit data of these regions from the database;

[0013] S4. Scoring of target charging and discharging behavior habits: According to the optimal level M3 charging and discharging behavior habits of the similar regions to the region where the target new energy operation vehicle is located finally selected, extract multiple groups of discharge habit characteristic sample data and multiple groups of charging habit characteristic sample data; preset the threshold range for each charging and discharging behavior habit characteristic that meets the scoring qualification; and calibrate the weight coefficient of the first score of each charging and discharging behavior habit based on the high or low overlap degree of each charging and discharging behavior habit finally selected; evaluate the first score of the discharge habit and the first score of the charging habit of the target new energy operation vehicle based on the preset threshold range and the weight coefficient of the first score of the charging and discharging behavior habit; at the same time, take the target new energy operation vehicle's own charging and discharging habits as the scoring standard to obtain the second score of the discharge habit and the second score of the charging habit as a supplementary comparison to the first score. This method more comprehensively evaluates the performance of the target new energy operation vehicle's charging and discharging behavior habits in the overall scoring and optimizes it more precisely;

[0014] S5. Optimizing the charging and discharging strategy based on the scores: By comparing the first scores of the charging and discharging behavior habits of the new energy operation vehicle with the second scores of the charging and discharging habits, select the scoring standard corresponding to the lower score as the optimization basis to generate the battery charging and discharging optimization strategy for the new energy operation vehicle. The driver adjusts the battery charging and discharging strategy by taking measures according to this strategy to achieve the purpose of extending the battery life and improving the sustainability and economy of the new energy operation vehicle.

[0015] Further, in step S1, a driving recorder and multiple sensors are used to monitor the discharging behavior habits of the new energy operation vehicle, and collect the following parameter real-time data: the number of sharp decelerations ρ and the number of sharp accelerations during the discharging process The situation of the driving speed level The driving road condition L, the outside temperature The affiliated range, forming the discharging behavior habits, and transmitting the collected data information to the cloud server; the specific content is as follows:

[0016] (A) The driving recorder detects the sharp deceleration and sharp acceleration conditions of the vehicle through the built-in acceleration sensor (G sensor); the acceleration sensor continuously collects the acceleration data of the vehicle, sets a specific acceleration threshold of 3m / s 2 , and a duration threshold of 2s. When the positive direction acceleration threshold and the time threshold are exceeded, trigger the event record of sharp acceleration. When the negative direction acceleration is less than the set value and the time threshold is exceeded, trigger the event record of sharp deceleration, determine the number of sharp braking times and sharp acceleration times of this new energy operation vehicle in a section of the journey, and record the sharp acceleration habit and sharp deceleration habit in the discharging behavior of the new energy operation vehicle;

[0017] (B) The driving recorder monitors the current position of the new energy commercial vehicle represented in longitude and latitude through the GPS module, and calculates the driving speed of the new energy commercial vehicle. The present invention sets two key parameters: speed threshold and duration threshold. If the monitored driving information is greater than the speed threshold of 40 km / h and the duration threshold of 1 min, it is determined to be driving at high speed; meanwhile, for speed ranges in (40 km / h, 50 km / h], (50 km / h, 60 km / h], (60 km / h, 70 km / h], etc., and the duration exceeding the duration threshold, the high-speed driving levels are classified as Ⅰ, Ⅱ, Ⅲ... in sequence. Within the range not exceeding the threshold, the normal driving levels are classified according to the speed and duration thresholds; for speed ranges in (30 km / h, 40 km / h], (20 km / h, 30 km / h], (10 km / h, 20 km / h], (0 km / h, 10 km / h], and when meeting or exceeding the duration threshold, the normal driving levels are classified as 4, 3, 2, 1 in sequence, so as to judge the level of the current driving speed and record the driving speed habits in the discharge behavior of the new energy commercial vehicle;

[0018] (C) Use the vehicle dynamics sensor and the road surface vibration sensor installed at the bottom of the vehicle to monitor the driving road conditions of the vehicle. The driving road conditions L consist of two aspects: road surface unevenness L1 and road surface slope L2; the present invention uses the International Roughness Index (IRI) as the evaluation index for the road surface unevenness level, establishes a calculation mathematical model for the road surface unevenness level by using the definition of IRI, and verifies its correctness through multiple sets of data; the vibration sensor can infer the load spectrum of the road section based on the vibration spectrum and amplitude generated when the vehicle passes, and find the corresponding road surface unevenness coefficient G after evaluating the road surface level through the road section load spectrum d (n0), when the new energy commercial vehicle drives at different vehicle speeds on roads of different grades, the road surface will generate time-domain excitation signals for the vehicle, and these signals change over time, reflecting the dynamic response characteristics of the vehicle to the road surface excitation;

[0019] The time-domain excitation signals corresponding to various road surface grades and various vehicle speeds are the road surface excitation time-frequency power spectral density G d (n), which can be expressed by Equation (1):

[0020]

[0021] v b is the vehicle speed driving on this road surface, Ω is the time frequency for measuring the vehicle driving on this road surface, and n0 is the reference spatial frequency, usually 0.1 m -1 ;

[0022] Calculate the current road surface unevenness IRI level value, and the formula is:

[0023]

[0024] Among them, q is the correction coefficient, which is usually 0.78 verified by multiple groups of real vehicle data;

[0025] The road surface unevenness level L1 is divided as shown in Table 1 below. Different IRIs correspond to different road surface unevenness levels

[0026] Table 2 Corresponding Table of Road Surface Unevenness Levels

[0027]

[0028] The estimated driving road surface slope L2 of the present invention is established based on the least - squares estimation model of longitudinal dynamics. When a new - energy commercial vehicle is driving on a slope, state variables at different times, such as speed, acceleration, steering angle, etc., and the vehicle position information provided by GPS are obtained by vehicle dynamics sensors. The dynamic system estimators of the system parameter w at the (i - 1) - th moment and the i - th moment are respectively The weights between the slope prediction value and the actual value at the (i - 1) - th moment and the i - th moment are P(i - 1) and P(i) respectively. The Kalman gain used at the i - th moment is K(i), and the system output at the i - th moment is F if (i), and the observable data vector at the i - th moment is μ(i). Then the road surface slope model based on the least - squares estimation of longitudinal kinematics is established as:

[0029]

[0030] Among them, the initial value P(0)=Z×E is determined. Z is a sufficiently large positive real number, and E usually represents the prediction error covariance matrix, that is, the uncertainty of slope estimation; τ Z is the forgetting factor of the least - squares estimation model of road surface slope based on longitudinal dynamics. The selection range of the forgetting factor is (0, 1]. Considering the variation characteristics of the road surface slope, the influence of old observation data on the new reference estimation is small. Therefore, the forgetting factor τ Z =0.9 is selected, and the value of is calculated, then the dynamic system estimator of the system parameter w at the i - th moment is obtained;

[0031] From w = sinα+(f0 + f1v p ), the road surface slope can be deduced as

[0032]

[0033] Among them, v p is the driving speed of the commercial vehicle at the i - th moment, and f0 and f1 are the constant term and the first - order coefficient of the speed fitting of the rolling resistance coefficient;

[0034] The present invention classifies the road surface slope L2 into five grades. Grade 1 has a slope of 0%-1%, Grade 2 has a slope of 1%-5%, Grade 3 has a slope of 5%-10%, Grade 4 has a slope of 10%-15%, and Grade 5 has a slope of more than 15%.

[0035] Respectively evaluate and record the grades of the road surface unevenness L1 and the road surface slope L2 in the driving road conditions information in the current discharge behavior habits of new energy commercial vehicles;

[0036] (D) The driving recorder records the external temperature during the discharge process of new energy commercial vehicles. The present invention sets every 5°C as a threshold range for the classification of the external temperature range. The temperature range of 0°C and above is divided into i intervals, which are respectively marked as r1, r2... r , and the temperature range below 0°C is also divided into i intervals, which are respectively marked as -r1, -r2... -r i , and classify the subordinate range of the current external temperature during the driving of new energy commercial vehicles into the corresponding intervals for recording; i , Based on monitoring the above-mentioned number of emergency brakes ρ, the number of rapid accelerations

[0037] The driving speed grade situation , the grade situations of the road surface unevenness L1 and the road surface slope L2 in the driving road condition L, and the external temperature The subordinate range situation, the discharge behavior habits of new energy commercial vehicles are obtained;

[0038] Taking an example recorded in Table 2, the habits of a certain new energy commercial vehicle in a certain discharge behavior in Area A are monitored;(E)

[0039] Table 2 Monitoring Table of the Discharge Behavior Habits of a Certain New Energy Commercial Vehicle in Area A

[0040]

[0041] This example monitoring method is also applicable to monitoring the discharge behavior habits of all new energy commercial vehicles in this area and other areas.

[0042] Furthermore, the use of the battery management system BMS to monitor the charging behavior habits of new energy commercial vehicles in step S1 includes: by monitoring the voltage, current, temperature, and charging time parameters during charging, estimating the battery SOC state, charging method QS, and charging temperature The charging duration These four charging habit indicators, and transmit the charging habit data information into the cloud server; the specific content is as follows:

[0043] (a) The BMS monitors the voltage, current, and temperature parameters of the battery during charging, and uses an improved Kalman filter algorithm to estimate the state of charge (SOC) of the battery. Compared with the traditional Kalman filter algorithm, this improved Kalman filter algorithm can improve the calculation accuracy of the SOC of the battery;

[0044] Using the monitored battery output voltage U C , the battery working current I, the target voltage U m of the battery, and the battery working duration T h , calculate I×T h to obtain the maximum available capacity C max of the battery, and then further calculate the value of SOC. The calculation formula for the state of charge SOC of the charging battery is:

[0045]

[0046] where, in the formula, SOC(t1) represents the real-time SOC value, SOC(t0) represents the initial battery charge, C max represents the maximum available capacity, η represents the battery Coulomb efficiency, and I(t) represents the function of current and time;

[0047] Divide the battery SOC state [0%, 100%] into 10 equal SOC threshold ranges. [0%, 10%] is the first range, [10%, 20%] is the second range, and so on to [90%, 100%] as the tenth range. Estimate in which range the SOC state of the new energy operating vehicle is when starting to charge, and record the corresponding data information;

[0048] (b) The BMS monitors the real-time voltage and real-time current data during the charging process, and calculates the average power P avg during charging. The formula is:

[0049]

[0050] where, represents the charging duration;

[0051] Set specific power thresholds. The power threshold for fast charging is 150 KW, and the power threshold for slow charging is 50 KW; when the new energy operating vehicle is charging, classify the charging method QS according to the calculated average power P avg ; if the average power is greater than the set fast charging power threshold, this charging is classified as fast charging; if the average power is less than the slow charging power threshold, this charging is classified as slow charging, and record the charging data information of this time; if the average power is within the range of [50 KW, 150 KW], this charging is classified as normal charging, abbreviated as general charging;

[0052] (c) Monitor the external temperature during charging through the temperature sensor in the BMS. For the division of the charging temperature range in this invention, set every 5°C as a threshold range, divide the temperature range above zero into i intervals, and mark them as g1, g2... g i , also divide the temperature range below zero into i intervals, and mark them as -g1, -g2... -g i , and classify the current charging temperature range of the new energy operation vehicle into the corresponding interval for recording;

[0053] (d) Record the start time of charging through the BMS, and monitor the change of current to determine the end time of charging, and then calculate the charging duration Set every 0.5h of charging duration as a threshold range, divide the charging duration into i intervals, and mark them as Q1, Q2... Q i , and classify the current charging duration range of the new energy operation vehicle into the corresponding interval for recording;

[0054] Based on monitoring the SOC state, charging method QS, temperature during charging range situation, charging duration range situation, obtain the charging habits of new energy operation vehicles;

[0055] (e) Taking an example recorded in Table 3, monitor the habits of a certain new energy operation vehicle in a certain charging behavior in Area A

[0056] Table 3 Monitoring Table of the Charging Behavior Habits of a Certain New Energy Operation Vehicle in Area A

[0057]

[0058] This example monitoring method is also applicable to monitoring the charging behavior habits of all new energy operation vehicle drivers in this area and other areas.

[0059] Furthermore, in step S2, establish a K-means evaluation model to evaluate the battery health level of new energy operation vehicles after one operation cycle. The implementation principle process is as follows:

[0060] (1) The current battery SOH, cycle life ε, capacity decay Q C , internal resistance R v , open circuit voltage U K , charging efficiency C R , discharge efficiency F RThese seven parameters are input into the K-means algorithm for clustering analysis, divided into K equally spaced clusters, so that each cluster has an initial center point;

[0061] (2) The input data is assigned to each cluster until all the input data is assigned to the new clusters, and the total number of samples S in the input dataset is assigned to K clusters. This is achieved by the K-means continuously adjusting the centroid positions of the clusters to minimize the sum of the distances from the points to the centroids, that is, minimizing the squared error J. For the minimized squared error J of the cluster partition γ obtained by clustering;

[0062]

[0063] where γ = {γ1, γ2,..., γ k} represents the clusters of the battery health level of K types, is the current battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charge efficiency C R , discharge efficiency F R of the feature matrix, represents the centroid of the i-th cluster;

[0064] (3) Randomly select K cluster centroids of the battery health level feature samples, measure the distances between the samples and each centroid, and assign each sample to the nearest cluster centroid. Iterate this n times. In each iteration process, update the centroids of each cluster using the mean. Repeat the above steps for the K cluster centroids until the cluster centroids are stable or the function converges;

[0065] Based on the above steps, the battery health assessment model can be divided into 3 battery health level grades: m1, m2, m3. Among them, m3 is the optimal grade, m2 is the medium grade, and m1 is the worst grade; thus, the battery health level of a certain new energy operating vehicle in a certain area after one operation cycle is evaluated. Since the high or low battery health level reflects the quality of the charge and discharge behavior habits, the battery health level of the optimal grade m3 corresponds to the optimal grade M3 of the charge and discharge behavior habits, and so on for M1 and M2. The present invention only extracts the data of the optimal grade M3 of the charge and discharge behavior habits recorded by the batteries whose battery health levels are evaluated as the optimal grade m3, and the others rated as M1 and M2 grades will not be used as references.

[0066] Furthermore, in step S3, in order to optimize the charge and discharge behavior habits of a certain new energy operating vehicle in area A, area A is determined as the target area, and the SVM model is used to screen the areas with geographical features similar to area A, and select the optimal grade M3 of the charge and discharge habits from them;

[0067] During the data screening process based on the SVM model algorithm, temperature, latitude, longitude, and altitude geographical features are used as key parameters for training the model; the SVM model classifies data with similar attributes based on these features; the steps for screening data using the SVM model are as follows:

[0068] Step 1) Establish a database divided by geographical location, which contains temperature, latitude, longitude, and altitude feature signals of each region, extract data with these features from it, and collect sample data with temperature, latitude, longitude, and altitude information of each region; each sample should include these feature values and a target label indicating the category to which the sample belongs;

[0069] Step 2) Preprocess the data, including handling missing values, normalizing features, etc. The goal of this process is to ensure that the dataset can be smoothly used for training the model; subsequently, the dataset is split into a training set and a test set;

[0070] Step 3) Use the support vector machine algorithm to train the model. During the training process, the algorithm will learn how to map the input features to the corresponding target labels; the support vector machine model performs type recognition on the samples, and it presets a kernel function θ; use the regression model samples to construct the cost function Its expression is

[0071]

[0072] Apply the constraint condition:

[0073]

[0074] δ i ≥0, i = 1, 2, 3, 4…

[0075] Among them, ω z Represents the classification interface vector; Represents the transpose of the classification interface vector matrix; δ i Represents the slack variable of the i-th sample; X i Represents the feature vector of the i-th sample in the sample set X; U i The category label of the i-th sample; b represents the intercept; Represents the penalty factor; Represents the slack variable of the i-th sample containing the kernel function θ

[0076] Construct the Lagrangian function, and the expression for determining the category to which the sample data in area A belongs is

[0077]

[0078] Among them, W z Represents the feature weight vector, αi and ζ i denote Lagrange multipliers;

[0079] Use the trained model to classify regions. Input the temperature, latitude, longitude, and altitude characteristics of the region, and based on the sign of, determine the category it belongs to. If the result is positive, it belongs to one category; if the result is negative, it belongs to another category. The model will output the predicted label;

[0080] According to the predicted label, find region B and region D with geographical characteristics similar to those of region A, and further extract the charge and discharge behavior habit data of the optimal level M3 in these regions.

[0081] Furthermore, in step S4, in order to optimize the charge and discharge behavior habits of a new energy operating vehicle in region A, using the optimal level M3 charge and discharge habits of region B and region D similar to region A as the scoring criteria, respectively score and evaluate the charging habit and discharging habit of this new energy operating vehicle. The obtained score is the first score of the charge and discharge behavior habits of the new energy operating vehicle;

[0082] The specific steps for scoring the first score of the discharging behavior habit and the first score of the charging behavior habit of the new energy operating vehicle are as follows:

[0083] ④ Extract the discharging behavior habit characteristics and charging habit characteristics from the charge and discharge behavior habit characteristic sample data of the optimal level M3 in region B and region D, and ensure that the charge and discharge habits correspond one by one to the characteristic sample data;

[0084] The discharging behavior habit characteristics of the optimal level M3 include: the number of sharp decelerations ρ, the number of sharp accelerations the situation of the driving speed level the situation of the unevenness L1 and the slope L2 levels of the road surface in the driving road condition L, the external temperature the situation of the belonging range;

[0085] The charging behavior habit characteristics of the optimal level M3 include: the SOC state, the charging method QS, the temperature during charging the situation of the belonging range, the charging duration the situation of the belonging range;

[0086] ⑤ According to each group of charging and discharging behavior habit characteristics in the charge and discharge behavior habit sample data of the optimal level M3 in region B and region D, determine the threshold range of each charge and discharge behavior habit characteristic that meets the scoring qualification:

[0087] Sort each group of discharge habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range; sort each group of charging habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range;

[0088] Taking the battery SOC charge state of the charging habit characteristics as an example, after sorting it, the highest value is 9 and the lowest value is 3. Set the threshold range [3, 9] that meets the scoring qualification. When evaluating the charging habit SOC state, if it meets this threshold condition, this charging habit enters the specific score evaluation link; if it does not meet this threshold condition, this charging behavior habit has no scoring qualification and is judged as 0 points;

[0089] ⑥According to the regions screened out that are geographically similar to the region where the target new energy operating vehicle is located, calibrate the weight coefficients of the first scores of each discharge behavior habit and charging behavior habit based on the overlap degrees of each discharge behavior habit and charging behavior habit at the optimal level M3 of its region;

[0090] ④Based on the preset threshold conditions and the weight coefficients of the first scores of each charge and discharge behavior habit, evaluate the first scores of the charging behavior habit and the first scores of the discharge behavior habit of the target new energy operating vehicle.

[0091] Further, step ③ is specifically:

[0092] (1) First, establish each piece of charge and discharge behavior habit characteristic data finally screened out, such as the SOC state at the start of charging in the charging habit, as vector characteristics; assume these characteristic data are h = [h1, h2, h3…h n , where n represents the number of characteristic data. For each piece of characteristic data, count its frequency p i that appears in the vector h, and then calculate the overlap degree З of each piece of characteristic data through the following formula i

[0093]

[0094] where p i is the frequency that the characteristic data h i appears in the vector h, and thus the overlap degree of each piece of charge and discharge behavior habit characteristic data can be calculated;

[0095] (2) The overlap degrees of each discharge behavior habit and the overlap degrees of each charging behavior habit calibrate the weight coefficients of the first scores of each charge and discharge habit. Select the highest overlap degree of each piece of habit characteristic data, sort them in descending order, and assign the corresponding first scores of each discharge behavior habit with weight coefficients W1, W2, W3, W4, W5, W6, W iThe discharge behavior characteristics corresponding to the weight coefficients are not repeated, i∈[1,6], and are integers; the corresponding first scores of each charging behavior habit are assigned weight coefficients from large to small, ω1, ω2, ω3, ω4, and ω I The charging behavior habit features corresponding to the weight coefficients are not repeated, I∈[1,4], and are integers;

[0096] (3) Based on the overlap between each item of charge and discharge behavior habit feature data in step (1), the overlap of each item of discharge behavior habit data is sorted from high to low, and the evaluation score of the corresponding discharge behavior habit feature data belongs to a range, level or number of times, and its weight coefficient A1, A2, A3 is assigned in descending order. A i1 The weight coefficient corresponds to the range, level or number of feature data that are marked, and they are not repeated. i1∈[1, n1] is an integer. The other five discharge habits are deduced in the same way. The evaluation scores of the range, level or number are assigned weight coefficient B1. C1, D1, E1, F1, B i2 , C i3 , D i4 、E i5 、F i6 The weight coefficients correspond to the ranges, levels or times of the characteristic data, which are not repeated, where i2, i3, i4, i5, and i6 belong to [1, n2], [1, n3], [1, n4], [1, n5], and [1, n6], respectively, and are integers; according to the same method, the overlapping degree of a charging behavior habit data is sorted from high to low, and the corresponding range, level or number of the charging behavior habit characteristic data is assigned with a weight coefficient a1, a2, a3, and so on from large to small. a i7 The weight coefficients correspond to the ranges, levels or times of the characteristic data that are identified, and they are not repeated. Among them, i7 belongs to [1, n7] and is an integer. The other three charging habits are deduced in the same way. The evaluation scores of the ranges, levels or times are assigned weight coefficients b1. c1, d1, b i8 、c i9 d i10 The weight coefficients correspond to the characteristic data that belong to the same range, level or number of times, where i8, i9, and i10 belong to [1, n8], [1, n9], [1, n 10 ], and it must be an integer.

[0097] Further, step ④ is specifically as follows:

[0098] (1) Clean and prepare multiple groups of characteristic data on the charging and discharging behavior habits of the target new energy operating vehicle to ensure the accuracy of the data, and enter the scoring link;

[0099] (2) Compare and analyze each charging and discharging behavior habit characteristic with the preset threshold range for scoring eligibility. If the habit characteristic meets the threshold range, enter the specific score evaluation link, with the scoring range being (0, 100]. Multiply the full score of 100 by the weight coefficient A corresponding to the level or frequency or range to which the habit characteristic data belongs i1 / a i7 , or B i2 / b i8 etc., and the result is the specific score of the habit data. If the habit characteristic does not meet the threshold range, it is determined to be 0 points;

[0100] (3) Multiply the specific score of the habit data by the weight coefficient W / ω corresponding to the first score of the habit to obtain the first score of the discharging behavior habit or charging behavior habit of this item;

[0101] (4) By analogy, obtain the first score of each discharging behavior habit or charging behavior habit, and perform a separate summation operation on them, then the first scores of the discharging behavior habit and charging behavior habit of the target new energy operating vehicle driver can be obtained;

[0102] In this example, the characteristic data of the discharging behavior habit of a new energy operating vehicle in area A is 6 groups, and the total score range is [0, 100]. Use the weighted summation formula to calculate the first total score of the discharging habit

[0103] RS f = Y1W1A i1 + Y2W2B i2 + Y3W3C i3 + Y4W4D i4 + Y5W5E i5 + Y6W6F i6 (11)

[0104] Y N is the score given to the Nth group of characteristic data of the discharging behavior habit of the target new energy operating vehicle according to the preset threshold conditions. If it meets the threshold conditions, the score is 100; if it does not meet the threshold conditions, the score is 0. N ∈ [1, 6] and is an integer;

[0105] The characteristic data of the charging behavior habit of a new energy operating vehicle in area A is 4 groups, and the total score range is [0, 100]. Use the weighted summation formula to calculate the first total score of the charging habit

[0106] RS c = β1ω1a i7 + β2ω2b i8 + β3ω3c i9 + β4ω4d i10 (12)

[0107] β и is the charging behavior habit of the target new energy commercial vehicle driver, and the score of the first group of characteristic data is based on the preset threshold conditions. If it meets the threshold conditions, the score is 100; if it does not meet the threshold conditions, the score is 0, and it is an integer;

[0108] From this, the first score of the charging behavior habit and the first score of the discharging behavior habit of a new energy commercial vehicle driver in Area A can be calculated;

[0109] From the above steps, it can be obtained that:

[0110] The first score of the discharging behavior habit of the new energy commercial vehicle: RS f

[0111] The first score of the charging behavior habit of the new energy commercial vehicle: RS c .

[0112] Furthermore, in step ④, using the charging and discharging habits of the target new energy commercial vehicle itself as the scoring standard, the second score of the discharging behavior habit and the second score of the charging behavior habit are obtained as supplementary references to the first score;

[0113] The present invention not only has the first score obtained by using the charging and discharging habits of the optimal level in similar regions as the scoring standard as described above, but also sets a reference score, that is, the second score of the charging and discharging habits. The scoring standard of this reference score is the second score scoring rule table established based on the historical charging and discharging behavior habits of the target new energy commercial vehicle, and the calculated score is used as the second score; the collection and utilization of the data in the rule scoring table no longer focus on the charging and discharging habits of the optimal level in similar regions, but are oriented to the charging and discharging habits of the target new energy commercial vehicle itself; the present invention does not rely solely on one scoring result, but combines the comprehensive judgments of two scoring modes to obtain a more accurate and reliable scoring result by making up for each other's strengths and weaknesses;

[0114] In the example, the second score of the charging and discharging behavior habit of a new energy commercial vehicle in Area A is obtained through the number of sharp decelerations ρ and the number of sharp accelerations during a large number of discharging processes of the target new energy commercial vehicle itself The driving speed level The driving road condition L, the outside temperature The range to which it belongs, the SOC state during charging, the charging method QS, the temperature during charging Scope of application, charging duration Historical data of the scope of application situation have different degrees of impact on battery health. A rule table is established. The discharge behavior habits and charging behavior habits of a new energy operating vehicle in area A are input. According to the corresponding scoring rules, the scores obtained for each habit are given, and the scores are added up to obtain respectively:

[0115] Second score of the discharge behavior habit of the new energy operating vehicle: rs f

[0116] Second score of the charging behavior habit of the new energy operating vehicle: rs c .

[0117] Furthermore, step S5 is specifically to select the lower score as the optimization basis to generate the battery charge and discharge optimization strategy for the new energy operating vehicle. The steps are as follows;

[0118] (3) Select the scoring standard corresponding to the lower score as the optimization basis; use the first score RS of the charge and discharge behavior habits of the new energy operating vehicle f 、RS c and the second score rs f 、rs c , conduct score comparison, and use the scoring standard corresponding to the lower score as the optimization basis for the charge and discharge strategy; in general, in the present invention, since the first score of the charge and discharge behavior habits of the new energy operating vehicle is based on the best-level charge and discharge habits in similar areas as the scoring standard, the requirements are more stringent and targeted, so the scores are generally lower. However, it cannot be excluded that when the second score is lower, instead of learning the best-level charge and discharge habits in similar areas, the special case of learning the better charge and discharge habits of the target new energy operating vehicle itself, then the situation of adding the scoring standard corresponding to the second score as the optimization basis is increased;

[0119] When RS f ≤rs f , select the scoring standard of the first score of the discharge behavior habit of the new energy operating vehicle as the optimization basis;

[0120] When RS f >rs f , select the scoring standard of the second score of the discharge behavior habit of the new energy operating vehicle as the optimization basis;

[0121] When RS c ≤rs c , then select the scoring standard of the first score of the charging behavior habit of the new energy operating vehicle as the optimization basis;

[0122] When RS c >rs c, then select the scoring criteria of the second score of the charging behavior habits of new energy commercial vehicles as the optimization basis;

[0123] (4) Generate the battery charge and discharge optimization strategy for new energy commercial vehicles; In the example, through the cloud server and intelligent devices such as smartphones, computers, tablets, etc., information is connected. The cloud server obtains the first score and the second score of the charge and discharge behavior habits of a new energy commercial vehicle in area A, and selects the scoring criteria corresponding to the lower score as the basis for optimizing the score of this charge and discharge behavior habit and the charge and discharge strategy. Through the generation module, a report form of the score of this charge and discharge behavior habit and a report form of the charge and discharge optimization strategy are formed, and they are sent to the driver's intelligent device through the transmission module;

[0124] Composition of the report form of the score of the charge and discharge behavior habit:

[0125] The score of this charging behavior habit and the score of the discharging behavior habit, and the specific score of each charge and discharge behavior habit;

[0126] Composition of the report form of the charge and discharge optimization strategy:

[0127] When the battery of a new energy commercial vehicle starts to charge, the optimal SOC state;

[0128] When the battery of a new energy commercial vehicle is charging, choose fast charging, slow charging or normal charging;

[0129] When the battery of a new energy commercial vehicle is charging, the optimal charging temperature The range it belongs to;

[0130] When the battery of a new energy commercial vehicle is charging, the optimal charging duration The range it belongs to;

[0131] When the battery of a new energy commercial vehicle is discharging, the optimal number of sharp decelerations ρ per 5 km of average distance;

[0132] When the battery of a new energy commercial vehicle is discharging, the optimal number of sharp accelerations per 5 km of average distance

[0133] When the battery of a new energy commercial vehicle is discharging, the optimal driving speed level situation per 5 km of average distance

[0134] When the battery of a new energy commercial vehicle is discharging, the roughness level L1 and the road slope level L2 of the road surface per 5 km of average distance;

[0135] When the battery of a new energy commercial vehicle is discharging, the optimal external temperature The range it belongs to.

[0136] Adopting the above technical solution, the present invention proposes a method for optimizing the battery charge and discharge strategy by evaluating the charge and discharge habits of new energy operating vehicles. Based on other regions with geographical characteristics similar to those of the target new energy operating vehicle, excellent charge and discharge behavior habits are selected as the scoring criteria to obtain the first score of the charge and discharge habits. At the same time, the charge and discharge behavior habits of the target new energy operating vehicle itself are used as the scoring criteria to obtain the second score of the charge and discharge habits. Then, according to the scoring, the battery charge and discharge strategy is optimized. The main benefits are as follows:

[0137] (1) It is possible to obtain the charge and discharge habits of the batteries of new energy operating vehicles in various regions. The driving recorder, battery management system BMS, and various sensors monitor the charge and discharge conditions of the battery, and uniformly transmit the monitored charge and discharge condition information to the cloud server, realizing detailed monitoring and acquisition of the charge and discharge behavior habits of new energy operating vehicles, which is convenient for analyzing and evaluating the charge and discharge behavior habits of new energy operating vehicles.

[0138] (2) It has the function of learning the charge and discharge behavior habits that have a better impact on the battery in similar regions. By continuously learning the excellent charge and discharge behavior habits of new energy operating vehicles in similar regions, it changes its own behavior habits that are not good for the battery life, and has a positive impact on the battery life of new energy operating vehicles.

[0139] (3) This method uses the excellent charge and discharge behavior habits of new energy operating vehicles in similar regions as the scoring criteria for the first score, and its own charge and discharge behavior habits as the scoring criteria for the second score. Based on these two scoring criteria, the function of scoring the charge and discharge behavior habits of the target new energy operating vehicle can be realized. The driver can optimize the charge and discharge strategy according to the score, adjust the charge and discharge behavior habits of the new energy operating vehicle, maintain good charge and discharge habits, improve the charge and discharge habits that are unfavorable to the battery, extend the service life of the battery, thereby reducing the use cost of new energy operating vehicles and improving the performance of new energy operating vehicles. Description of the Drawings

[0140] Figure 1 It is the general flowchart for optimizing the battery charge and discharge strategy of new energy operating vehicles in the present invention;

[0141] Figure 2 It is the device connection diagram for obtaining the charge and discharge behavior habits of new energy operating vehicles in the present invention;

[0142] Figure 3 It is the flowchart for evaluating the battery health level after one operation cycle of new energy operating vehicles in the present invention;

[0143] Figure 4 It is the method flowchart for screening the optimal level charge and discharge habits of regions similar to the target region based on the SVM model in the present invention;

[0144] Figure 5 It is the functional flowchart of the first score scoring module for evaluating the charging and discharging behavior habits of the target of the present invention;

[0145] Figure 6 It is the flowchart of the first score scoring link for evaluating the discharging behavior habit and the charging behavior habit of the present invention;

[0146] Figure 7 It is the functional flowchart of the second score scoring module for evaluating the charging and discharging behavior habits of the target of the present invention;

[0147] Figure 8 It is the flowchart of optimizing the charging strategy based on the score of the present invention;

[0148] Figure 9 It is the flowchart of optimizing the discharging strategy based on the score of the present invention. Specific embodiments

[0149] The following further describes in detail the embodiments of the present invention in conjunction with the drawings and embodiments.

[0150] As Figure 1 shown, a method for optimizing the battery charging and discharging strategy of a new energy commercial vehicle includes the following steps:

[0151] S1. Monitor the charging and discharging habits: Use a driving recorder, a battery management system BMS, and a variety of sensors to fully and comprehensively monitor the charging and discharging behavior habits of the new energy commercial vehicle, obtain the charging and discharging habit data information of the new energy commercial vehicle and transmit it to the cloud server;

[0152] S2. Evaluate the battery health level after one operation cycle: By monitoring the charging and discharging behavior habits of a new energy commercial vehicle in a certain area and recording it with one year of operation as a cycle, according to the battery parameter indicators, establish a K-means evaluation model to evaluate the battery health level of the new energy commercial vehicle, which is divided into three grades: m1, m2, and m3. Among them, m3 is the optimal grade, m2 is the medium grade, and m1 is the worst grade. Since the high or low battery health level reflects the quality of the charging and discharging habits, only the optimal grade M3 charging and discharging behavior habit data recorded by the batteries with the optimal grade m3 are extracted;

[0153] S3. Screen the optimal grade charging and discharging habits of areas similar to the target area: Determine the area where the target new energy commercial vehicle whose charging and discharging habits need to be optimized is located, and use it as the target area. Use the optimal grade M3 charging and discharging behavior habit data as the database, preset the threshold ranges of the temperature, latitude, longitude, and altitude of the target area, and use the SVM model algorithm for classification. Through this classification process, select the areas that meet the conditions and extract the optimal grade M3 charging and discharging behavior habit data of these areas from the database;

[0154] S4. Target charging and discharging behavior habit scoring: According to the optimal level M3 charging and discharging behavior habits of the regions similar to the region where the target new energy operation vehicle is located finally selected, extract multiple groups of discharge habit characteristic sample data and multiple groups of charging habit characteristic sample data therefrom; preset the threshold range for each charging and discharging habit characteristic that meets the scoring qualification; calibrate the weight coefficient of the first score of each charging and discharging behavior habit based on the high or low overlap degree of each finally selected charging and discharging behavior habit; evaluate the first score of the discharge behavior habit and the first score of the charging behavior habit of the target new energy operation vehicle based on the preset threshold range and the weight coefficient of the first score of the charging and discharging behavior habit; at the same time, use the target new energy operation vehicle's own charging and discharging habits as the scoring standard to obtain the second score of the discharge behavior habit and the second score of the charging behavior habit as a supplementary comparison to the first score. This method can more comprehensively evaluate the performance of the target driver's charging and discharging behavior habits in the overall scoring and optimize it more accurately;

[0155] S5. Optimize the charging and discharging strategy based on the score: By comparing the first score of the charging and discharging behavior habits of the new energy operation vehicle with the second score of the charging and discharging behavior habits, select the scoring standard corresponding to the lower score as the optimization basis to generate the battery charging and discharging optimization strategy for the new energy operation vehicle. The driver takes measures to adjust the battery charging and discharging strategy according to this strategy to achieve the purpose of extending the battery life and improving the sustainability and economy of the new energy operation vehicle.

[0156] As Figure 2 shown, in step S1, a driving recorder and multiple sensors are used to monitor the discharge behavior habits of the new energy operation vehicle, and collect the following parameter real-time data: the number of emergency brakes (sharp decelerations) ρ during discharge, the number of sharp accelerations The situation of the driving speed level The driving road condition L, the external temperature The belonging range, form the discharge behavior habit, and transmit the collected data information into the cloud server; the specific content is as follows:

[0157] (A) The driving recorder detects the sharp deceleration and sharp acceleration conditions of the vehicle through the built-in acceleration sensor (G sensor); the acceleration sensor continuously collects the acceleration data of the vehicle, sets a specific acceleration threshold of 3m / s 2 , and a duration threshold of 2s. When the positive acceleration threshold and the time threshold are exceeded, the event record of sharp acceleration is triggered. When the negative acceleration is less than the set value and greater than the time threshold, the event record of sharp deceleration is triggered. Determine the number of emergency brakes and sharp accelerations of this new energy operation vehicle in a section of the road, and record the sharp acceleration habit and sharp deceleration habit in the discharge behavior of the new energy operation vehicle.

[0158] (B) The driving recorder monitors the current position of the vehicle in the form of longitude and latitude through the GPS module, calculates the driving speed of the vehicle. The present invention sets two key parameters: the speed threshold and the duration threshold. If the monitored driving information is greater than the speed threshold of 40 km / h and the duration threshold of 1 minute, it is determined to be high-speed driving. At the same time, for speed ranges such as (40 km / h, 50 km / h], (50 km / h, 60 km / h], (60 km / h, 70 km / h], etc., and the duration exceeding the duration threshold, the high-speed driving levels are classified as I, II, III... in sequence. Within the range not exceeding the threshold, the normal driving levels are classified according to the speed and duration thresholds; for speed ranges in (30 km / h, 40 km / h], (20 km / h, 30 km / h], (10 km / h, 20 km / h], (0 km / h, 10 km / h], and meeting or exceeding the duration threshold, the normal driving levels are classified as 4, 3, 2, 1 in sequence, so as to judge the level of the current driving speed and record the driving speed habits in the discharge behavior of new energy commercial vehicles.

[0159] (C) The vehicle dynamics sensor and the road surface vibration sensor installed at the bottom of the vehicle are used to monitor the driving road conditions of the vehicle. The driving road conditions L consist of two aspects: the road surface unevenness L1 and the road surface slope L2. The present invention uses the International Roughness Index (IRI) as an index to evaluate the road surface unevenness level, establishes a calculation mathematical model for the road surface unevenness level using the definition of IRI, and verifies its correctness through multiple sets of data. The vibration sensor can infer the load spectrum of the road section based on the vibration spectrum and amplitude generated when the vehicle passes. After evaluating the road surface level through the road section load spectrum, the corresponding road surface unevenness coefficient G d (n0) is found. On roads with different levels, when new energy commercial vehicles drive at different speeds, the road surface will generate time-domain excitation signals for the vehicle, and these signals change over time, reflecting the dynamic response characteristics of the vehicle to road surface excitation;

[0160] The time-domain excitation signals corresponding to various road surface levels and various vehicle speeds are the road surface excitation time-frequency power spectral density G d (n), which can be expressed by Equation (1):

[0161]

[0162] v b is the vehicle speed driving on this road surface, Ω is the time frequency for measuring the vehicle driving on this road surface, and n0 is the reference spatial frequency, usually 0.1 m -1 .

[0163] Calculate the current road surface unevenness IRI level value, and the formula is:

[0164]

[0165] Among them, q is the correction coefficient, which is usually 0.78 verified by multiple groups of real vehicle data;

[0166] The classification of the road surface unevenness level L1 is shown in Table 1, and different IRIs correspond to different road surface unevenness levels

[0167] Table 1 Corresponding table of road surface unevenness levels

[0168]

[0169] The estimation of the driving road surface slope L2 of the present invention is established based on the least square estimation model of longitudinal dynamics. When a new energy commercial vehicle is driving on a slope, the state variables at different moments, such as speed, acceleration, steering angle, etc., and the vehicle position information provided by GPS are obtained by the vehicle dynamics sensor. The dynamic system estimators of the system parameter w at the i-1th moment and the ith moment are respectively defined as The weights between the predicted value and the actual value of the slope at the i-1th moment and the ith moment are P(i - 1) and P(i) respectively. The Kalman gain used at the ith moment is K(i), and the system output at the ith moment is F if (i). The observable data vector at the ith moment is μ(i). Then, the road surface slope model based on the least square estimation of longitudinal kinematics is established as:

[0170]

[0171] Among them, the initial value P(0) = Z×E is determined. Z is a sufficiently large positive real number, and E usually represents the prediction error covariance matrix, that is, the uncertainty of slope estimation; τ Z is the forgetting factor of the least square estimation model of the road surface slope based on longitudinal dynamics. The selection range of the forgetting factor is (0, 1]. Considering the change characteristics of the road surface slope, the influence of old observation data on the new reference estimation is small. Therefore, the forgetting factor τ Z = 0.9 is selected, and the value of is calculated, then the dynamic system estimator of the system parameter w at the ith moment is obtained;

[0172] From w = sinα+(f0 + f1v p ), the road surface slope can be deduced as

[0173]

[0174] Among them, v p is the driving speed of the commercial vehicle at the ith moment, and f0 and f1 are the constant term and the first-order coefficient of the speed fitting of the rolling resistance coefficient;

[0175] The present invention classifies the road surface gradient L2 into five levels: level 1 (0%-1%), level 2 (1%-5%), level 3 (5%-10%), level 4 (10%-15%), and level 5 (above 15%).

[0176] Respectively evaluate and record the levels of the road surface unevenness L1 and the road surface gradient L2 in the driving road conditions information in the current discharge behavior habits of new energy commercial vehicles;

[0177] (D) The driving recorder records the external temperature during the discharge of new energy commercial vehicles. For the division of the applicable range of the external temperature in the present invention, every 5°C is set as a threshold range. The temperature range of 0°C and above is divided into i intervals, which are respectively marked as r1, r2... r , and the temperature range below 0°C is also divided into i intervals, which are respectively marked as -r1, -r2... -r i , and classify the subordinate range of the current external temperature when the new energy commercial vehicle is driving into the corresponding interval for recording; i , Based on monitoring the above-mentioned number of hard brakes ρ, the number of hard accelerations

[0178] The driving speed level situation , The levels of the road surface unevenness L1 and the road surface gradient L2 in the driving road conditions L, and the situation of the external temperature The subordinate range situation, to obtain the discharge behavior habits of new energy commercial vehicles;

[0179] (E) Taking an example recorded in Table 2, the habits of a certain new energy commercial vehicle in a certain area A during a certain discharge behavior were monitored.

[0180] Table 2 Monitoring Table of the Discharge Behavior Habits of a Certain New Energy Commercial Vehicle in Area A

[0181]

[0182] This example monitoring method is also applicable to monitoring the discharge behavior habits of all new energy commercial vehicles in this area and other areas.

[0183] In step S1, using the battery management system BMS to monitor the charging behavior habits of new energy commercial vehicles includes: by monitoring the voltage, current, temperature, and charging time parameters during charging, estimating the battery SOC state, charging method QS, and charging temperature Charging duration Four charging habit indicators, and transmitting the charging habit data information into the cloud server; the specific content is as follows:

[0184] (a) During charging, the BMS monitors voltage, current, and temperature parameters and uses an improved Kalman filtering algorithm to estimate the state of charge (SOC) of the battery. Compared with the traditional Kalman filtering algorithm, this algorithm can improve the calculation accuracy of the battery SOC.

[0185] Using the monitored battery output voltage U C , the battery working current I, the target voltage U m of the battery, and the battery working duration T h , calculate I×T h to obtain the maximum available capacity C max of the battery, and then further calculate the value of SOC. The calculation formula for the state of charge (SOC) of the charging battery is:

[0186]

[0187] where SOC(t1) in the formula represents the real-time SOC value, SOC(t0) represents the initial battery charge, C max represents the maximum available capacity, η represents the battery Coulomb efficiency, and I(t) represents the function of current and time;

[0188] Divide the battery SOC state [0%, 100%] into 10 equal SOC threshold ranges. [0%, 10%] is the first range, [10%, 20%] is the second range, and so on to [90%, 100%] as the tenth range. Estimate in which range the SOC state is when the new energy commercial vehicle starts charging, and record the corresponding data information;

[0189] (b) During charging, the BMS monitors the real-time voltage and real-time current data and calculates the average power P avg during charging. The formula is:

[0190]

[0191] where represents the charging duration;

[0192] Set specific power thresholds. The power threshold for fast charging is 150KW, and the power threshold for slow charging is 50KW. When the new energy commercial vehicle is charging, classify the charging method QS according to the calculated average power P avg . If the average power is greater than the set fast charging power threshold, this charging is classified as fast charging; if the average power is less than the slow charging power threshold, this charging is classified as slow charging, and record the charging data information for this time; if the average power is in the range of [50KW, 150KW], this charging is classified as normal charging (general charging).

[0193] (c) Monitor the temperature during charging through the temperature sensor in the BMS. For the division of the charging temperature range in the present invention, set every 5°C as a threshold range. Divide the temperature range of 0°C and above into i intervals, which are respectively marked as g1, g2... g , and divide the temperature range below 0°C into i intervals as well, which are respectively marked as -g1, -g2... -g i , and classify the current charging temperature range of the new energy operation vehicle into the corresponding interval for recording; i ,

[0194] (d) Record the start time of charging through the BMS, and monitor the change of current to determine the end time of charging, and then calculate the charging duration Set every 0.5h of charging duration as a threshold range, divide the charging duration into i intervals, which are respectively marked as Q1, Q2... Q i , and classify the current charging duration range of the new energy operation vehicle into the corresponding interval for recording;

[0195] Based on monitoring the SOC state, charging method QS, the temperature range during charging , and the charging duration range , obtain the charging behavior habits of new energy operation vehicles;

[0196] (e) Taking an example recorded in Table 3, monitor the habits of a certain new energy operation vehicle in a certain charging behavior in Area A.

[0197] Table 3 Monitoring Table of the Charging Behavior Habits of a Certain New Energy Operation Vehicle Driver in Area A

[0198]

[0199] This example monitoring method is also applicable to monitoring the charging behavior habits of all new energy operation vehicles in this area and other areas.

[0200] In step S2, establish a K-means evaluation model to evaluate the battery health level of new energy operation vehicles after one operation cycle. The implementation principle process is as follows:

[0201] (1) Input the seven parameters of the current battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charging efficiency C R , and discharge efficiency F R into the K-means algorithm for clustering analysis, and divide them into K equidistant clusters, so that each cluster has an initial center point;​​

[0202] (2) The input data is distributed into each cluster until all the input data is assigned to the new clusters. And the total number of samples S in the input dataset is assigned to K clusters. This is achieved by the K-means algorithm continuously adjusting the centroid positions of the clusters to minimize the sum of the distances from the points to the centroids, that is, minimizing the squared error J. For the minimized squared error J of the cluster partition γ obtained by clustering;

[0203]

[0204] where γ = {γ1, γ2, …, γ k} represents the clusters of the battery health level of K types, is the current battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charging efficiency C R , discharge efficiency F R feature matrix, represents the centroid of the i-th cluster;

[0205] (3) Randomly select K cluster centroids of the battery health level feature samples, measure the distances between the samples and each centroid, and assign each sample to the nearest cluster centroid. Iterate n times in this way. During each iteration, update the centroids of each cluster using the mean. Repeat the above steps for the K cluster centroids until the cluster centroids are stable or the function converges.

[0206] As Figure 3 shown, the example process of evaluating the battery health level after an operation cycle of a new energy operating vehicle in a certain area specifically includes the following steps:

[0207] Step 1: Train the K-means algorithm. Input the battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charging efficiency C R , discharge efficiency F R of a new energy operating vehicle in a certain area after an operation cycle. During the training process, the algorithm will use these parameters to learn the features of the data and adjust the relevant parameters according to the training results;

[0208] Step 2: If the input parameter data is training data, perform data classification detection, calculate the corresponding grade label for each data, and thus achieve grade division. Calculate the mean of all the data in each grade, and the obtained result is the optimal point of the corresponding grade; Use the optimal point as the initial cluster center of the K-means clustering algorithm;

[0209] Step 3: If the input parameter data is not training data, directly use the K-means algorithm with initial cluster centers for clustering. According to the distance between the class and the data, divide the data into the nearest class to generate a new clustering result. If the data in a certain class changes, the algorithm will recalculate the division level until the optimal value is found and then stop.

[0210] Based on the above steps, the battery health assessment model can be divided into three battery health level grades: m1, m2, and m3. Among them, m3 is the optimal grade, m2 is the medium grade, and m1 is the worst grade. Use this to evaluate the battery health level of a new energy operating vehicle in a certain area after one operation cycle. Since the high or low battery health level reflects the quality of the charging and discharging habits, the battery health level of the optimal grade m3 corresponds to the charging and discharging behavior habits of the optimal grade M3. And so on for M1 and M2. The present invention only extracts the data of the charging and discharging behavior habits of the optimal grade M3 recorded by the batteries whose battery health levels are evaluated as the optimal grade m3, and those rated as M1 and M2 grades will not be used as references.

[0211] In step S3, in order to optimize the charging and discharging behavior habits of a new energy operating vehicle in area A, determine area A as the target area, and use the SVM model to screen areas with geographical features similar to those of area A, and select the charging and discharging habits of the optimal grade M3 from them.

[0212] In the process of data screening based on the SVM model algorithm, geographical features such as temperature, latitude, longitude, and altitude are used as key parameters for training the model. The SVM model classifies data with similar attributes through these features. The specific steps for screening data using the SVM model are as follows Figure 4 as shown

[0213] Step 1) Establish a database divided according to geographical location, which contains the temperature, latitude, longitude, and altitude characteristic signals of each area. Extract the data of these features from it, and collect the sample data with temperature, latitude, longitude, and altitude information of each area. Each sample should include these feature values and a target label for indicating the category to which the sample belongs.

[0214] Step 2) Preprocess the data, including handling missing values, normalizing features, etc. The goal of this process is to ensure that the data set can be smoothly used for training the model. Subsequently, the data set is split into a training set and a test set.

[0215] Step 3) Use the support vector machine algorithm to train the model. During the training process, the algorithm will learn how to map the input features to the corresponding target labels. The support vector machine model identifies the type of the sample, and it presets a kernel function θ. Use the regression model samples to construct a cost function whose expression is

[0216]

[0217] Apply constraint conditions:

[0218]

[0219] δ i ≥0, i = 1, 2, 3, 4…

[0220] where ω z represents the classification interface vector; represents the transpose of the classification interface vector matrix; δ i represents the slack variable of the i-th sample; X i represents the feature vector of the i-th sample in the sample set X; U i is the class label of the i-th sample; b represents the intercept; represents the penalty factor; represents the slack variable of the i-th sample containing the kernel function θ.

[0221] Construct the Lagrangian function, and the expression for judging the class of the sample data in area A is

[0222]

[0223] where W z represents the feature weight vector, α i and ζ i represent the Lagrange multipliers.

[0224] Use the trained model to classify the area, input the temperature, latitude, longitude and altitude features of the area, and judge the class according to the sign of, if the result is positive, it belongs to one class, if the result is negative, it belongs to another class, and the model will output the predicted label;

[0225] According to the predicted label, find areas B and D with geographical features similar to area A, and further extract the charge and discharge behavior habit data of the optimal M3 level in these areas.

[0226] In step S4, in order to optimize the charge and discharge behavior habits of a new energy operating vehicle in area A, taking the charge and discharge habits of the optimal M3 level in areas B and D similar to area A as the scoring criteria, the charging habits and discharge habits of this driver are scored and evaluated respectively, and the obtained score is the first score of the charge and discharge behavior habits of the new energy operating vehicle; at the same time, taking the charge and discharge habits of the target new energy operating vehicle itself as the scoring criteria, the second score of the discharge behavior habit and the second score of the charging behavior habit are obtained as a supplementary comparison to the first score.

[0227] As Figure 5 shown, the specific steps for scoring the first score of the discharge behavior habit and the first score of the charging behavior habit of new energy operation vehicles are as follows:

[0228] ① Extract the discharge behavior habit characteristics and charging behavior habit characteristics from the sample data of the charging and discharging behavior habits of the optimal level M3 in Region B and Region D, and ensure that the charging and discharging habits correspond one by one to the characteristic sample data;

[0229] The discharge behavior habit characteristics of new energy operation vehicles at the optimal level M3 include: the number of hard brakes ρ, the number of hard accelerations The situation of the driving speed level The situation of the unevenness L1 and the slope L2 levels of the road surface in the driving road condition L, the outside temperature The situation of the affiliated range;

[0230] The charging behavior habit characteristics of new energy operation vehicles at the optimal level M3 include: the SOC state, the charging method QS, the temperature during charging The situation of the affiliated range, the charging duration The situation of the affiliated range;

[0231] ② According to each group of charging and discharging behavior habit characteristics in the sample data of the charging and discharging behavior habits of the optimal level M3 in Region B and Region D, determine the threshold range of each charging and discharging behavior habit characteristic that meets the scoring qualification;

[0232] Sort each group of discharge habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range; sort each group of charging habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range;

[0233] Taking the battery SOC charge state of the charging habit characteristic as an example, after sorting it, the highest value is 9 and the lowest value is 3. Set the threshold range [3, 9] that meets the scoring qualification. When evaluating the SOC state of the charging habit, if it meets this threshold condition, this charging habit enters the specific score evaluation link. If it does not meet this threshold condition, this charging behavior habit has no scoring qualification and is judged as 0 points;

[0234] ③ According to the regions with geographical characteristics similar to those of the target new energy operation vehicle, calibrate the weight coefficients of the first scores of each discharge behavior habit and charging behavior habit based on the high and low degrees of overlap of the discharge behavior habits and charging behavior habits of the optimal level M3 in its region;

[0235] ⑦ Based on the preset threshold conditions and the weight coefficients of the first scores of each charging and discharging behavior habit, evaluate the first score of the charging behavior habit and the first score of the discharge behavior habit of the target new energy operation vehicle.

[0236] Step ③ is as follows:

[0237] (1) First, each of the finally selected charging and discharging behavior habit feature data, such as the SOC state at the beginning of charging in the charging habit, is established as a vector feature. Assume that these feature data are represented by h = [h1, h2, h3 ... h n ] Where n represents the number of feature data, for each feature data, count its frequency p in vector h i , and then calculate the overlap of each feature data using the following formula i

[0238]

[0239] Its i is the characteristic data h i The frequency of occurrence in vector h, from which the overlap of each charging and discharging behavior habit feature data can be calculated;

[0240] (2) The weight coefficients of the first scores of each charging and discharging habit are calibrated by the overlap of each discharge behavior habit and the overlap of each charging behavior habit. The highest overlap of each habit feature data is selected and sorted from high to low. The corresponding first scores of each discharge behavior habit are assigned weight coefficients from large to small as W1, W2, W3, W4, W5, W6, W7, W8, W9, W10, W11, W12, W13, W14, W15, W16, W17, W18, W19, W20, W21, W22, W33, W4, W5, W6, W7, W8, W9, W10, W11, W21 ... i The discharge behavior characteristics corresponding to the weight coefficients are not repeated, i∈[1,6], and are integers; the corresponding first scores of each charging behavior habit are assigned weight coefficients from large to small, ω1, ω2, ω3, ω4, and ω I The charging behavior habit features corresponding to the weight coefficients are not repeated, I∈[1,4], and are integers;

[0241] (3) Based on the overlap between each item of charge and discharge behavior habit feature data in step (1), the overlap of each item of discharge behavior habit data is sorted from high to low, and the evaluation score of the corresponding discharge behavior habit feature data belongs to a range, level or number of times, and its weight coefficient A1, A2, A3 is assigned in descending order. A i1 The weight coefficient corresponds to the range, level or number of feature data that are marked, and they are not repeated. i1∈[1, n1] is an integer. The other five discharge habits are deduced in the same way. The evaluation scores of the range, level or number are assigned weight coefficient B1. C1, D1, E1, F1, B i2 , C i3 , Di4 , E i5 , F i6 The ranges, levels, or frequencies to which the characteristic data corresponding to the weight coefficients belong do not repeat each other, where i2, i3, i4, i5, and i6 belong to [1, n2], [1, n3], [1, n4], [1, n5], and [1, n6] respectively, and are integers; in the same way, the overlap degrees of a set of charging behavior habit data are sorted from high to low, and the ranges, levels, or frequencies to which the corresponding characteristic data of this charging behavior habit belong, and their evaluations The estimated scores are assigned with weight coefficients a1, a2, a3 from large to small in turn a i7 Characteristic data corresponding to the weight coefficient identifier The ranges, levels, or frequencies do not repeat each other, where i7 belongs to [1, n7], and is an integer. By analogy for the other three charging habits, the evaluation scores of the ranges, levels, or frequencies are assigned the weight coefficient b1, c1, d1, b i8 , c i9 , d i10 The ranges, levels, or frequencies to which the characteristic data corresponding to the weight coefficients belong do not repeat each other, where i8, i9, and i10 belong to [1, n8], [1, n9], and [1, n 10 , and are integers;

[0242] For example Figure 6 , step ④ is specifically as follows:

[0243] (1) Clean and prepare multiple sets of characteristic data on the charging and discharging behavior habits of the target new energy operating vehicle to ensure the accuracy of the data and enter the scoring process;

[0244] (2) Compare and analyze each charging and discharging behavior habit characteristic with the preset threshold range for scoring eligibility. If the habit characteristic meets the threshold range, enter the specific score evaluation process. The scoring range is (0, 100], and use 100 full marks multiplied by the weight coefficient A i1 / a i7 , or B i2 / b i8 etc., and the result is the specific score of this habit data. If the habit characteristic does not meet the threshold range, it is judged as 0 points;

[0245] (3) Multiply the specific score of this habit data by the weight coefficient W / ω corresponding to the first score of this habit to obtain the first score of this discharging behavior habit or charging behavior habit;

[0246] (4) By analogy with this method, the first score of each discharging behavior habit or charging behavior habit is obtained. By performing a separate summation operation on them, the first scores of the discharging behavior habit and charging behavior habit of the target new energy operating vehicle driver can be obtained;

[0247] In this example, the characteristic data of the discharging behavior habit of a new energy operating vehicle in Area A is 6 groups, and the total score range is [0, 100]. The weighted summation formula is used to calculate the first total score of the discharging habit:

[0248] RS f = Y1W1A i1 + Y2W2B i2 + Y3W3C i3 + Y4W4D i4 + Y5W5E i5 + Y6W6F i6 (11)

[0249] Y N is the score given to the Nth group of characteristic data of the discharging behavior habit of the target new energy operating vehicle according to the preset threshold conditions. If it meets the threshold conditions, the score is 100; if it does not meet the threshold conditions, the score is 0, where N ∈ [1, 6] and is an integer;

[0250] The characteristic data of the charging behavior habit of a new energy operating vehicle in Area A is 4 groups, and the total score range is [0, 100]. The weighted summation formula is used to calculate the first total score of the charging habit

[0251] RS c = β1ω1a i7 + β2ω2b i8 + β3ω3c i9 + β4ω4d i10 (12)

[0252] is the score given to the charging behavior habit of the target new energy operating vehicle driver for the group of characteristic data according to the preset threshold conditions. If it meets the threshold conditions, the score is 100; if it does not meet the threshold conditions, the score is 0, and is an integer;

[0253] Thus, the first score of the charging behavior habit and the first score of the discharging behavior habit of a new energy operating vehicle in Area A can be calculated.

[0254] From the above steps, it can be obtained that

[0255] The first score of the discharging behavior habit of the new energy operating vehicle: RS f

[0256] The first score for the charging behavior habits of new energy commercial vehicles: RS c

[0257] The present invention not only has the first score obtained by using the optimal charging and discharging habits in other regions similar to the target region as the scoring standard, but also sets a control score, that is, the second score for charging and discharging habits. The scoring standard for this control score is the second score scoring rule table established based on the historical charging and discharging behavior habits of the target new energy commercial vehicle, and the calculated score is used as the second score. The collection and utilization of the data in the rule scoring table no longer focus on the optimal charging and discharging habits in other regions similar to the target region, but are oriented to the charging and discharging habits of the target new energy commercial vehicle. The present invention combines two scoring modes for comprehensive judgment, so as to obtain a more accurate and reliable scoring result by making up for each other's strengths and weaknesses.

[0258] Such as Figure 7 , in the example, the second score for the charging and discharging behavior habits of a new energy commercial vehicle in Area A is obtained through the number of sharp deceleration times ρ, the number of sharp acceleration times The driving speed level situation The driving road condition L, the external temperature The range to which it belongs, the SOC state during charging, the charging method QS, the temperature during charging The range to which it belongs, the charging duration The historical data of the range to which it belongs, which has different degrees of influence on the battery health, establish a rule table as shown in Table 4, input the discharging behavior habits and charging behavior habits of a new energy commercial vehicle in Area A, give the scores obtained for each habit according to the corresponding scoring rules, and sum up the scores to respectively obtain:

[0259] The second score for the discharging behavior habits of the driver of the new energy commercial vehicle: rs f

[0260] The second score for the charging behavior habits of the driver of the new energy commercial vehicle: rs c

[0261] Table 4 Scoring rule table for the second score of charging and discharging behavior habits

[0262]

[0263]

[0264]

[0265] Step S5 is specifically as follows: Based on selecting the scoring standard corresponding to the lower score as the optimization basis, generate an optimization strategy for the battery charging and discharging of new energy commercial vehicles, and the steps are as follows;

[0266] (1) Select the scoring criteria corresponding to the lower score as the optimization basis. Utilize the first score RS of the charging and discharging behavior habits of new energy commercial vehicles f 、RS c and the second score rs f 、rs c , conduct a score comparison, and take the scoring criteria corresponding to the lower score as the optimization basis for the charging and discharging strategy. In general, for the present invention, since the first score of the charging and discharging behavior habits of new energy commercial vehicles is based on the charging and discharging habits of the optimal level in similar regions as the scoring criteria, the requirements are more stringent and targeted, so the scores are generally lower. However, it cannot be excluded that when the second score is lower, it may not be necessary to learn the charging and discharging habits of the optimal level in similar regions, but instead it may be more appropriate to learn the charging and discharging habits of the target new energy commercial vehicle itself. In this special case, the scoring criteria corresponding to the second score need to be selected as the optimization basis.

[0267] When RS f ≤ rs f , select the scoring criteria of the first score of the discharging behavior habits of new energy commercial vehicles as the optimization basis;

[0268] When RS f > rs f , select the scoring criteria of the second score of the discharging behavior habits of new energy commercial vehicles as the optimization basis;

[0269] When RS c ≤ rs c , select the scoring criteria of the first score of the charging behavior habits of new energy commercial vehicles as the optimization basis;

[0270] When RS c > rs c , select the scoring criteria of the second score of the charging behavior habits of new energy commercial vehicles as the optimization basis.

[0271] (2) Generate the battery charging and discharging optimization strategy for new energy commercial vehicles. In the example, through information connection between the cloud server and intelligent devices such as smartphones, computers, tablets, etc., the cloud server obtains the scoring criteria of the lower score selected from the first score and the second score of the charging and discharging behavior habits of a certain new energy commercial vehicle in area A as the score of this charging and discharging behavior habit and the optimization basis for the charging and discharging strategy. Through the generation module, a score report form of this charging and discharging behavior habit and a charging and discharging optimization strategy report form are formed, and through the sending module, they are sent to the intelligent device of the driver.

[0272] Composition of the score report form of the charging and discharging behavior habit:

[0273] The score of this charging behavior habit and the score of the discharging behavior habit, and the specific score value of each charging and discharging behavior habit.

[0274] Report on Charge and Discharge Optimization Strategy Composition:

[0275] When the battery of a new energy operating vehicle starts charging, the optimal SOC state;

[0276] When the battery of a new energy operating vehicle is charging, whether to choose fast charging, slow charging or normal charging;

[0277] When the battery of a new energy operating vehicle is charging, the optimal charging temperature Range;

[0278] When the battery of a new energy operating vehicle is charging, the optimal charging duration Range;

[0279] When the battery of a new energy operating vehicle is discharging, the optimal number of sharp decelerations ρ for an average 5-km journey;

[0280] When the battery of a new energy operating vehicle is discharging, the optimal number of sharp accelerations for an average 5-km journey

[0281] When the battery of a new energy operating vehicle is discharging, the optimal driving speed level for an average 5-km journey

[0282] When the battery of a new energy operating vehicle is discharging, the roughness level L1 and slope level L2 of the road surface for an average 5-km journey;

[0283] When the battery of a new energy operating vehicle is discharging, the optimal external temperature Range.

[0284] The charge optimization strategy for the battery of a new energy operating vehicle is as Figure 8 .

[0285] If the scoring standard of the first score of the charging behavior habit of a new energy operating vehicle is selected as the optimization basis, then learn from and refer to the battery charging behavior habits of the optimal level M3 in regions B and D with geographical characteristics similar to those of region A, calculate the overlap degree of the four groups of data of the optimal charging habits in regions similar to region A, and select the data with the highest overlap degree for each charging habit as the best data, which is used as the optimized charging strategy. Taking an example, if the data with the highest overlap degree of the SOC state is 8, then for this new energy operating vehicle driver in region A, the optimal SOC state when the battery starts charging is 70%-80%.

[0286] If the scoring standard of the second score of the charging behavior habit of a new energy operating vehicle is selected as the optimization basis, then learn from and refer to its own charging habits, and select the data with the highest score for each charging habit in the second score scoring rule table of charge and discharge habits as the best data, which is used as the optimized charging strategy. Taking an example, the optimal charging temperature The data with the highest score in the range is g2 - g5. Then, for the drivers of new energy operation vehicles in Area A during battery charging, the optimal charging temperature is in the range of 11°C - 25°C.

[0287] The optimization strategy for the battery discharge of new energy operation vehicles is as Figure 9 .

[0288] If the scoring standard of the first score of the discharge behavior habit of new energy operation vehicles is selected as the optimization basis, learn from and refer to the battery discharge behavior habits of the optimal level of M3 in Area B and Area D with geographical locations similar to that of Area A, calculate the overlap degree of six groups of data of the optimal level discharge habits in areas similar to Area A, and select the data with the highest overlap degree for each discharge habit as the best data, which is used as the optimized discharge strategy. Taking an example, the data with the highest overlap degree in the driving speed level situation is 3. Then, for this new energy operation vehicle in Area A during battery discharge, the optimal driving speed for an average 5 - km journey is 20 km / h - 30 km / h.

[0289] If the scoring standard of the second score of the discharge behavior habit of new energy operation vehicles is selected as the optimization basis, learn from and refer to its own discharge habits, and select the data with the highest score for each discharge habit in the second - score scoring rule table of charge - discharge habits as the best data, which is used as the optimized discharge strategy. Taking an example, the data with the highest score for the best hard - deceleration times is 0 - 4. Then, for the battery discharge of this new energy operation vehicle in Area A, the optimal hard - deceleration times for an average 5 - km journey is 0 - 4 times.

[0290] (3) According to the feedback of the charge - discharge behavior habit score report and the charge - discharge optimization strategy report, the driver of this new energy operation vehicle can judge the quality of this charge - discharge behavior habit through the score. Through the charge - discharge optimization strategy report, adopt the optimization strategy in the report to improve the bad charge - discharge behavior habits. Good charge - discharge behavior habits need to be continued to maintain, so as to improve the charge - discharge behavior habit score, extend the battery life, improve energy efficiency, and reduce operating costs.

[0291] The present invention is applicable to all regions, not limited to Area A, Area B, etc. in the examples given. At the same time, the optimization method of the present invention is applicable to other charge - discharge habits.

Claims

1. A method for optimizing the battery charging and discharging strategy of a new energy operation vehicle, characterized in that: It includes the following steps: S1. Monitor the charging and discharging habits: Use a driving recorder, a battery management system (BMS), and a variety of sensors to fully and comprehensively monitor the charging and discharging behavior habits of new energy commercial vehicles, obtain the data information of the charging and discharging behavior habits, and transmit it to the cloud server; S2. Evaluate the battery health level after one operation cycle: By monitoring the charging and discharging behavior habits of a new energy commercial vehicle in a certain area and recording it with one year of operation as a cycle, according to the battery parameter indicators, establish a K-means evaluation model to evaluate the battery health level of new energy commercial vehicles, which is divided into three grades: m1, m2, and m3. Among them, m3 is the best grade, m2 is the medium grade, and m1 is the worst grade. Since the high or low battery health level reflects the good or bad charging and discharging behavior habits, only extract the optimal grade M3 charging and discharging behavior habit data recorded by the batteries with the best grade m3; S3. Screen the optimal grade charging and discharging habits of areas similar to the target area: Determine the area where the target new energy commercial vehicle that needs to optimize the charging and discharging habits is located, and use it as the target area. Use the optimal grade M3 charging and discharging behavior habit data as the database, preset the threshold ranges of the temperature, latitude, longitude, and altitude of the target area, and use the SVM model algorithm for classification. Through this classification process, screen out the areas that meet the conditions, and extract the optimal grade M3 charging and discharging behavior habit data of these areas from the database; S4. Score the target charging and discharging behavior habits: According to the optimal grade M3 charging and discharging behavior habits of the areas similar to the target new energy commercial vehicle finally screened out, extract multiple groups of discharge habit characteristic sample data and multiple groups of charging habit characteristic sample data from them; Preset the threshold ranges of each charging and discharging behavior habit characteristic that meets the scoring qualifications; And calibrate the weight coefficient of the first score of each charging and discharging behavior habit based on the high or low overlap degree of each charging and discharging behavior habit finally screened out; Evaluate the first score of the discharge habit and the first score of the charging habit of the target new energy commercial vehicle based on the preset threshold ranges and the weight coefficient of the first score of the charging and discharging behavior habits; At the same time, use the charging and discharging habits of the target new energy commercial vehicle itself as the scoring standard to obtain the second score of the discharge habit and the second score of the charging habit as a supplementary comparison to the first score. This method more comprehensively evaluates the performance of the charging and discharging behavior habits of the target new energy commercial vehicle in the overall scoring and optimizes it more accurately; S5. Optimize the charging and discharging strategy based on the scores: By comparing the first scores of the charging and discharging behavior habits of the new energy commercial vehicle with the second scores of the charging and discharging habits, select the scoring standard corresponding to the lower score as the optimization basis, generate the battery charging and discharging optimization strategy for the new energy commercial vehicle. The driver takes measures to adjust the battery charging and discharging strategy according to this strategy to achieve the purpose of extending the battery life and improve the sustainability and economy of the new energy commercial vehicle.

2. The optimization method for the battery charge and discharge strategy of a new energy operation vehicle according to claim 1, wherein: In step S1, a driving recorder and a variety of sensors are used to monitor the discharging behavior habits of new energy commercial vehicles, and the following parameter real-time data are collected: the number of sudden decelerations ρ and the number of sudden accelerations during the discharging process Driving speed level conditions Driving road conditions L, outside temperature The scope to which it belongs, form the discharging behavior habits, and transmit the collected data information to the cloud server; the specific content is as follows: (A) The driving recorder detects the sudden deceleration and sudden acceleration of the vehicle through the built-in acceleration sensor (G sensor); the acceleration sensor continuously collects the acceleration data of the vehicle, and a specific acceleration threshold of 3 m / s 2 is set, and a duration threshold of 2 s is set. When the positive acceleration threshold and the time threshold are exceeded, the event record of sudden acceleration is triggered. When the negative acceleration is less than the set value and the time threshold is exceeded, the event record of sudden deceleration is triggered. Determine the number of sudden brakes and sudden accelerations of this new energy commercial vehicle during a certain journey, and record the sudden acceleration habits and sudden deceleration habits in the discharge behavior of the new energy commercial vehicle; (B) The driving recorder monitors the current position of the new energy commercial vehicle in the form of longitude and latitude through the GPS module, and calculates the driving speed of the new energy commercial vehicle. The present invention sets two key parameters: speed threshold and duration threshold. If the monitored driving information is greater than the speed threshold of 40 km / h and the duration threshold of 1 min, it is determined to be driving at high speed. At the same time, for speed ranges in (40 km / h, 50 km / h], (50 km / h, 60 km / h], (60 km / h, 70 km / h], etc., and the duration exceeding the duration threshold, the high-speed driving levels are classified as Ⅰ, Ⅱ, Ⅲ... in sequence. Within the range not exceeding the threshold, the normal driving levels are classified according to the speed and duration thresholds. For speed ranges in (30 km / h, 40 km / h], (20 km / h, 30 km / h], (10 km / h, 20 km / h], (0 km / h, 10 km / h], and when the duration meets or exceeds the duration threshold, the normal driving levels are classified as 4, 3, 2, 1 in sequence, so as to judge the level of the current driving speed and record the driving speed habit in the discharge behavior of the new energy commercial vehicle; (C) Utilize vehicle dynamics sensors and road surface vibration sensors installed at the bottom of the vehicle to monitor the driving conditions of the vehicle. The driving conditions L consist of two aspects: road surface unevenness L1 and road surface gradient L2. In this invention, the international roughness index (IRI) is used as an index to evaluate the level of road surface unevenness. A calculation mathematical model for the road surface unevenness level is established based on the definition of IRI, and its correctness is verified through multiple sets of data. The vibration sensor can infer the load spectrum of the road section based on the vibration spectrum and amplitude generated when the vehicle passes through. After evaluating the road surface grade through the road section load spectrum, the corresponding road surface unevenness coefficient G is found. d (n0), On road surfaces of different grades, when new energy commercial vehicles drive at different speeds, the road surface will generate time-domain excitation signals for the vehicle. These signals change over time and reflect the dynamic response characteristics of the vehicle to road surface excitation. The time-domain excitation signal corresponding to various pavement grades and various vehicle speeds is the pavement excitation time-frequency power spectral density G d (n), which can be expressed by Equation (1): v b is the vehicle speed on this road surface, Ω is the time frequency for measuring the vehicle driving on this road surface, and n0 is the reference spatial frequency, usually 0.1 m -1 ; Calculate the current road surface unevenness IRI level value, and the formula is: Where q is the correction coefficient, which is usually 0.78 verified by multiple sets of actual vehicle data; The road surface unevenness level L1 is divided as shown in Table 1 below. Different IRIs correspond to different road surface unevenness levels Table 1 Road surface unevenness level correspondence table The estimation of the driving road surface gradient L2 of the present invention is established based on a least-squares estimation model of longitudinal dynamics. When a new energy commercial vehicle is driving on a slope, state variables at different times, such as speed, acceleration, steering angle, etc., and vehicle position information provided by GPS are obtained by vehicle dynamics sensors. The estimated quantities of the dynamic system of the system parameter w at the (i - 1)-th moment and the i-th moment are respectively The weights between the predicted value and the actual value of the gradient at the (i - 1)-th moment and the i-th moment are P(i - 1) and P(i) respectively. The Kalman gain used at the i-th moment is K(i), and the system output at the i-th moment is F if (i). The observable data vector at the i-th moment is μ(i). Then, the road surface gradient model based on the least-squares estimation of longitudinal kinematics is established as follows: Among them, the initial value P(0) = Z×E is determined, where Z is a sufficiently large positive real number, and E usually represents the prediction error covariance matrix, that is, the uncertainty of slope estimation; τ Z is the forgetting factor of the least squares estimation model of road surface slope based on longitudinal dynamics. The selection range of the forgetting factor is (0, 1]. Considering the variation characteristics of road surface slope, the influence of old observation data on new reference estimation is small. Therefore, the forgetting factor γ Z = 0.9 is selected, and the value of is calculated, then the estimation quantity of the dynamic system of system parameter w at the i-th moment is obtained; From w = sinα+(f0 + f1v p )cosα, the road surface slope can be deduced as where v p is the driving speed of the operating vehicle at the i-th moment, and f0 and f1 are the constant term and the first-order coefficient of the speed fitting of the rolling resistance coefficient; The present invention divides the road surface slope L2 level as follows: level 1 slope is 0%-1%, level 2 slope is 1%-5%, level 3 slope is 5%-10%, level 4 slope is 10%-15%, and level 5 slope is above 15%; Respectively evaluate the levels of the road surface unevenness L1 and the road surface slope L2 in the driving road condition information in the discharge behavior habit of the current new energy commercial vehicle, and record them; (D) The driving recorder records the external temperature of new energy operating vehicles during the discharging process. For the division of the external temperature range of the present invention, every 5°C is set as a threshold range. The temperature range of 0°C and above is divided into i intervals, which are respectively marked as r1, r2... r For the division of the range, every 5°C is set as a threshold range. The temperature range of 0°C and above is divided into i intervals, which are respectively marked as r1, r2... r i , and the temperature range below 0°C is also divided into i intervals, which are respectively marked as -r1, -r2... -r i , and the external temperature when the current new energy operating vehicle is driving is classified into the corresponding interval for recording; Based on monitoring the above-mentioned emergency braking times ρ, emergency acceleration times Driving speed level conditions The unevenness L1 and road surface slope L2 level conditions of the driving road condition L, as well as the external temperature The range of ownership, and obtain the discharge behavior habits of new energy commercial vehicles; (E) Taking an example recorded in Table 2, the habit of a certain new energy commercial vehicle in a certain discharge behavior in Area A was monitored; Table 2 Monitoring table of the discharge behavior habit of a certain new energy commercial vehicle in Area A This example monitoring method is also applicable to monitoring the discharge behavior habits of all new energy commercial vehicles in this area and other areas.

3. The optimization method for the battery charge and discharge strategy of a new energy operating vehicle according to claim 2, characterized in that: In step S1, the battery management system (BMS) is used to monitor the charging behavior habits of new energy commercial vehicles, including: estimating the state of charge (SOC) of the battery during charging, the charging method QS, and the temperature during charging by monitoring the voltage, current, temperature, and charging time parameters during charging Charging duration These four charging habit indicators are used to transmit the charging habit data information into the cloud server. The specific content is as follows: (a) Monitor the voltage, current, and temperature parameters of the battery during charging through the BMS, and use the improved Kalman filter algorithm to estimate the state of charge SOC of the battery. Compared with the traditional Kalman filter algorithm, this improved Kalman filter algorithm can improve the calculation accuracy of the state of charge SOC of the battery; Using the monitored battery output voltage U C , the battery operating current I, the target voltage U of the battery m and the battery operating duration T h , calculate I×T h to obtain the maximum available capacity C of the battery max , and further calculate the value of SOC. The calculation formula for the state of charge SOC of the rechargeable battery is: Among them, SOC(t1) in the formula represents the real-time SOC value, SOC(t0) represents the initial battery charge, C max represents the maximum available capacity, η represents the battery Coulomb efficiency, and I(t) represents the function of current and time; Divide the battery SOC state [0%, 100%] into 10 equal SOC threshold ranges. [0%, 10%] is the first range, [10%, 20%] is the second range, and so on to [90%, 100%] as the tenth range. Estimate in which range the SOC state of the new energy commercial vehicle is when it starts charging, and record the corresponding data information; (b) Monitor the real-time voltage and real-time current data during charging through the BMS, and calculate the average power P during charging avg , and the formula is: Among them, indicates the charging duration; Set specific power thresholds, where the power threshold for fast charging is 150KW and the power threshold for slow charging is 50KW; when a new energy commercial vehicle is charging, according to the calculated average power P avg Classify the charging method QS; if the average power is greater than the set fast charging power threshold, this charging is classified as fast charging; if the average power is less than the slow charging power threshold, this charging is classified as slow charging, and record the charging data information for this time; if the average power is within the range of [50KW, 150KW], this charging is classified as normal charging, abbreviated as general charging; (c) Monitor the external temperature during charging through the temperature sensor in the BMS. For the division of the temperature range to which the present invention pertains, set every 5°C as a threshold range, divide the temperature range of zero degree and above into i intervals, and label them as g1, g2... g respectively. Divide the temperature range below zero degree into i intervals as well, and label them as -g1, -g2... -g i respectively. Classify the temperature range to which the current charging temperature of new energy operation vehicles belongs into the corresponding intervals for recording; i ​ (d) The BMS records the start time of charging and monitors the change of current to determine the end time of charging, and then calculates the charging duration. Set a threshold range of 0.5h for each charging duration, divide the charging duration into i intervals, and label them as Q1, Q2... Q i , and classify the current charging duration of new energy commercial vehicles into the corresponding intervals for recording; Based on monitoring the SOC state, charging method QS, and temperature during charging as described above the range situation, and charging duration the range situation, to obtain the charging habits of new energy commercial vehicles; (e) Taking an example recorded in Table 3, the habit of a certain new energy commercial vehicle in a certain charging behavior in Area A was monitored Monitoring Table of a New Energy Operating Vehicle's Charging Behavior Habits in Area A in Table 3 This instance monitoring method is also applicable to monitoring the charging behavior habits of all new energy operating vehicle drivers in this area and other areas.

4. The optimization method for the battery charging and discharging strategy of a new energy operation vehicle according to claim 2, characterized in that: In step S2, a K-means evaluation model is established to evaluate the battery health level of new energy operating vehicles after one operating cycle. The implementation principle process is as follows: (1) Input the current battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charge efficiency C R , discharge efficiency F R into the K-means algorithm for clustering analysis, and divide them into K equally spaced clusters, so that each cluster has an initial center point; (2) Allocate the input data to each cluster until all the input data is allocated to the new clusters, and allocate the total number of samples S in the input data set to K clusters. This is achieved by continuously adjusting the centroid positions of the clusters by K-means to minimize the sum of the distances from the points to the centroids, that is, to minimize the squared error J. For the minimized squared error J of the cluster division γ obtained by clustering; where γ = {γ1, γ2, …, γ k} represents K clusters of battery health levels, is the current battery SOH, cycle life ε, capacity attenuation Q C , internal resistance R v , open circuit voltage U K , charge efficiency C R , discharge efficiency F R feature matrix, represents the cluster centroid of the i-th cluster; (3) Randomly select K cluster centroids of the battery health level characteristic samples, measure the distances between the samples and each centroid, and assign each sample to the nearest cluster centroid. Iterate n times in this way. In each iteration process, update the centroids of each type of cluster using the mean value. Repeat the above steps for the K cluster centroids until the cluster centroids are stable or the function converges; Based on the above steps, the battery health assessment model can be divided into 3 battery health level grades: m1, m2, m3. Among them, m3 is the optimal grade, m2 is the medium grade, and m1 is the worst grade; use this to evaluate the battery health level of a new energy operating vehicle in a certain area after one operating cycle; since the level of battery health reflects the quality of charging and discharging behavior habits, the battery health level of the optimal grade m3 corresponds to the charging and discharging behavior habits of the optimal grade M3, and so on for M1 and M2; The present invention only extracts the data of the charging and discharging behavior habits of the optimal grade M3 recorded by the batteries whose battery health levels are evaluated as the optimal grade m3, and those rated as M1 and M2 grades will not be used as references.

5. The optimization method for the battery charge and discharge strategy of a new energy operation vehicle according to claim 4, characterized in that: In step S3, in order to optimize the charging and discharging behavior habits of a new energy operating vehicle in Area A, Area A is determined as the target area, and the SVM model is used to screen areas with geographical characteristics similar to those of Area A and select the charging and discharging habits of the optimal grade M3 from them; In the data screening process based on the SVM model algorithm, geographical characteristics such as temperature, latitude, longitude, and altitude are used as key parameters for training the model; the SVM model classifies data with similar attributes through these characteristics; the steps for screening data using the SVM model are specifically as follows: Step 1) Establish a database divided according to geographical location, which contains the temperature, latitude, longitude, and altitude characteristic signals of each area, extract the data of these characteristics from it, and collect the sample data of each area with temperature, latitude, longitude, and altitude information; each sample should include these characteristic values and a target label for indicating the category to which the sample belongs; Step 2) Preprocess the data, including handling missing values, normalizing features, etc. The goal of this process is to ensure that the data set can be smoothly used for training the model; subsequently, the data set is divided into a training set and a test set; Step 3) Train the model using the support vector machine algorithm. During the training process, the algorithm will learn how to map the input features to the corresponding target labels; the support vector machine model performs type recognition on the samples, and it presets a kernel function θ; Construct a cost function using regression model samples Its expression is Apply the constraint conditions: δ i ≥ 0, i = 1, 2, 3, 4… Among them, ω z represents the classification interface vector; represents the transpose of the classification interface vector matrix; δ i represents the slack variable of the i-th sample; X i represents the feature vector of the i-th sample in the sample set X; U i the class label of the i-th sample; b represents the intercept; represents the penalty factor; represents the slack variable of the i-th sample containing the kernel function θ Construct the Lagrangian function. The expression for determining the category to which the sample data in Area A belongs is Among them, W z represents the feature weight vector, α i and ζ i represent Lagrange multipliers; Use the trained model to classify regions, input the temperature, latitude, longitude, and altitude characteristics of the regions, and determine the category to which they belong based on the symbol. If the result is positive, it belongs to one category; if the result is negative, it belongs to another category. The model will output the predicted label. According to the predicted label, find Area B and Area D that are geographically similar to the geographical features of Area A, and further extract the charge and discharge behavior habit data of the optimal level M3 in these areas.

6. The optimization method for the battery charging and discharging strategy of a new energy operation vehicle according to claim 5, characterized in that: In step S4, in order to optimize the charge and discharge behavior habits of a new energy operating vehicle in Area A, using the charge and discharge habits of the optimal level M3 in Area B and Area D that are similar to Area A as the scoring criteria, score and evaluate the charging habit and discharging habit of this new energy operating vehicle respectively. The obtained score is the first score of the charge and discharge behavior habits of the new energy operating vehicle; The specific steps for scoring the first score of the discharging behavior habit and the first score of the charging behavior habit of the new energy operating vehicle are as follows: ① Extract the discharging behavior habit characteristics and charging habit characteristics from the charge and discharge behavior habit characteristic sample data of the optimal level M3 in Area B and Area D, and ensure that the charge and discharge habits correspond one by one to the characteristic sample data; The discharge behavior characteristics of the optimal level M3 include: the number of sharp decelerations ρ, the number of sharp accelerations The situation of the driving speed level The level situations of the road surface unevenness L1 and the road surface slope L2 in the driving road condition L, and the outside temperature The situation of the belonging range; The charging behavior habit characteristics of the optimal level M3 include: SOC status, charging method QS, temperature during charging Range situation, charging duration Range situation; ② According to each group of charging and each group of discharging behavior habit characteristics in the charge and discharge behavior habit sample data of the optimal level M3 in Area B and Area D, determine the threshold range of each charge and discharge behavior habit characteristic that meets the scoring qualification: Sort each group of discharging habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range; sort each group of charging habit characteristics, set the highest value after sorting as the upper limit of the threshold range, and the lowest value after sorting as the lower limit of the threshold range; Taking the battery SOC charge state of the charging habit characteristic as an example, sort it. The highest value is 9 and the lowest value is 3. Set the threshold range [3, 9] that meets the scoring qualification. When evaluating the SOC state of the charging habit, if it meets this threshold condition, this charging habit enters the specific score evaluation link. If it does not meet this threshold condition, this charging behavior habit has no scoring qualification and is judged as 0 points; ③ According to the areas selected that are geographically similar to the area where the target new energy operating vehicle is located, calibrate the weight coefficients of the first scores of each discharging behavior habit and each charging behavior habit based on the overlap degrees of the optimal level M3's discharging behavior habits and charging behavior habits in these areas; ④ Based on the preset threshold conditions and the weight coefficients of the first scores of each charge and discharge behavior habit, evaluate the first score of the charging behavior habit and the first score of the discharging behavior habit of the target new energy operating vehicle.

7. The optimization method for the battery charging and discharging strategy of a new energy operation vehicle according to claim 6, characterized in that: Step ③ is specifically: (1) First, establish each charge and discharge behavior habit characteristic data finally screened out, such as the SOC state at the start of charging in the charging habit, as a vector feature; assume that these characteristic data are represented by h = [h1, h2, h3…h n , where n represents the number of characteristic data. For each characteristic data, count the frequency p i with which it appears in the vector h, and then calculate the overlap degree of each characteristic data through the following formula Its p i is the frequency of occurrence of the characteristic data h i in the vector h, from which the overlap degree of each charge and discharge behavior habit characteristic data can be calculated; (2) The overlap degrees of various discharging behavior habits and the overlap degrees of various charging behavior habits calibrate the weight coefficients of the first scores of various charging and discharging habits. Select the highest overlap degree of each habit characteristic data, sort them in descending order, and assign the corresponding first scores of various discharging behavior habits to the weight coefficients W1, W2, W3, W4, W5, W6, W in descending order. i The discharging behavior habit characteristics corresponding to the weight coefficients are non-repetitive, i ∈ [1, 6], and is an integer; the corresponding first scores of various charging behavior habits are assigned to the weight coefficients ω1, ω2, ω3, ω4 in descending order, and ω I The charging behavior habit characteristics corresponding to the weight coefficients are non-repetitive, I ∈ [1, 4], and is an integer; (3) Based on the overlap degree among each charge and discharge behavior habit characteristic data in step (1), sort the overlap degree of one discharge behavior habit data from high to low, and assign the weight coefficients from large to small to the evaluation scores corresponding to the range, level or frequency of this discharge behavior habit characteristic data A i1 The ranges, levels or frequencies of the characteristic data corresponding to the weight coefficients do not repeat each other, where i1 ∈ [1, n1] and is an integer. The same method is applied to the other five discharge habits, and the evaluation scores of the ranges, levels or frequencies are assigned weight coefficients B i2 、C i3 、D i4 、E i5 、F i6 The ranges, levels or frequencies of the characteristic data corresponding to the weight coefficients do not repeat each other, where i2, i3, i4, i5, i6 belong to [1, n2], [1, n3], [1, n4], [1, n5], [1, n6] respectively and are integers; in the same way, sort the overlap degree of one charge behavior habit data from high to low, and assign the weight coefficients from large to small to the evaluation scores corresponding to the range, level or frequency of this charge behavior habit characteristic data a i7 The ranges, levels or frequencies of the characteristic data corresponding to the weight coefficients do not repeat each other, where i7 belongs to [1, n7] and is an integer. The same method is applied to the other three charge habits, and the evaluation scores of the ranges, levels or frequencies are assigned weight coefficients b i8 、c i9 、d i10 The ranges, levels or frequencies of the characteristic data corresponding to the weight coefficients do not repeat each other, where i8, i9, i10 belong to [1, n8], [1, n9], [1, n 10 , and are integers.

8. The optimization method for the battery charging and discharging strategy of a new energy operating vehicle according to claim 6, characterized in that: Step ④ is specifically: (1) Clean and prepare the multi-group characteristic data of the charge and discharge behavior habits of the target new energy operating vehicle to ensure the accuracy of the data and enter the scoring link; (2) Compare and analyze each charge and discharge behavior habit characteristic with the preset threshold range that meets the scoring qualification. If this habit characteristic meets the threshold range, enter the specific score evaluation link. The scoring range is (0, 100]. Multiply the full score of 100 by the weight coefficient A corresponding to the level or frequency or range to which the data of this habit characteristic belongs i1 / a i7 , or B i2 / b i8 etc. The result is the specific score of this habit data. If this habit characteristic does not meet the threshold range, it is determined to be 0 points; (3) Multiply the specific score of this habit data by the weight coefficient W / ω corresponding to the first score of this habit, thereby obtaining the first score of this discharging behavior habit or charging behavior habit; (4) By analogy with this method, the first score of each discharging behavior habit or charging behavior habit is obtained, and by performing a separate summation operation on them, the first scores of the discharging behavior habit and charging behavior habit of the target new energy operating vehicle can be obtained; In this example, there are 6 groups of characteristic data of the discharging behavior habit of a new energy operating vehicle in area A, and the total score range is [0, 100]. The weighted summation formula is used to calculate the first total score of the discharging habit RS f = Y1W1A i1 + Y2W2B i2 + Y3W3C i3 + Y4W4D i4 + Y5W5E i5 + Y6W6F i6 (11) Y N The score given to the Nth group of characteristic data of the discharge behavior habits of the target new energy operating vehicle according to the preset threshold conditions. If it meets the threshold conditions, the score is 100; if it does not meet the threshold conditions, the score is 0. N ∈ [1, 6] and is an integer; There are 4 groups of characteristic data of the charging behavior habit of a new energy operating vehicle in area A, and the total score range is [0, 100]. The weighted summation formula is used to calculate the first total score of the charging habit RS c = β1ω1a i7 + β2ω2b i8 + β3ω3c i9 + β4ω4d i10 (12) For the charging behavior habits of drivers of target new energy operation vehicles, the score of the nth group of characteristic data is based on a preset threshold condition. If it meets the threshold condition, the score is 100; if it does not meet the threshold condition, the score is 0, and it is an integer; From this, the first scores of the charging behavior habit and discharging behavior habit of a driver of a new energy operating vehicle in area A can be calculated; From the above steps, it can be obtained that: The first score of the discharging behavior habit of new energy commercial vehicles: RS f The first score of the charging behavior habits of new energy operating vehicles: RS c .

9. The optimization method for the battery charging and discharging strategy of a new energy operation vehicle according to claim 6, characterized in that: In step ④, the self-charging and discharging habits of the target new energy operating vehicle are used as the scoring criteria, and the second scores of the discharging behavior habit and the charging behavior habit are obtained as a supplementary comparison to the first scores; The present invention not only has the first scores obtained based on the charging and discharging habits of the optimal level in similar areas as described above, but also sets a control score, that is, the second score of the charging and discharging habits. The scoring criteria for this control score are based on the second score scoring rule table established based on the historical charging and discharging behavior habits of the target new energy operating vehicle, and the calculated score is used as the second score; The collection and utilization of the data in its rule scoring table no longer focus on the charging and discharging habits of the optimal level in similar areas, but are oriented to the charging and discharging habits of the target new energy operating vehicle itself; the present invention does not rely solely on one scoring result, but combines the comprehensive judgments of two scoring modes to obtain a more accurate and reliable scoring result by complementing each other's strengths and weaknesses; In the example, the second score of the charging and discharging behavior habits of a new energy operating vehicle in Area A is obtained through the number of sharp decelerations ρ and sharp accelerations during the large-scale discharging process of the target new energy operating vehicle itself Driving speed level Driving road condition L, outside temperature Scope of belonging, SOC state during charging, charging method QS, temperature during charging Scope of belonging, charging duration Historical data of the scope of belonging have different degrees of impact on battery health. A rule table is established. The discharging behavior habits and charging behavior habits of a new energy operating vehicle in Area A are input. The scores obtained for each habit are given according to the corresponding scoring rules, and the scores are added up to obtain respectively: The second score of the discharge behavior habit of new energy operating vehicles: rs f The second score of the charging behavior habit of new energy operation vehicles: rs c .

10. The optimization method for the battery charging and discharging strategy of a new energy operation vehicle according to claim 9, characterized in that: Step S5 is specifically to select the lower score as the optimization basis to generate the battery charging and discharging optimization strategy for the new energy operating vehicle, and the steps are as follows; (1) Select the scoring criteria corresponding to the lower score as the optimization basis; utilize the first score RS of the charging and discharging behavior habits of new energy operating vehicles f , RS c and the second score rs f , rs c , conduct a score comparison, and use the scoring criteria corresponding to the lower score as the optimization basis for the charging and discharging strategy; in general, since the first score of the charging and discharging behavior habits of new energy operating vehicles is based on the charging and discharging habits of the optimal level in similar regions as the scoring criteria, the requirements are more stringent and targeted, so the scores are generally lower. However, it cannot be excluded that when the second score is lower, instead of learning the charging and discharging habits of the optimal level in similar regions, the special case of learning the better charging and discharging habits of the target new energy operating vehicle itself is considered, then the scoring criteria corresponding to the second score is added as the optimization basis; When RS f ≤ rs f , select the scoring criteria of the first score of the discharge behavior habit of new energy operation vehicles as the optimization basis; When RS f >rs f , select the scoring criteria of the second score of the discharge behavior habit of new energy operating vehicles as the optimization basis; When RS c ≤ rs c , the scoring criteria for the first score of the charging behavior habits of new energy operation vehicles are used as the optimization basis; When RS c >rs c , then select the scoring criteria of the second score of the charging behavior habit of new energy operation vehicles as the optimization basis; (2) Generate the battery charging and discharging optimization strategy for the new energy operating vehicle; in the example, through information connection between the cloud server and intelligent devices such as smartphones, computers, and tablets, the cloud server obtains the first scores and second scores of the charging and discharging behavior habits of a new energy operating vehicle in area A. Select the scoring criteria corresponding to the lower score as the basis for optimizing the charging and discharging behavior habit scores and charging and discharging strategies this time. Through the generation module, generate the score report form of the charging and discharging behavior habit and the charging and discharging optimization strategy report form this time, and send them to the driver's intelligent device through the transmission module; The composition of the score report form of the charging and discharging behavior habit: The charging behavior habit score and discharging behavior habit score this time, and the specific score of each charging and discharging behavior habit; The composition of the charging and discharging optimization strategy report form: When the battery of the new energy operating vehicle starts charging, the optimal SOC state; When the battery of the new energy operating vehicle is charging, choose fast charging, slow charging or normal charging; When the battery of a new energy operating vehicle is charging, the optimal temperature during charging Range; Optimal charging duration when the battery of a new energy operation vehicle is charging Scope of belonging; When the battery of the new energy operating vehicle is discharging, the optimal number of sharp decelerations ρ for an average of 5 km journey; Optimal number of sharp accelerations for an average 5-km journey during battery discharge of new energy commercial vehicles When the battery of a new energy operating vehicle discharges, the best driving speed level for an average 5-km journey When the battery of a new energy commercial vehicle discharges, the average roughness level L1 of the road surface for every 5 km and the road surface gradient level L2; when the battery of a new energy commercial vehicle discharges, the optimal ambient temperature The applicable range.

Images