Electric vehicle battery output power optimization method and device and storage medium

Abstract

The invention provides an electric vehicle battery output power optimization method and device and a storage medium, and the method comprises the steps: calculating an average output power value of a battery under a historical road condition similar to a current position road condition feature vector according to the extracted battery output power data, determining the optimal battery output power corresponding to the current position road condition feature vector in combination with the battery charge state, the battery temperature, the battery internal resistance and the battery capacity attenuation parameter; and controlling a battery management system to adjust the actual battery output power of the target automobile, enabling the actual battery output power to gradually approach to the optimal battery output power, and performing battery temperature control, charging and discharging power adjustment, battery equalization management and battery over-charging and over-discharging protection in real time in the adjustment process, thereby dynamically optimizing the battery efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of power batteries, and particularly to a method, device, and storage medium for optimizing the output power of an electric vehicle battery. Background Art

[0002] During the actual driving process of an electric vehicle, it faces a complex and changeable road condition environment, which poses a huge challenge to the battery management system. How to accurately control the battery output power under different road conditions, which can not only meet the driving requirements but also maximize the battery efficiency, has become a technical problem to be solved urgently. Summary of the Invention

[0003] The present invention provides a method for optimizing the output power of an electric vehicle battery, mainly including:

[0004] Obtain a list of vehicles with similar models to the target vehicle, and use the on-vehicle system to collect historical driving data from the similar vehicles. The historical driving data includes driving routes, vehicle speeds, battery output powers, road types, road surface conditions, traffic flows, weather conditions, accelerations, steering angles, and gradient change information, and construct a historical driving record library of similar vehicles based on the historical driving data;

[0005] Obtain the real-time position coordinates of the target vehicle through the on-vehicle positioning system, obtain the surrounding road information from the map database according to the real-time position coordinates, and use the on-vehicle sensors to detect the real-time road condition information of the target vehicle. The real-time road condition information includes road surface conditions, weather conditions, vehicle speed change trends, accelerations, steering angles, and gradient changes, and generate a current position road condition feature vector based on the surrounding road information and the real-time road condition information;

[0006] Match the current position road condition feature vector with the historical road conditions in the historical driving record library of similar vehicles, obtain the historical driving record with the highest matching degree with the current position road condition feature vector, and extract battery output power data, state of charge of the battery, battery temperature, charge and discharge power, battery internal resistance, and battery capacity attenuation parameters from the historical driving record with the highest matching degree;

[0007] According to the extracted battery output power data, calculate the average output power value of the battery under the historical road conditions similar to the current position road condition feature vector, and combine the state of charge of the battery, battery temperature, battery internal resistance, and battery capacity attenuation parameters to determine the optimal battery output power corresponding to the current position road condition feature vector;

[0008] Control the battery management system to adjust the actual battery output power of the target vehicle, so that the actual battery output power gradually approaches the optimal battery output power. During the adjustment process, perform real-time battery temperature control, charge and discharge power regulation, battery balancing management, and battery overcharge and over-discharge protection, thereby dynamically optimizing battery efficiency;

[0009] During the driving process of the target vehicle, continuously collect data on actual battery output power, driving route, vehicle speed, road type, road surface condition, traffic flow, weather condition, acceleration, steering angle, and slope change, and update the collected data to the similar vehicle driving record library to continuously improve the historical data set of the similar vehicle driving record library, so as to improve the accuracy of subsequent battery output power optimization.

[0010] The technical solution provided by the embodiment of the present invention may include the following beneficial effects:

[0011] The present invention discloses a method for optimizing the battery output power of an electric vehicle. The method constructs a driving record library by collecting historical driving data of similar vehicle models, obtains the location and road condition information of the target vehicle in real time, matches the current road condition with the historical records, extracts the corresponding battery parameters, and calculates the optimal output power. The present invention dynamically adjusts the battery output power to gradually approach the optimal value, and at the same time performs protection measures such as temperature control, charge and discharge regulation, etc. During the driving process, continuously collect actual data and update the record library to continuously improve the historical data set. The present invention realizes the intelligent optimization of the battery output power by matching similar historical road conditions and combining various battery parameters, improves the battery usage efficiency, and extends the battery life. At the same time, through real-time data collection and update, the optimization accuracy is continuously improved, forming a self-improving dynamic optimization system. Brief Description of the Drawings

[0012] Figure 1 It is a flowchart of a method for optimizing the battery output power of an electric vehicle according to the present invention.

[0013] Figure 2 It is a schematic block diagram of a computer device provided by an embodiment of the present invention. Detailed Embodiments

[0014] To further understand the content of the present invention, the present invention will be described in detail in combination with the drawings and embodiments. The following further describes the present application in detail with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related invention, rather than limiting the invention. In addition, it should be noted that, for the sake of description, only the parts related to the invention are shown in the drawings.

[0015] As Figure 1-2 , a method for optimizing the battery output power of an electric vehicle in this embodiment may specifically include:

[0016] Step S101, obtain the historical driving data of similar vehicles according to the model of the target vehicle. The historical driving data includes driving route, vehicle speed, battery output power, road type, road surface condition, traffic flow, weather condition, acceleration, steering angle, and gradient change information.

[0017] According to the model parameters of the target vehicle, retrieve a list of similar vehicle models from the vehicle model database to obtain a set of similar vehicle models; for the set of similar vehicle models, obtain the in-vehicle system interface information of each model vehicle, and establish a mapping relationship data table between the vehicle model and the in-vehicle system; through the in-vehicle system interface, collect the historical driving data of each similar vehicle, including driving route, vehicle speed, battery output power, road type, road surface condition, traffic flow, weather condition, acceleration, steering angle, gradient change and other information; preprocess the collected historical driving data, clean outliers and invalid data, and standardize the data to convert it into a unified data format and unit; according to the driving route information, segment the vehicle speed, battery output power, acceleration, steering angle, and gradient change data of each section of the route, and extract section features; associate the environmental information such as road type, road surface condition, traffic flow, and weather condition with the section features to construct an association matrix between the section features and environmental factors; based on the association matrix, use a clustering algorithm to perform clustering analysis on the driving data of different sections to obtain a set of sections with similar driving characteristics, and construct a similar vehicle driving record library.

[0018] Specifically, in order to mine valuable information from the historical driving data of similar vehicles, the following methods can be adopted: First, according to the model parameters of the target vehicle, such as vehicle length, vehicle width, wheelbase, etc., use the cosine similarity algorithm to retrieve vehicle models with a similarity exceeding 8 from the vehicle model database to form a set of similar vehicle models. Then, through the API interface provided by the vehicle enterprise, obtain the in-vehicle system interface information of each vehicle model, such as interface type, protocol, data format, etc., and establish a mapping relationship data table between the vehicle model and the in-vehicle system. Next, connect to the in-vehicle systems of each similar vehicle through the MQTT protocol, collect its historical driving data at a frequency of 1 Hz, and the data fields include GPS trajectory, vehicle speed, battery output power, road type, road surface condition, traffic flow, weather condition, acceleration, steering angle, and slope change, etc., and store them in the MongoDB database. Clean the collected data, remove abnormal data such as vehicle speed exceeding 220 km / h and negative battery output power, and normalize the data into a unified JSON format with the dimension unified to the International System of Units. According to the GPS trajectory information, segment each route at intervals of 100 m, and extract features such as average vehicle speed, average battery output power, average acceleration, average steering angle, and average slope change of each section. Use the logistic regression algorithm to associate environmental information such as road type, road surface condition, traffic flow, and weather condition with the section features to construct a 15×8-dimensional association matrix. Finally, use the K-means clustering algorithm to perform clustering analysis on the association matrix, stop the iteration when the silhouette coefficient is greater than 7, obtain 12 sets of sections with similar driving characteristics, store them in the PostgreSQL database, and construct a similar vehicle driving record library to provide data support for subsequent applications such as driving condition analysis and energy consumption prediction.

[0019] Step S102: Obtain the real-time position coordinates of the target vehicle through the vehicle-mounted positioning system, obtain the surrounding road information from the map database according to the real-time position coordinates, detect the real-time road condition information of the target vehicle using the vehicle-mounted sensors, and generate a current position road condition feature vector based on the surrounding road information and the real-time road condition information. The real-time road condition information includes road surface condition, weather condition, vehicle speed change trend, acceleration, steering angle, and slope change.

[0020] Obtain the real-time position coordinates of the target vehicle through the vehicle-mounted positioning system, and use the real-time position coordinates as the query condition to obtain the surrounding road information of this location from the pre-constructed map database. Use vehicle-mounted sensors to detect the driving state of the target vehicle in real time, and obtain real-time road condition information including road surface conditions, weather conditions, vehicle speed change trends, acceleration values, steering angles, and slope changes, etc. According to the obtained surrounding road information and real-time road condition information, construct a road condition feature vector centered on the real-time position coordinates. Use a convolutional neural network to perform feature extraction and representation learning on the road condition feature vector to obtain a compressed road condition feature vector. Input the compressed road condition feature vector into a long short-term memory network, and realize the temporal modeling of the road condition features through the gated recurrent unit to capture the dynamic change trend of the road condition features. At the output layer of the long short-term memory network, map the learned road condition features to the energy consumption prediction value of the target vehicle through a fully connected layer. If the predicted energy consumption value exceeds the preset safety threshold, send a warning message to the target vehicle to prompt the driver to adjust the driving strategy to reduce the energy consumption risk. At the same time, match the predicted energy consumption value with the remaining battery power to determine the location of the nearest charging station and guide the target vehicle to reach safely.

[0021] Specifically, the real-time position coordinates of the target vehicle are obtained through an in-vehicle positioning system. Using GPS / Beidou dual-mode positioning technology, the positioning accuracy can reach within 1 meter. Taking the real-time position coordinates as the query condition, and using a spatial index algorithm such as the R-tree, etc., the surrounding road information such as the road topology structure and road attributes within a radius of 1 kilometer around this position can be quickly obtained from the pre-constructed map database PostgreSQL / PostGIS. The in-vehicle sensors collect vehicle driving state data in real time at a frequency of 100Hz, including detecting road surface conditions such as water accumulation depth and road surface friction coefficient using lidar, detecting weather conditions such as temperature, humidity, and visibility using an in-vehicle weather station, calculating the vehicle speed change trend using a wheel speed sensor, obtaining the triaxial acceleration value using an accelerometer, obtaining the steering angle using a steering angle sensor, and obtaining the slope change using an inclination sensor, etc. According to the obtained surrounding road information and real-time road condition information, each road unit is vectorized and represented, and a 128-dimensional road condition feature vector centered on the real-time position coordinates is constructed. A convolutional neural network composed of 3 convolutional layers and 2 pooling layers is used to perform feature extraction and representation learning on the road condition feature vector. Each convolutional layer contains 64 3*3 convolutional kernels, the activation function uses ReLU, and the pooling layer uses 2*2 max pooling, which can compress the road condition feature vector to 32 dimensions. The compressed road condition feature vector is input into an LSTM layer containing 128 neurons, and the temporal modeling of the road condition features is realized through 3 gated recurrent units. With a time step of 10 seconds, the dynamic change trend of the road condition features is captured. At the output layer of the long short-term memory network, the learned road condition features are mapped to the energy consumption prediction value of the target vehicle through a fully connected layer containing 64 neurons and a Sigmoid activation function. If the predicted energy consumption value exceeds the safety threshold of 20% of the battery capacity, an alarm prompt is sent to the target vehicle, and according to the real-time traffic flow information, using the Dijkstra shortest path algorithm, a low-energy driving route to the nearest charging station is planned on the in-vehicle navigation map to guide the target vehicle to reach safely.

[0022] Step S103, determine the historical driving record with the highest matching degree according to the current position road condition feature vector and the historical driving data, and the historical driving record with the highest matching degree includes battery output power data, state of charge of the battery, battery temperature, charge and discharge power, battery internal resistance, and battery capacity attenuation parameters.

[0023] An eigenvector is an important concept in linear algebra and is closely related to an eigenvalue. Given a square matrix A, if there exists a non-zero vector v and a scalar λ such that Av = λv, then v is an eigenvector of matrix A, and λ is the corresponding eigenvalue. Obtain the road condition eigenvector at the current position. The road condition eigenvector includes the road directions that a computer can recognize and information on road changes. Match it with a pre-established historical road condition feature library, calculate the similarity between each historical road condition eigenvector and the current road condition eigenvector, and select the historical driving record corresponding to the historical road condition eigenvector with the highest similarity; extract parameters related to battery performance from the historical driving record with the highest matching degree, including battery output power data, state of charge (SOC) of the battery, battery temperature, charge and discharge power, battery internal resistance, and battery capacity attenuation parameters; according to the battery temperature and battery internal resistance, use the Kalman filter algorithm to filter the battery output power data to obtain a filtered battery output power curve; according to the filtered battery output power curve and the SOC of the battery, use the least squares fitting algorithm to obtain a functional relationship between the battery output power and the SOC; substitute the charge and discharge power and the battery capacity attenuation parameters into the above functional relationship to obtain the estimated available mileage of the battery under the current road conditions; if the estimated available mileage is less than a preset threshold, it is determined that the current battery power is not sufficient to support the vehicle to drive to the destination, and a low battery warning is issued to the driver; if the estimated available mileage is greater than or equal to the preset threshold, it is determined that the current battery power can support the vehicle to drive to the destination, and no warning is issued.

[0024] Specifically, first, traffic condition feature data of the current vehicle position is collected through in-vehicle sensors, including road type, road surface condition, traffic flow, etc., to form a 20-dimensional feature vector. Then, this feature vector is matched with a feature library pre-established with 1000 historical traffic condition feature vectors. The Euclidean distance is used to calculate the similarity, and the historical driving record corresponding to the historical traffic condition feature vector with the highest similarity (the closest distance) is selected. Next, parameters related to battery performance are extracted from the matched historical driving record, including battery output power data (sampling frequency is 1Hz, duration is 1 hour), state of charge SOC of the battery (accuracy is 1%), battery temperature (accuracy is 1°C), charge and discharge power (accuracy is 1kW), battery internal resistance (accuracy is 1mΩ), and battery capacity attenuation parameter (accuracy is 1%). According to the battery temperature and internal resistance, the Kalman filter algorithm is used to filter the battery output power data. The initial value of the estimated error covariance matrix is set to 100, the process noise covariance matrix is set to 10, and the measurement noise covariance matrix is set to 5. After 20 iterations, the filtered power curve is obtained. Then, based on this curve and SOC, the least squares method is used to fit a cubic polynomial function (the goodness of fit RR is greater than 9) to obtain the functional relationship P = f(SOC) between power P and SOC. Substitute the current charge and discharge power of the battery (such as 3kW) and the capacity attenuation parameter (such as 5%) into this function to calculate the estimated available mileage of the battery under the current traffic conditions (such as 80km). Finally, if the estimated mileage is less than the preset threshold (such as 50km), it is determined that the current battery power is not sufficient to support the vehicle to reach the destination, and a low battery warning is issued to the driver; otherwise, it is determined that the battery power is sufficient and no warning is issued. Through the above method, the remaining mileage of the electric vehicle can be accurately estimated based on the real-time traffic conditions and battery status, effectively avoiding the situation of the vehicle breaking down halfway due to battery depletion and improving the user travel experience.

[0025] Step S104, determine the optimal battery output power according to the historical driving record with the highest matching degree and the current position traffic condition feature vector.

[0026] Obtain the road condition feature vector of the current location, and search for historical road condition data similar to the current road condition feature vector in the historical database; according to the historical road condition data, extract the corresponding battery output power data, and calculate the average output power value of the battery under this road condition; obtain the state parameters of the current battery, such as state of charge, temperature, internal resistance, and capacity attenuation; based on the battery state parameters, establish a battery output power optimization model, and optimize and calculate the optimal battery output power under the current road condition with the average output power value as the benchmark; use the optimal battery output power as a control command to adjust the actual output power of the battery to make it optimal under the current road condition; monitor the changes in the output power and state parameters of the battery in real time. If the deviation from the optimal value exceeds the preset threshold, re-execute the above steps to dynamically adjust the battery output power; store the optimized battery output power data and the corresponding road condition feature vector in the historical database, and continuously update and improve the optimization model.

[0027] Specifically, first, the road condition feature data of the current location of the vehicle is collected in real time through in-vehicle sensors, including road type, slope, curvature, traffic flow, etc., to form a 20-dimensional feature vector. Then, the K-nearest neighbor algorithm is used to retrieve 100 pieces of historical road condition data with the closest Euclidean distance to the current feature vector in the historical database. Next, the battery output power data corresponding to these 100 pieces of data is extracted, and its arithmetic mean is calculated as the average output power of the battery under this road condition, assumed to be 50 kW. At the same time, through the in-vehicle battery management system, the SOC of the current battery is obtained as 60%, the temperature is 30°C, the internal resistance is 5 mΩ, and the capacity attenuation is 10%. On this basis, a multiple linear regression model is established, with the battery state parameters as independent variables and the battery output power as the dependent variable, and optimized and solved with 50 kW as the target value to obtain the optimal output power of 45 kW under the current road condition. Subsequently, 45 kW is used as a control command to be sent to the battery management system to adjust the duty cycle of the DC / DC converter to make the actual output power of the battery reach the optimal value. During the driving of the vehicle, the battery output power and state parameters are sampled at a period of 100 ms. If it is found that the output power deviates from the optimal value

[0028] ±5 kW or the SOC changes by more than 5%, then re-optimization is triggered. Finally, the optimized output power data and the corresponding road condition feature vector are stored in the historical database according to the timestamp, and the optimization model is incrementally trained every day at midnight using the newly collected data to continuously improve the accuracy and robustness of the model.

[0029] Step S105, adjust the actual battery output power of the target vehicle to gradually approach the optimal battery output power.

[0030] Obtain the actual battery output power of the electric vehicle and the optimal battery output power under the current working conditions, calculate the difference between the two as the target value for battery output power adjustment. According to the real-time temperature data collected by the battery temperature sensor, judge whether the battery temperature exceeds the preset safe range. If it exceeds, start the battery temperature control strategy, and adjust the battery temperature to the safe range by controlling the battery cooling system or heating system. Obtain the real-time charge and discharge current and voltage data of the battery. According to the preset charge and discharge power curve, judge whether the current charge and discharge power exceeds the safe range. If it exceeds, start the charge and discharge power adjustment strategy, and control the battery management system to adjust the charge and discharge current to control the power within the safe range. Adopt an equalization management algorithm to obtain the voltage data of each battery cell, calculate the mean and variance of the cell voltages, and judge whether they exceed the preset equalization threshold. If they exceed, start the equalization control strategy, and adjust the cell voltages to the equalized state through the equalization circuit. Obtain the real-time voltage data of the battery, and judge whether it exceeds the preset overcharge voltage threshold. If it exceeds, start the overcharge protection strategy, and control the battery management system to stop charging to avoid safety hazards caused by overcharging the battery. Obtain the real-time voltage data of the battery, and judge whether it is lower than the preset over-discharge voltage threshold. If it is lower, start the over-discharge protection strategy, and control the battery management system to stop discharging to avoid performance degradation caused by over-discharging the battery. According to the battery output power adjustment target value, use the gradient descent algorithm to adjust the battery output power in real time to make it gradually approach the optimal value. At the same time, during the adjustment process, perform real-time battery temperature control, charge and discharge power adjustment, battery equalization management, and overcharge and over-discharge protection. Through the multi-objective optimization algorithm, dynamically balance various indicators, and finally realize the dynamic optimization of battery efficiency.

[0031] Specifically, first obtain the actual battery output power of the electric vehicle and the optimal battery output power under the current working conditions. For example, if the actual output power is 80 kW and the optimal output power is 100 kW, the calculated difference is 20 kW, which is used as the target value for battery output power adjustment. At the same time, based on the real-time temperature data collected by the battery temperature sensor, determine whether the battery temperature exceeds the preset safety range. For example, if the battery temperature is 45°C, which exceeds the preset safety range of 0°C to 40°C, start the battery temperature control strategy and adjust the battery temperature to 35°C by controlling the battery cooling system. Obtain the real-time charge and discharge current and voltage data of the battery, and judge whether the current charge and discharge power exceeds the safety range according to the preset charge and discharge power curve. For example, if the current charging power is 120 kW, which exceeds the preset maximum charging power of 100 kW, start the charge and discharge power adjustment strategy and control the battery management system to adjust the charging current from 200 A to 150 A, and control the charging power within 90 kW. Adopt the equalization management algorithm to obtain the voltage data of each battery cell, calculate the average value of the cell voltages to be 8 V and the variance to be 0.5, which exceeds the preset equalization threshold of 0.2, then start the equalization control strategy and adjust the cell voltages to the equalized state of 8 V ± 0.1 V through the equalization circuit. Obtain the real-time voltage data of the battery and judge whether it exceeds the preset overcharge voltage threshold of 2 V. If it exceeds, start the overcharge protection strategy and control the battery management system to stop charging. Obtain the real-time voltage data of the battery and judge whether it is lower than the preset over-discharge voltage threshold of 5 V. If it is lower, start the over-discharge protection strategy and control the battery management system to stop discharging. Finally, according to the battery output power adjustment target value of 20 kW, adopt the gradient descent algorithm to adjust the battery output power in real time. For example, the adjustment step size is 5 kW each time. After 4 iterative adjustments, the battery output power gradually approaches the optimal value of 100 kW from 80 kW. At the same time, during the adjustment process, real-time battery temperature control, charge and discharge power adjustment, battery equalization management, and overcharge and over-discharge protection are carried out. Through the multi-objective weighted summation algorithm, the weights of various indicators are dynamically balanced, and finally the dynamic optimization of battery efficiency is achieved.

[0032] In one embodiment, the method for optimizing the battery output power of the electric vehicle further includes: during the driving process of the target vehicle, continuously collect the actual battery output power, driving route, vehicle speed, road type, road surface condition, traffic flow, weather condition, acceleration, steering angle, and slope change data, and update the collected data to the similar vehicle driving record library.

[0033] Obtain the real-time driving data of the target vehicle, including information such as battery output power, driving route, vehicle speed, road conditions, traffic flow, weather conditions, acceleration, steering angle, and slope change. Transmit the obtained real-time driving data to the cloud server and store it in the driving record library of similar vehicles for merging with historical data. Preprocess the historical data and real-time data in the driving record library of similar vehicles, including operations such as data cleaning and data normalization, to obtain a standardized data set. Based on the standardized data set, use machine learning algorithms (such as support vector machines, random forests, etc.) to establish a battery output power prediction model, and use the cross-validation method to train and optimize the model. Deploy the trained battery output power prediction model to the cloud server to receive the driving data of the target vehicle in real time and predict the battery output power. According to the predicted battery output power, combined with information such as the battery status and remaining power of the target vehicle, use the dynamic programming algorithm to obtain an optimized battery output power scheduling strategy. Send the optimized battery output power scheduling strategy to the target vehicle to adjust the output power of the battery in real time to achieve the purpose of energy conservation and efficiency improvement and increase the driving range of electric vehicles.

[0034] Specifically, by installing sensors and in-vehicle terminal devices on the target vehicle, driving data such as battery output power, GPS positioning information, vehicle speed, acceleration, and steering angle can be collected in real time, and the data can be uploaded to the cloud server through a wireless communication module (such as 4G / 5G network). At the same time, real-time weather data of the area where the target vehicle is located is obtained from the meteorological department, and road conditions and traffic flow information are obtained from the traffic management department, and are stored in the similar vehicle driving record library in the cloud together with the vehicle driving data. Use big data processing platforms such as Hadoop to perform ETL processing on massive historical data and real-time data, remove outliers and missing values through data cleaning, and use the Min-Max normalization method to normalize the data, mapping the data to the interval [0,1] to eliminate the influence of different metric dimensions. Based on the standardized data set, use the Scikit-learn machine learning library in Python and adopt the support vector regression (SVR) algorithm to establish a battery output power prediction model. Select the radial basis kernel function (RBF), and optimize the model hyperparameters such as the penalty coefficient C and the kernel function parameter γ through grid search and 5-fold cross-validation to improve the prediction accuracy. Package the trained SVR model as a RESTful API interface and deploy it on the cloud server. When receiving the driving data uploaded in real time by the target vehicle, call the prediction interface and input the feature vector, and the predicted value of the battery output power can be returned within 50 milliseconds. According to the predicted battery output power, obtain the battery state parameters (such as battery temperature, voltage, current, SOC, etc.) and remaining power information of the target vehicle, and use the dynamic programming algorithm to optimize the scheduling of the battery output power. The objective function is to minimize the total energy consumption of the entire driving process, and the constraints include the upper and lower limits of the battery output power, the SOC change range, the motor efficiency characteristics, etc., to solve the optimal battery output power sequence. Send the optimization result to the battery management system (BMS) and vehicle controller (VCU) of the target vehicle through the CAN bus to achieve real-time regulation of the battery output power. Simulation tests show that this method can effectively reduce the energy consumption of electric vehicles and increase the cruising range by 10% - 20%.

[0035] See Figure 2 , Figure 2 is a schematic block diagram of a computer device provided by an embodiment of the present application. The computer device 500 can be a terminal or a server. Among them, the terminal can be an electronic device with communication functions such as a smart phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant, and a wearable device. The server can be an independent server or a server cluster composed of multiple servers.

[0036] The computer device 500 includes a processor 502, a memory, and a network interface 505 connected via a system bus 501. Among them, the memory may include a non-volatile storage medium 503 and an internal memory 504.

[0037] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. When the computer program 5032 is executed, it can cause the processor 502 to execute a method for identifying parsing results.

[0038] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500.

[0039] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, it can cause the processor 502 to execute a method for identifying parsing results.

[0040] The network interface 505 is used for network communication with other devices. Those skilled in the art can understand that the above structure is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device 500 to which the solution of this application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0041] It should be understood that in the embodiment of this application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits

[0042] (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0043] It is understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a storage medium, which is a computer-readable storage medium. The computer program is executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0044] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program.

[0045] The storage medium is a physical, non-transient storage medium, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which can store program codes. The computer-readable storage medium can be non-volatile or volatile.

[0046] It should be noted that the above examples are only some specific embodiments of the present invention. Obviously, the present invention is not limited to the above embodiments, and there are many variations. All variations that can be directly derived or associated with the content disclosed by a person skilled in the art should be considered as the protection scope of the present invention.

Claims

1. A method for optimizing the output power of an electric vehicle battery, characterized in that: The method comprises: Acquiring historical driving data of similar vehicles according to the model of the target vehicle, wherein the historical driving data includes driving route, vehicle speed, battery output power, road type, road surface condition, traffic flow, weather conditions, acceleration, steering angle, and slope change information; The real-time position coordinates of the target vehicle are obtained through a vehicle positioning system, and the surrounding road information is obtained from a map database according to the real-time position coordinates. The real-time road condition information of the target vehicle is detected by a vehicle sensor, and a current position road condition feature vector is generated based on the surrounding road information and the real-time road condition information, wherein the real-time road condition information includes road surface conditions, weather conditions, vehicle speed change trends, acceleration, steering angles, and slope changes; Determine the historical driving record with the highest matching degree according to the current location road condition feature vector and the historical driving data, wherein the historical driving record with the highest matching degree includes battery output power data, battery state of charge, battery temperature, charge and discharge power, battery internal resistance, and battery capacity attenuation parameters; Determine the optimal battery output power according to the historical driving record with the highest matching degree and the current location road condition feature vector; The actual battery output power of the target vehicle is adjusted to gradually approach the optimal battery output power.

2. The method according to claim 1, characterized in that The method of obtaining historical driving data of similar vehicles according to the model of the target vehicle includes: According to the model of the target car, a list of car models similar to the target car is retrieved from a car model database to obtain a set of similar car models; For a set of similar car models, obtain the in-vehicle system interface information of each model of car, and establish a data table of mapping relationships between the model and the in-vehicle system; The historical driving data of each similar car is collected through the vehicle system interface.

3. The method according to claim 1, characterized in that The method comprises: obtaining the real-time position coordinates of the target vehicle through the vehicle positioning system, obtaining the surrounding road information from the map database according to the real-time position coordinates, detecting the real-time road condition information of the target vehicle by using the vehicle-mounted sensor, and generating the current position road condition feature vector based on the surrounding road information and the real-time road condition information. The real-time position coordinates of the target vehicle are obtained through the vehicle positioning system, and the real-time position coordinates are used as query conditions to obtain the surrounding road information of the location from a pre-built map database; Using on-board sensors to detect the driving status of the target vehicle in real time, and obtain real-time road condition information including road conditions, weather conditions, vehicle speed change trends, acceleration values, steering angles and slope changes; Based on the acquired surrounding road information and real-time traffic information, a traffic feature vector centered on the real-time location coordinates is constructed.

4. The method according to claim 1, characterized in that The determining the historical driving record with the highest matching degree according to the current location road condition feature vector and the historical driving data includes: Extracting parameters related to battery performance from the historical driving records with the highest matching degree, including battery output power data, battery state of charge SOC, battery temperature, charge and discharge power, battery internal resistance, and battery capacity attenuation parameters; According to the battery temperature and battery internal resistance, the Kalman filter algorithm is used to filter the battery output power data to obtain a filtered battery output power curve; According to the filtered battery output power curve and battery state of charge SOC, the least squares fitting algorithm is used to obtain the functional relationship between the battery output power and SOC; Substituting the charge and discharge power and battery capacity attenuation parameters into the above functional relationship, the estimated available mileage of the battery under the current road conditions is obtained; If the estimated available mileage is less than a preset threshold, it is determined that the current battery power is insufficient to support the vehicle to reach the destination, and a low-battery warning is issued to the driver.

5. The method according to claim 1, characterized in that The determining the optimal battery output power according to the historical driving record with the highest matching degree and the current location road condition feature vector comprises: Obtain the traffic feature vector of the current location, and search the historical traffic data similar to the current traffic feature vector from the historical database; According to the historical road condition data, the corresponding battery output power data is extracted to calculate the average output power value of the battery under the road condition; A battery output power optimization model is established based on battery status parameters, and the optimal battery output power under current road conditions is optimized and calculated based on the average output power value. The battery status parameters include state of charge, temperature, internal resistance and capacity attenuation.

6. The method according to claim 1, characterized in that The step of adjusting the actual battery output power of the target vehicle to gradually approach the optimal battery output power includes: Obtain the actual battery output power of the electric vehicle and the optimal battery output power under the current working conditions, and calculate the difference between the two as the target value for adjusting the battery output power; Based on the real-time temperature data collected by the battery temperature sensor, determine whether the battery temperature exceeds the preset safety range. If so, activate the battery temperature control strategy to adjust the battery temperature to a safe range by controlling the battery cooling system or heating system; Obtain the real-time charge and discharge current and voltage data of the battery, and determine whether the current charge and discharge power exceeds the safe range according to the preset charge and discharge power curve. If it exceeds, start the charge and discharge power adjustment strategy to control the battery management system to adjust the charge and discharge current to control the power within the safe range; The balancing management algorithm is used to obtain the voltage data of each battery cell, calculate the mean and variance of the cell voltage, and determine whether it exceeds the preset balancing threshold. If it exceeds, the balancing control strategy is activated to adjust the cell voltage to a balanced state through the balancing circuit; Acquire the real-time voltage data of the battery, determine whether the real-time voltage data exceeds a preset overcharge voltage threshold, and if so, control the battery management system to stop charging; determine whether the real-time voltage data is lower than a preset over-discharge voltage threshold, and if so, control the battery management system to stop discharging; The target value is adjusted according to the battery output power, and the gradient descent algorithm is used to adjust the battery output power in real time so that it gradually approaches the optimal value.

7. The method according to claim 1, characterized in that Also includes: During the driving process of the target vehicle, actual battery output power, driving route, vehicle speed, road type, road condition, traffic flow, weather conditions, acceleration, steering angle and slope change data are continuously collected, and the collected data is updated to a similar vehicle driving record library.

8. The method according to claim 7, characterized in that During the driving process of the target vehicle, the actual battery output power, driving route, vehicle speed, road type, road surface condition, traffic flow, weather conditions, acceleration, steering angle and slope change data are continuously collected, and the collected data is updated to the similar vehicle driving record library, including: Acquiring real-time driving data of the target vehicle, the real-time driving data including battery output power, driving route, vehicle speed, road conditions, traffic flow, weather conditions, acceleration, steering angle, and slope change information; The acquired real-time driving data is transmitted to the cloud server and stored in the similar car driving record library and merged with the historical data; Preprocessing the historical data and real-time data in the similar automobile driving record library to obtain a standardized data set, wherein the preprocessing includes data cleaning and data normalization; Based on the standardized data set, a machine learning algorithm is used to establish a battery output power prediction model, and the cross-validation method is used to train and optimize the model; Deploy the trained battery output power prediction model to the cloud server, receive the driving data of the target vehicle in real time, and predict the battery output power; Based on the predicted battery output power, combined with the target vehicle’s battery status and remaining power information, a dynamic programming algorithm is used to obtain an optimized battery output power scheduling strategy. The optimized battery output power scheduling strategy is sent to the target vehicle to adjust the battery output power in real time.

9. A computer device, characterized in that: It includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, for implementing the steps of the method described in any one of claims 1 to 8 when executing a program stored in a memory.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images