Electric vehicle intelligent charging method and device based on battery health and medium

Abstract

The invention discloses an electric vehicle intelligent charging method and device based on battery health and a medium, and relates to the field of new energy automobiles, and the method comprises the steps: collecting a charging data flow in the charging process of an electric vehicle through a multi-source sensor device based on a preset collection frequency corresponding to a current charging stage; transmitting the charging data flow to a cloud service platform in real time; outputting a current battery health assessment result in real time through a deep learning model according to the charging data flow; and generating a corresponding optimized charging curve according to the current battery health assessment result, generating a charging parameter adjustment instruction based on the optimized charging curve, and sending the charging parameter adjustment instruction to the charging pile terminal so as to adjust the charging parameters of the charging terminal. Through rich charging data collected by a multi-source sensor in real time and in combination with the analysis capability of a cloud deep learning model, dynamic and non-intrusive monitoring of the health state of the battery is realized, and a charging scheme with higher adaptability is provided for batteries in different health states.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of new energy vehicles, in particular to an intelligent charging method for electric vehicles based on battery health, a device and a medium. BACKGROUND

[0002] With the rapid popularization of electric vehicles, the battery as the core component of the electric vehicle directly affects the endurance mileage, safety and service life of the vehicle. Accurate assessment of the battery health status and reasonable charging strategy based on the battery health status are crucial for ensuring the normal operation of electric vehicles and the interests of users.

[0003] The traditional battery health status evaluation method relies on the data provided by the vehicle-mounted battery management system (BMS) for evaluation or offline evaluation. The BMS data usually has the problems of low collection frequency, limited data dimension, etc., which easily leads to insufficient evaluation accuracy. The offline evaluation method needs to disassemble the battery from the vehicle and use professional equipment for detection, which is not only invasive and cumbersome to operate, but also difficult to perform in real time in daily use, resulting in a lag in the evaluation results.

[0004] Moreover, most existing charging piles use standardized charging curves, i.e., the same charging parameters are used for batteries of different health statuses and different types, without considering the individual differences of batteries caused by battery health status, thereby causing problems such as overcharging or undercharging of some batteries, affecting the service life of the battery. SUMMARY

[0005] To solve the above problems, the present application provides an intelligent charging method for electric vehicles based on battery health, comprising: Through a multi-source sensor device, charging data streams in the charging process of an electric vehicle are collected based on a preset collection frequency corresponding to the current charging stage; The charging data streams are preprocessed, and the preprocessed charging data streams are transmitted to a cloud service platform in real time; According to the charging data streams, a pre-trained deep learning model in the cloud service platform is used to output the current battery health evaluation result of the electric vehicle in real time; According to the current battery health evaluation result, a corresponding optimized charging curve is generated, a charging parameter adjustment instruction is generated based on the optimized charging curve, and the charging terminal is sent to the charging terminal to adjust the charging parameters of the charging terminal.

[0006] On the other hand, the present application also provides an intelligent charging device for electric vehicles based on battery health, comprising: at least one processor; and a memory in communication connection with the at least one processor; wherein The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform a battery health-based intelligent charging method for an electric vehicle as described in the above examples.

[0007] In another aspect, the application also provides a non-volatile computer storage medium storing computer executable instructions configured to implement a battery health-based intelligent charging method for an electric vehicle as described in the above examples.

[0008] The battery health-based intelligent charging method for an electric vehicle proposed in the application can bring the following beneficial effects: In terms of battery health state monitoring, through the rich charging data collected by the multi-source sensor in real time and the analysis capability of the cloud deep learning model, the subtle state changes of the battery during the charging process can be continuously captured, so that the health assessment is always synchronized with the real-time working condition of the battery, without relying on offline disassembly or invasive detection operation, which neither interferes with the normal charging process nor makes the evaluation results more comprehensive to reflect the real health status of the battery, providing accurate and real-time state basis for subsequent charging strategy adjustment, helping to perceive the state fluctuation of the battery in advance, ensuring the long-term stable operation of the battery, and realizing dynamic and non-invasive monitoring of the battery health state.

[0009] In terms of charging experience and device function upgrade, the optimized charging curve generated based on the real-time health evaluation result fully matches the health characteristics of the individual battery, avoids the adaptation deviation of the standardized curve to the batteries with different health states, reduces unnecessary battery loss and delays the aging process, and at the same time makes the charging process more consistent with the carrying capacity of the current battery, improves the adaptation and efficiency of charging, and provides a more adaptive charging solution for batteries with different health states.

[0010] In addition, by issuing the charging parameter adjustment instruction to the charging pile in real time, the charging pile is no longer limited to a single energy injection function, but can be dynamically regulated in conjunction with battery health monitoring, giving the charging pile more intelligent regulation and control capabilities, promoting the transformation of the charging device from passive power supply to active adaptation of intelligent services, and improving the intelligent level of the overall charging system. BRIEF DESCRIPTION OF DRAWINGS

[0011] The accompanying drawings described herein are used to provide further understanding of the application, form a part of the application, and the illustrative embodiments of the application and their descriptions are used to explain the application, and do not constitute an improper limitation on the application. In the drawings: Figure 1 A flowchart of a battery health-based intelligent charging method for an electric vehicle in the embodiments of the application is shown in the figure; Figure 2A schematic diagram of an intelligent charging device for an electric vehicle based on battery health in an embodiment of the present application. DETAILED DESCRIPTION

[0012] In order to make the objectives, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in detail with reference to the embodiments of the present application and corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0013] The technical solutions provided by the embodiments of the present application will be described in detail below with reference to the drawings.

[0014] As shown in Figure 1 The embodiments of the present application provide an intelligent charging method for an electric vehicle based on battery health, which comprises: S101: Through a multi-source sensor device, charging data flow in the charging process of the electric vehicle is collected based on a preset collection frequency corresponding to a current charging stage.

[0015] When the electric vehicle is connected to the charging pile and charging, through the multi-source sensor device deployed at the terminal of the charging pile, the charging data flow in the charging process of the electric vehicle is collected based on a preset collection frequency corresponding to a current charging stage.

[0016] Among them, the multi-source sensor device includes a voltage sensor, a current sensor, a temperature sensor, a SOC sensor, a pressure sensor, etc., the voltage sensor is used to collect the battery monomer voltage and the total voltage; the current sensor is used to collect the real-time current of the charging loop; the temperature sensor is used to collect the battery monomer temperature and the battery pack environment temperature; the SOC sensor is used to estimate the remaining power; the pressure sensor is used to monitor the battery expansion degree and is used for aging evaluation.

[0017] In the embodiments of the present application, the charging of the electric vehicle is divided into four stages, including the initial charging stage, the trickle charging stage, the constant current charging stage and the constant voltage charging stage. The battery state changes at different rates in different stages, so the collection frequency needs to be set differently. The battery internal resistance is high in the initial charging stage, and the voltage rises at the fastest speed in several stages, so the collection frequency is set to high frequency; the battery internal resistance is high in the trickle charging stage, and the voltage rises fast, so the collection frequency is set to a higher frequency (such as 100 ms / time), to avoid overcharging risk; the current is stable in the constant current charging stage, and the battery state changes gently, so the collection frequency is set to medium frequency (such as 1 s / time), to balance the data volume and real-time performance; the current gradually decreases in the constant voltage charging stage, and the polarization effect is enhanced, so the collection frequency is set to high frequency (such as 200 ms / time), to accurately capture voltage fluctuations.

[0018] When the electric vehicle is just connected to the charging pile and charging, it is in the initial charging stage. Based on the first preset frequency corresponding to the initial charging stage, the initial charging data stream is collected. The initial charging data stream is calculated based on the corresponding time stamp from the collected raw charging data. The specific process is as follows: through the real-time voltage raw data, current raw data, and temperature raw data of the multi-source sensor, the data is aligned according to the time stamp, and a discrete data set corresponding to the voltage, current, and temperature parameters under the same time stamp is constructed.

[0019] For the charging voltage curve, the abnormal values in the over-rated range of the original voltage data are removed, the consistency deviation data of the single voltage is verified and completed, and the total voltage (or the average voltage of the single cell) is selected as the characteristic parameter. The preprocessed data is sorted by time, and is fitted into a continuous curve by moving average (stationary stage) or polynomial interpolation (rapid change stage).

[0020] For the charging current curve, low-pass filtering is used on the original current data to suppress high-frequency noise and remove invalid data during charging interruption. The filtered data is sorted by time, and is fitted into a continuous curve by sliding window average (constant current stage) or linear interpolation (conversion stage).

[0021] For the temperature rise curve, the original temperature data is verified and abnormal values are removed, and the average battery temperature is calculated based on the average temperature of the previous n minutes. By comparing the current average battery temperature with the reference temperature, the real-time temperature rise is obtained, which is sorted by time and fitted into a continuous curve by polynomial.

[0022] Further, by monitoring whether these initial curves reach the preset determination range corresponding to each stage (such as the voltage curve reaching the threshold of the constant current stage, the current curve tending to the stable interval, and the temperature rise curve not exceeding the upper limit of the initial stage), it is determined whether the charging stage conversion event is triggered. When the curve characteristics continuously fall into the determination range of the next stage (such as the voltage curve rising above 350V, and the current curve stabilizing in the 200A interval), the conversion event is confirmed. When any of the initial charging voltage curve, the initial charging current curve, and the initial temperature rise curve is in the corresponding preset determination range, the charging stage conversion event is triggered.

[0023] According to the determination range to which the initial curve belongs, the current charging stage is determined (such as from the trickle stage to the constant current stage), and the data is collected at the second preset frequency corresponding to the stage to generate the current charging data curve. For example, in the constant current stage, the current curve remains stable, the voltage curve rises linearly, and the temperature rise curve has a stable slope. If it is converted to the constant voltage stage, it is collected at a high frequency, the current curve shows a decay trend, and the voltage curve remains stable. In this way, the characteristics of the stage are matched, and the dynamic adaptation of data collection and curve generation is realized.

[0024] In addition, when the electric vehicle is connected to the charging pile, a handshake is performed with the vehicle battery management system of the electric vehicle in response to the electric vehicle being connected to the charging pile terminal, vehicle VIN information and maximum charging parameters allowed by the battery pack are obtained based on the vehicle model of the electric vehicle, and the electric vehicle is initially charged based on a default charging curve corresponding to the maximum charging parameters through the charging pile terminal, so that the electric vehicle starts charging by using a built-in and conservative default charging curve. The purpose of this initial stage is to safely obtain an initial charging data stream.

[0025] In the embodiment of the present application, the charging data stream is collected by the data collection module in the electric vehicle battery online health state evaluation and individualized charging system. The data collection module is integrated in the charging pile and includes voltage sensors, current sensors, temperature sensors, etc. During the charging process of the electric vehicle, the charging voltage curve, the charging current curve and the temperature rise curve are collected in real time at high frequency. The charging voltage curve includes voltage change data in different charging stages (constant current charging, constant voltage charging and trickle charging); The charging current curve corresponds to current change data in different charging stages. The temperature rise curve represents temperature change data of the battery during the charging process, which is collected by a temperature sensor installed near the charging interface or a position that can indirectly reflect the temperature of the battery.

[0026] S102: The charging data stream is preprocessed, and the preprocessed charging data stream is transmitted to the cloud service platform in real time.

[0027] The original data stream is cleaned and converted, including outlier removal, missing value filling, unit standardization, feature extraction, etc. The outlier removal is to remove the jump value caused by sensor failure by 3σ principle or sliding window method. The missing value filling is to supplement the missing data caused by temporary communication interruption by linear interpolation method before and after the time. The unit standardization is to convert the data such as voltage (V), current (A) and temperature (℃) into standard range for model training. The feature extraction is to calculate derived features (such as "current change rate ΔI / Δt" and "temperature gradient ΔT / ΔSOC") to enhance the representation ability of data to the battery state.

[0028] The preprocessed data is uploaded to the cloud service platform in real time through the Internet of Vehicles (T-BOX) or the charging pile communication module (such as CAN bus, 4G / 5G). Encryption protocols (such as TLS) are used in the transmission process to ensure data security and avoid tampering.

[0029] In the embodiment of the present application, the data transmission and preprocessing module in the electric vehicle battery online health state evaluation and individualized charging system performs data preprocessing and data transmission.

[0030] The collected raw data is transmitted to the cloud platform in real time through wired (such as Ethernet) or wireless (such as 4G, 5G, Wi-Fi) communication mode. In order to ensure the safety and reliability of data transmission, an encrypted transmission protocol is used.

[0031] The raw data transmitted to the cloud is processed, including data cleaning (removing outliers and noise), data normalization (converting data of different magnitudes to the same magnitude), data dimension reduction (reducing data dimension to improve subsequent analysis efficiency), etc.

[0032] S103: According to the charging data stream, the pre-trained deep learning model in the cloud service platform is used to output the current battery health assessment result of the electric vehicle in real time.

[0033] Specifically, based on the collection timestamp, the charging data stream is sorted, and a charging time sequence vector is constructed and input into the pre-trained deep learning model in the cloud service platform. Through the deep learning model, the multi-dimensional time sequence features of the charging time sequence vector are extracted, and based on the multi-dimensional time sequence features, the current battery health state of the electric vehicle is evaluated, and the current battery health assessment result is output.

[0034] In the embodiments of the present application, the charging data stream is sorted according to the collection time stamp of each data point in chronological order, ensuring the time sequence continuity of the data, and the sorted multi-dimensional parameters (such as voltage value, current value, temperature rise value at a certain time) are organized into a structured charging time sequence vector according to the time sequence.

[0035] For example, 100 parameters at 100 consecutive time points are combined into a vector [V1, I1, ΔT1, V2, I2, ΔT2, …, V100, I100, ΔT100] with an interval of 100 ms, which is input to the cloud pre-trained deep learning model.

[0036] The deep learning model extracts multi-dimensional time sequence features from the charging time sequence vector through a multi-layer network structure, including short-term features (such as instantaneous fluctuation amplitude of voltage, mutation slope of current), long-term features (such as overall rising trend of voltage in constant current stage, cumulative change rate of temperature rise curve) and stage transition features (such as current decay mode from constant current to constant voltage stage).

[0037] Based on the multi-dimensional time sequence features, the model combines the time sequence features learned in the pre-training process and the correlation rules of battery health (such as slow voltage rise of aged battery, high temperature rise rate), and evaluates the current battery health state (such as SOH health degree, internal resistance state, potential aging risk). Finally, the quantitative health assessment result is output, including internal resistance change, capacity attenuation (SOH), consistency and other health state parameters of the battery.

[0038] It should be noted that the deep learning model can be a time series model such as LSTM, TCN, etc.

[0039] In the embodiment of the present application, the pre-training process of the deep learning model is as follows: collecting historical charging data streams containing time series data such as voltage, current, temperature rise of electric vehicles in different health states, and matching corresponding battery health state true values such as SOH values obtained by offline detection, internal resistance, etc., to construct a training data set covering the whole life cycle health state, ensuring that the data can reflect the complete change rule of the battery from new to aging.

[0040] The training data set is preprocessed, such as eliminating outliers, standardizing parameter ranges, sorting time stamps to construct time series vectors, etc., to eliminate data noise and unify the format; the preprocessed data is input into the deep learning model to be trained, and the model outputs the battery health state prediction value based on the current parameters. The difference (loss value) between the prediction value and the true value is calculated by a loss function (such as mean square error), and the loss signal is transmitted to each layer of the model using the back propagation algorithm to adjust the weight, bias and other parameters to reduce the loss. Repeat the above data input, prediction, loss calculation and parameter adjustment process until the loss value stabilizes at a low level, and the model converges. At this time, the model has learned the mapping relationship between the charging time series characteristics and the battery health state, and can be used for subsequent real-time evaluation.

[0041] In the embodiment of the present application, the model training and health evaluation process is realized through the online health state evaluation of the electric vehicle battery and the cloud AI analysis and health evaluation module in the individualized charging system. Among them, the model training is based on a large number of batteries of different types and different health states in the charging process to construct and train a deep learning model (such as an LSTM neural network model). The input of the model is the preprocessed charging voltage curve, current curve, temperature rise curve and other data, and the output is the health state parameters of the battery such as internal resistance change, capacity attenuation (SOH), consistency, etc. The online evaluation inputs the real-time collected and preprocessed current charging data into the trained deep learning model, and the model outputs the evaluation result of the current health state of the battery in real time.

[0042] S104: According to the current battery health evaluation result, the corresponding optimized charging curve is generated, the charging parameter adjustment instruction is generated based on the optimized charging curve, and is sent to the charging pile terminal to adjust the charging parameter of the charging terminal.

[0043] According to the health evaluation result, the charging parameter is dynamically adjusted to maximize the charging efficiency under the premise of ensuring the health of the battery. The key health state parameter values such as SOH health degree, internal resistance, maximum allowable charging current are extracted from the current battery health evaluation result, and the key health state parameter values are compared with the preset multiple health state region thresholds to determine the health state region to which the current battery belongs.

[0044] The optimization charging strategy template corresponding to the health state area is called, and the template pre-defines the core charging parameter rules suitable for the battery in the area, including the maximum current limit in the constant current stage, the target voltage value in the constant voltage stage, and the trigger condition of the trickle charging. Combined with the current specific health state parameter value, an optimized charging curve suitable for the current battery health state is generated, achieving the goal of both ensuring charging efficiency and protecting battery life.

[0045] For example, if the battery health is high (SOH≥90%), the curve is biased towards high-efficiency charging, maintaining a high current in the constant current stage to shorten the charging time; if the battery health is medium (80%≤SOH<90%), the curve limits the maximum current and reduces the current after SOC80% to reduce polarization; if the battery health is low (SOH<80%), the curve adopts a small numerical multi-stage ladder descending mode to avoid overcharging stress.

[0046] Further, the instructions are sent to the charging pile terminal through the communication interface of the cloud and the charging pile (such as the OCPP protocol), and the charging pile controller adjusts the output parameters in real time according to the instructions to complete the closed-loop control.

[0047] Further, new charging data streams are continuously collected at the preset collection frequency corresponding to the current charging stage, and these dynamically updated data are uploaded to the cloud service platform in real time to ensure that the cloud can obtain the latest state information of the battery during charging. After receiving the new data stream, the cloud service platform inputs it into the pre-trained deep learning model, which extracts the time sequence features based on the latest data and updates the evaluation of the battery health state. For example, if the new data shows that the temperature rise rate suddenly increases, the model may lower the health and safety threshold of the current battery and output an updated health state evaluation result, such as adjusting the maximum allowed charging current from 200A to 180A.

[0048] According to the updated battery health evaluation result, the original optimized charging curve is adjusted in real time: if the evaluation result shows that the current heating risk of the battery increases, the current value in the constant current stage will be lowered and the constant voltage stage will be triggered earlier; if the evaluation shows that the battery state is stable, the charging parameters can be appropriately maintained or fine-tuned. The adjusted optimized charging curve is converted into updated charging parameter adjustment instructions and is sent to the charging pile terminal in real time, so that the output parameters of the charging pile are always matched with the real-time health state of the battery, forming a dynamic closed loop of data collection, evaluation update, and strategy adjustment, which not only ensures charging safety, but also maximizes the charging efficiency of the current state of the battery.

[0049] In the embodiments of the present application, through the online health state evaluation of the electric vehicle battery and the individualized charging strategy generation module in the individualized charging system, the battery health state evaluation result output by the cloud AI analysis and health evaluation module is combined with the type, service life and other basic information of the battery to generate an individualized optimal charging curve. It can include: optimizing the current size and duration of the constant current charging phase, appropriately reducing the constant current charging current for batteries with serious capacity attenuation. Adjust the voltage value of the constant voltage charging phase, and according to the consistency of the battery, the equalization charging adjustment is carried out for the battery with poor consistency. Optimize the trickle charging phase, shorten the trickle charging time for the battery with large internal resistance change, and avoid overcharging.

[0050] Through the execution and feedback module in the online health state evaluation of the electric vehicle battery and the individualized charging system, the charging curve parameters output by the individualized charging strategy generation module are received, and the charging operation is performed according to the parameters. In the charging process, the data acquisition module continuously acquires charging data and transmits it to the cloud. The cloud deep learning model dynamically corrects the battery health state evaluation result according to the new data, and at the same time, the individualized charging strategy generation module also adjusts the charging curve in real time according to the corrected evaluation result, forming a closed loop control.

[0051] In the battery health state monitoring layer, through the rich charging data collected by the multi-source sensor in real time, combined with the analysis ability of the cloud deep learning model, the subtle state change of the battery in the charging process can be continuously captured, so that the health evaluation is always synchronized with the real-time working condition of the battery. Without relying on offline disassembly or invasive detection operation, it neither interferes with the normal charging process, nor makes the evaluation result more fully reflect the real health status of the battery, providing accurate and real-time state basis for subsequent charging strategy adjustment, helping to perceive the state fluctuation of the battery in advance, ensuring the long-term stable operation of the battery, and realizing the dynamic and non-invasive monitoring of the battery health state.

[0052] In the charging experience and device function upgrading layer, the optimized charging curve generated based on the real-time health evaluation result fully fits the health characteristics of the individual battery, avoids the adaptation deviation of the standardized curve to different health state batteries, reduces unnecessary battery loss, delays the aging process, and at the same time makes the charging process more consistent with the current battery carrying capacity, improves the adaptation and efficiency of charging, and provides a more adaptive charging scheme for batteries with different health states.

[0053] In addition, by issuing the charging parameter adjustment instruction in real time to the charging pile, the charging pile is no longer limited to a single energy injection function, but can be linked with the dynamic regulation and control of the battery health monitoring, giving the charging pile more rich intelligent regulation and control ability, promoting the transformation of the charging equipment from passive power supply to active adaptation intelligent service, and improving the intelligent level of the overall charging system.

[0054] AsFigure 2 As shown in the embodiments of the present application, the present application also proposes an intelligent charging device for electric vehicles based on battery health, comprising: at least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method for intelligent charging of electric vehicles based on battery health as described in any of the above embodiments.

[0055] The embodiments of the present application also provide a non-volatile computer storage medium, which stores computer executable instructions, and the computer executable instructions are configured to perform the method for intelligent charging of electric vehicles based on battery health as described in any of the above embodiments.

[0056] Each of the embodiments of the present application adopts a progressive manner for description, and the same or similar parts between the embodiments can be referred to each other. Each of the embodiments mainly describes the difference from other embodiments. In particular, the device and medium embodiments are basically similar to the method embodiments, and thus the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

[0057] The device and medium provided by the embodiments of the present application are one-to-one corresponding to the method, and thus the device and medium also have the similar beneficial technical effects as the method. Since the beneficial technical effects of the method have been described in detail above, the beneficial technical effects of the device and medium will not be described here.

[0058] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes.

[0059] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0060] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0061] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks Figure 1 means for functionally implementing the steps listed in the flowchart block or blocks.

[0062] In one typical configuration, the computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0063] The memory can include non-persistent memory and / or volatile memory, such as random access memory (RAM) and / or cache memory, non-volatile memory, etc. in the form of computer-readable media. The memory is an example of computer-readable media.

[0064] Computer-readable media includes permanent and non-permanent, movable and non-movable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape, disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information accessible to a computing device. According to the definition herein, computer-readable media does not include transitory media such as modulated data signals and carriers.

[0065] It should also be noted that the terms "comprising", "containing", or any other variant thereof are intended to cover a non-exclusive inclusion, such that a process, method, article or apparatus that comprises a list of elements does not only include those elements, but can also include other elements not expressly listed or inherent to such process, method, article or apparatus. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of additional identical elements in the process, method, article or apparatus that includes the element.

[0066] The above only describes the embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the scope of claims of the present application.

Claims

1. A battery health based intelligent charging method for electric vehicles, characterized in that, The method comprises the following steps: acquiring, by a multi-source sensor device, a charging data stream in a charging process of an electric vehicle based on a preset acquisition frequency corresponding to a current charging stage; preprocessing the charging data stream and transmitting the preprocessed charging data stream to a cloud service platform in real time; outputting, by a pre-trained deep learning model in the cloud service platform, a current battery health evaluation result of the electric vehicle in real time based on the charging data stream; generating an optimized charging curve corresponding to the current battery health evaluation result, generating a charging parameter adjustment instruction based on the optimized charging curve, and sending the charging parameter adjustment instruction to a charging pile terminal to adjust the charging parameters of the charging terminal.

2. The battery health based intelligent charging method for electric vehicles as claimed in claim 1 wherein, Before the step of acquiring, by a multi-source sensor device, a charging data stream in a charging process of an electric vehicle based on a preset acquisition frequency corresponding to a current charging stage, the method further comprises the following steps: performing a handshake with a vehicle battery management system of the electric vehicle in response to the electric vehicle accessing a charging pile terminal; acquiring a maximum charging parameter allowed by a battery of the electric vehicle based on a vehicle model of the electric vehicle; initially charging the electric vehicle based on a default charging curve corresponding to the maximum charging parameter by the charging pile terminal.

3. The battery health based smart charging method for electric vehicles as claimed in claim 2, wherein, The charging data stream comprises a charging voltage curve, a charging current curve, and a temperature rise curve. The step of acquiring, by a multi-source sensor device, a charging data stream in a charging process of an electric vehicle based on a preset acquisition frequency corresponding to a current charging stage comprises the following steps: acquiring an initial charging data stream based on a first preset frequency corresponding to an initial charging stage by a multi-source sensor device; monitoring the initial charging data stream in real time based on a preset determination range corresponding to each charging stage to determine whether a charging stage conversion event is triggered; if yes, determining a current charging stage based on a preset determination range to which the initial charging data stream belongs; acquiring a current charging data stream based on a second preset acquisition frequency corresponding to the current charging stage.

4. The battery health based smart charging method for electric vehicles as claimed in claim 3, wherein, The step of monitoring the initial charging data stream in real time based on a preset determination range corresponding to each charging stage to determine whether a charging stage conversion event is triggered comprises the following steps: acquiring a preset determination range corresponding to each charging stage; the preset determination range comprises a voltage variation range, a current variation range, and a temperature variation range; monitoring the initial charging data stream in real time, comparing an initial charging voltage curve, an initial charging current curve, and an initial temperature rise curve in the initial charging data stream with a corresponding preset determination range, respectively; when any of the initial charging voltage curve, the initial charging current curve, and the initial temperature rise curve is within the corresponding preset determination range, triggering a charging stage conversion event.

5. The battery health based intelligent charging method for electric vehicles as claimed in claim 1 wherein, The step of outputting, by a pre-trained deep learning model in the cloud service platform, a current battery health evaluation result of the electric vehicle in real time based on the charging data stream comprises the following steps: sorting the charging data stream based on an acquisition timestamp, constructing a charging time sequence vector, and inputting the charging time sequence vector into the pre-trained deep learning model in the cloud service platform; The deep learning model extracts multi-dimensional time sequence features of the charging time sequence vector, evaluates the current battery health state of the electric vehicle based on the multi-dimensional time sequence features, and outputs a current battery health evaluation result.

6. The battery health based intelligent charging method for electric vehicles as claimed in claim 1 wherein, The pre-training process of the deep learning model specifically includes: Obtain historical charging data streams and corresponding battery health state true values to construct a training data set; the historical charging data streams include charging data streams corresponding to electric vehicles with different health states; Preprocess the training data set, input the preprocessed training data set into the deep learning model, and obtain battery health state prediction values output by the deep learning model; Based on the battery health state true values, calculate the loss value of the battery health state prediction values, adjust the model parameters according to the loss value, and stop until the model converges.

7. The battery health based smart charging method for electric vehicles as claimed in claim 1 wherein, The current battery health evaluation result is used to generate a corresponding optimized charging curve, specifically including: Obtain the health state parameter value in the current battery health evaluation result, compare the health state parameter value with the threshold value corresponding to a plurality of preset health state regions, and determine the preset health state region to which the health state parameter value belongs; Call the optimized charging strategy template corresponding to the preset health state region, generate a corresponding optimized charging curve according to the health state parameter value; the strategy template is used to define the current value of the constant current charging stage, the voltage value of the constant voltage charging stage, and the trigger condition of the trickle charging according to the health state parameter value.

8. The battery health based smart charging method for electric vehicles as claimed in claim 1, wherein, Based on the optimized charging curve, generate a charging parameter adjustment instruction and send it to a charging pile terminal to adjust the charging parameters of the charging terminal, and the method further includes: Continuously collect new charging data streams, upload the new charging data streams to the cloud service platform, output updated battery health state evaluation results through the deep learning model; According to the updated evaluation results, real-time adjust the optimized charging curve, generate an updated charging parameter adjustment instruction and send it to the charging pile terminal.

9. A battery health based intelligent charging device for electric vehicles, characterized in that, It includes: At least one processor; And The memory is in communication connection with the at least one processor; wherein The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the battery health-based electric vehicle intelligent charging method according to any one of claims 1-8.

10. A non-transitory computer storage medium storing computer-executable instructions, the computer-executable instructions comprising instructions for: receiving a request to access a file; determining whether the file is stored in a cache; and in response to determining that the file is stored in the cache, providing access to the file from the cache. The computer executable instructions are configured to execute the battery health-based electric vehicle intelligent charging method according to any one of claims 1-8.

Images