Battery life prediction method and device and computer equipment
By determining the half-peak width integral area value and peak voltage in lithium-ion batteries, combining machine learning models and particle filtering algorithms, dynamically updating parameters, the problems of complexity and low efficiency of battery life prediction in the prior art are solved, and accurate and real-time prediction of the remaining battery life are achieved.
Patent Information
- Application Number
- CN202510895203.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-30
- Publication Date
- 2025-08-15
AI Technical Summary
Prior Art In the fields of new energy vehicles, the life prediction method of lithium-ion batteries is complex and cannot meet the continuous changing working conditions, resulting in low battery life prediction efficiency.
By determining the half-peak width integral area value and peak voltage of the battery cell during the charging cycle, the prediction is performed using a pre-trained machine learning model, combined with a particle filtering algorithm, the parameters are dynamically updated to achieve the prediction of battery capacity and internal resistance, and then the remaining service life is determined.
It improves the prediction accuracy and real-time of the remaining service life of the battery, and provides timely and accurate information for battery maintenance and management.
Smart Images
Figure CN120490839A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of electronic technology, and in particular to a battery life prediction method, device, and computer equipment. Background Art
[0002] With the development and research of lithium-ion batteries, they have found widespread application in many fields, such as electric vehicles, unmanned systems, and energy storage. Currently, new energy vehicles are increasing in number, and the majority of their battery systems are lithium-ion batteries. As battery capacity declines, their output power decreases, causing the overall performance of lithium-ion batteries to gradually deteriorate, until they are unable to complete their intended tasks.
[0003] In related technologies, life prediction is achieved by establishing battery degradation empirical equations or electrochemical principles.
[0004] However, the above method requires more complex battery testing or degradation modeling, which limits the application scenarios of this method and cannot meet the application scenarios of new energy vehicles with continuously changing operating conditions, thereby reducing the prediction efficiency of the remaining battery life. Summary of the Invention
[0005] This application provides a battery life prediction method, device, and computer equipment to effectively improve the accuracy of predicting the remaining battery life. The technical solution is as follows:
[0006] According to one aspect of the present application, a battery life prediction method is provided, the method comprising:
[0007] Determine a half-height peak width integrated area value and a peak voltage of a battery cell during a charging cycle, wherein the half-height peak width integrated area value is used to describe a change in battery capacity of the battery cell during the charging cycle, and the peak voltage refers to the maximum voltage of the battery cell during the charging cycle;
[0008] Inputting the half-height peak width integrated area value and the peak voltage into a battery attenuation prediction model to predict a half-height peak width integrated area value prediction value and a peak voltage prediction value of the battery cell after the charging cycle;
[0009] Inputting the half-height peak width integrated area prediction value and the peak voltage prediction value into a battery capacity and internal resistance prediction model to predict a battery capacity prediction value and a battery internal resistance prediction value;
[0010] determining a remaining service life of the battery cell based on the battery capacity prediction value and the battery internal resistance prediction value;
[0011] Based on the battery capacity prediction value and the battery internal resistance prediction value, a particle filter algorithm is used to determine a confidence result of the remaining service life, where the confidence result is used to evaluate the accuracy of the remaining service life;
[0012] Among them, the battery attenuation prediction model and the battery capacity internal resistance prediction model are pre-trained machine learning models.
[0013] According to one aspect of the present application, a battery life prediction device is provided, the device comprising:
[0014] a determination module, configured to determine a half-height peak width integrated area value and a peak voltage of a battery cell during a charging cycle, wherein the half-height peak width integrated area value is used to describe a change in battery capacity of the battery cell during the charging cycle, and the peak voltage refers to a maximum voltage of the battery cell during the charging cycle;
[0015] An input module, configured to input the half-height peak width integrated area value and the peak voltage into a battery attenuation prediction model to predict a predicted value of the half-height peak width integrated area value and a predicted value of the peak voltage of the battery cell after the charging cycle;
[0016] The input module is further used to input the half-peak width integrated area prediction value and the peak voltage prediction value into the battery capacity and internal resistance prediction model to predict the battery capacity prediction value and the battery internal resistance prediction value;
[0017] The determination module is further configured to determine the remaining service life of the battery unit based on the battery capacity prediction value and the battery internal resistance prediction value;
[0018] The determination module is further configured to determine a confidence result of the remaining useful life using a particle filter algorithm based on the battery capacity prediction value and the battery internal resistance prediction value, wherein the confidence result is used to evaluate the accuracy of the remaining useful life;
[0019] Among them, the battery attenuation prediction model and the battery capacity internal resistance prediction model are pre-trained machine learning models.
[0020] According to another aspect of the present application, a computer device is provided, which includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by the processor to implement the battery life prediction method as described in any of the above embodiments of the present application.
[0021] According to another aspect of the present application, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is loaded and executed by a processor to implement the above battery life prediction method.
[0022] According to another aspect of the present application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the above-described battery life prediction method.
[0023] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:
[0024] By nesting different prediction models, the battery capacity and internal resistance of a battery cell after a charge cycle are predicted. The remaining useful life of the battery cell is then determined based on these values. A particle filtering algorithm is then used to determine the confidence level of this remaining useful life. This confidence level determines whether the predicted remaining useful life is reliable, further ensuring the accuracy of the remaining useful life prediction. Furthermore, the particle filtering algorithm dynamically updates the parameters involved in the particle filtering process based on the predicted battery capacity and internal resistance values and the current state of the battery cell, enabling real-time prediction and verification of the battery's remaining useful life, providing timely and accurate information for battery cell maintenance and management. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0026] Figure 1 is a structural block diagram of a terminal provided by an exemplary embodiment of the present application;
[0027] Figure 2 This is a flowchart of an execution flow of a battery life prediction method provided by an exemplary embodiment of the present application;
[0028] Figure 3 This is a flowchart of the execution flow of model training provided by an exemplary embodiment of the present application;
[0029] Figure 4 This is a structural block diagram of a battery life prediction device provided by an exemplary embodiment of the present application;
[0030] Figure 5 is a structural block diagram of a battery life prediction device provided by another exemplary embodiment of the present application;
[0031] Figure 6 It is a structural block diagram of a computer device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION
[0032] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.
[0033] Figure 1 A block diagram of a terminal provided by an exemplary embodiment is shown. Based on this block diagram, the execution process of the battery life prediction method provided by an embodiment of the present application is described. The block diagram includes a terminal 10, which integrates a battery degradation prediction model 100 and a battery capacity and internal resistance prediction model 110.
[0034] The following will briefly introduce the data flow of the terminal 10 during the execution of the battery life prediction method. For the contents of the battery degradation prediction model 100 and the battery capacity internal resistance prediction model 110, please refer to the following embodiments.
[0035] The terminal 10 executes a data acquisition process, in which the terminal 10 acquires the half-peak width integrated area value and the peak voltage generated by the battery cell during the charging cycle.
[0036] The terminal 10 inputs the obtained half-peak width integrated area value and peak voltage into the battery attenuation prediction model 100. The battery attenuation prediction model 100 predicts the first predicted value of the battery cell after the charging cycle based on the half-peak width integrated area value and the peak voltage. The first predicted value includes the half-peak width integrated area value predicted value and the peak voltage predicted value.
[0037] The terminal 10 inputs the first prediction value into the battery capacity and internal resistance prediction model 110 , which predicts the battery capacity and internal resistance of the battery cell to obtain a second prediction value. The second prediction value includes a battery capacity prediction value and a battery internal resistance prediction value.
[0038] Terminal 10 obtains the remaining useful life of the battery cell based on the first prediction value. It then processes the battery capacity prediction value and the battery internal resistance prediction value using a particle filter algorithm to obtain a confidence result for the remaining useful life. The accuracy and reliability of the remaining useful life of the battery cell are determined based on the confidence result.
[0039] In some embodiments, the terminal 10 includes but is not limited to smart phones, tablet computers, laptop computers, desktop computers, smart home appliances, smart car terminals, smart speakers, digital cameras, smart voice interaction devices, smart home appliances, aircraft, etc.
[0040] In the embodiment of the present application, the terminal 10 is implemented as an intelligent vehicle terminal. Optionally, the terminal 10 is deployed with a battery degradation prediction model 100 and a battery capacity and internal resistance prediction model 110, and the terminal 10 alone realizes the purpose of battery cell life prediction.
[0041] Illustratively, the terminal 10 receives battery charging data to be predicted (such as the half-peak width integrated area value and peak voltage, etc.), and calls the battery attenuation prediction model 100 and the battery capacity internal resistance prediction model 110 to execute the corresponding process of the battery life prediction method.
[0042] In other embodiments, the battery life prediction method provided in the embodiments of the present application is jointly implemented by the terminal 10 and the server. A battery attenuation prediction model 100 and a battery capacity internal resistance prediction model 100 are deployed in the server. The terminal receives the battery charging data to be predicted and sends the battery charging data to the server. The server receives the battery charging data to be predicted sent by the terminal 10, and calls the battery attenuation prediction model 100 and the battery capacity internal resistance prediction model 100 to execute the corresponding process of the battery life prediction method. After the server obtains the confidence result corresponding to the remaining service life of the battery cell, it transmits the confidence result back to the terminal 10. The terminal 10 receives the confidence result sent by the server and displays the confidence result in the terminal 10.
[0043] It's worth noting that a server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), big data, and artificial intelligence platforms. Cloud technology refers to a hosting technology that unifies hardware, software, and network resources within a wide area network or local area network to enable data computing, storage, processing, and sharing. Cloud technology is a general term for network technology, information technology, integration technology, management platform technology, and application technology based on the cloud computing business model. It can form a resource pool that can be used on demand and flexibly traversed. Cloud computing technology will become a key support. The backend servers of technical network systems require extensive computing and storage resources. Examples include video websites, image networks, and more portals. With the rapid development and application of the internet industry, every item in the future will likely have its own unique identification mark, which will need to be transmitted to the backend system for logical processing. Different levels of data will be processed separately. All kinds of industry data require a strong system backend, which can only be achieved through cloud computing. Optionally, the server can also be implemented as a node in the blockchain system.
[0044] It should be noted that this application can display a prompt interface, pop-up window or output voice prompt information before collecting the user's relevant data and during the process of collecting the user's relevant data. The prompt interface, pop-up window or voice prompt information is used to remind the user that its relevant data is currently being collected, so that this application only starts to execute the relevant steps of obtaining the user's relevant data after obtaining the user's confirmation operation on the prompt interface or pop-up window. Otherwise (that is, when the user's confirmation operation on the prompt interface or pop-up window is not obtained), the relevant steps of obtaining the user's relevant data are terminated, that is, the user's relevant data is not obtained. In other words, all user data collected by this application are collected with the user's consent and authorization, and the collection, use and processing of relevant user data need to comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0045] Example 1:
[0046] like Figure 2 As shown, Figure 2 The following is a flowchart of the execution flow of a battery life prediction method provided by an exemplary embodiment of the present application. Figure 1 The terminal 10 is shown for description.
[0047] Step 200: Determine the half-peak width integrated area value and peak voltage of the battery cell during a charging cycle.
[0048] Optionally, a battery cell is a device that converts chemical energy into electrical energy, and current is generated through chemical reactions inside the battery cell.
[0049] In one embodiment, the battery cell is any sub-unit in a battery pack, that is, the battery pack includes at least one battery cell.
[0050] There are many types of battery cells, which can be classified according to any one or more of the following classification standards.
[0051] 1) Classify according to whether it is rechargeable.
[0052] Divided into primary batteries and secondary batteries.
[0053] Primary batteries are designed to be used only once and cannot be recharged after discharge. Primary batteries are defined as zinc-manganese dry cells, alkaline zinc-manganese batteries, or lithium primary batteries.
[0054] A secondary battery is a battery that can be used multiple times through charge and discharge cycles. It can be implemented as a lithium-ion battery, a nickel-metal hydride battery, or a lead-acid battery.
[0055] 2) Classification according to electrolyte status.
[0056] It is classified into at least one of liquid electrolyte batteries and solid electrolyte batteries.
[0057] Liquid electrolyte batteries are batteries whose electrolyte is liquid and has good ionic conductivity. Liquid electrolyte batteries include but are not limited to lithium-ion batteries and lead-acid batteries.
[0058] Solid-state electrolyte batteries refer to batteries whose electrolytes are solid materials. Solid-state electrolyte batteries include but are not limited to lithium solid-state batteries and sodium solid-state batteries.
[0059] 3) Classify according to battery shape.
[0060] It is classified into at least one of cylindrical batteries, square batteries, and soft-pack batteries.
[0061] Cylindrical batteries are cylindrical in shape. Cylindrical batteries include but are not limited to lithium-ion batteries, alkaline batteries, etc.
[0062] Prismatic batteries are square in shape. They include, but are not limited to, prismatic lithium-ion batteries and prismatic lead-acid batteries.
[0063] The shell of the soft-pack battery is encapsulated with flexible materials. Soft-pack batteries include but are not limited to soft-pack lithium-ion batteries, etc.
[0064] 4) Classify according to battery usage.
[0065] It is divided into at least one of consumer electronic batteries, automotive batteries, energy storage batteries, medical batteries and industrial batteries.
[0066] Consumer electronic batteries are mainly used in portable electronic devices such as mobile phones, laptops, tablets, etc. Consumer electronic batteries include but are not limited to lithium-ion batteries and polymer lithium-ion batteries, etc.
[0067] Automotive batteries are primarily used in the starting, lighting, and ignition systems of vehicles, as well as the power systems of electric vehicles. These batteries include, but are not limited to, lead-acid batteries, lithium-ion batteries, and nickel-metal hydride batteries.
[0068] Energy storage batteries are primarily used for grid energy storage, home energy storage, and industrial energy storage to balance power supply and demand and improve grid stability. Energy storage batteries include, but are not limited to, lithium-ion batteries, sodium-sulfur batteries, and flow batteries.
[0069] Medical batteries are mainly used in medical devices such as pacemakers, hearing aids, etc. Medical batteries include but are not limited to lithium primary batteries and zinc-air batteries.
[0070] Industrial batteries are mainly used in industrial equipment such as forklifts, power tools, etc. Industrial batteries include but are not limited to lead-acid batteries and lithium-ion batteries.
[0071] 5) Classify according to the working principle of the battery.
[0072] It is classified into at least one of chemical batteries, physical batteries and biological batteries.
[0073] Chemical batteries are batteries that convert chemical energy into electrical energy through chemical reactions. Chemical batteries include but are not limited to lithium-ion batteries, nickel-metal hydride batteries, and lead-acid batteries.
[0074] A physical battery is one that converts other forms of energy into electricity through a physical process. For example, a fuel cell generates electricity through a chemical reaction between hydrogen and oxygen, which is a type of physical battery.
[0075] Biobatteries are batteries that use chemical reactions within living organisms to generate electricity. For example, microbial fuel cells use the metabolic processes of microorganisms to generate electricity.
[0076] The above classification standards are only illustrative examples. Battery cells can also be classified according to other classification standards, which are not limited in this application.
[0077] Schematically, the working principle of a lithium-ion battery cell is introduced by taking a lithium-ion battery cell as an example: when a lithium-ion battery cell is used and charged, it is divided into a charging process and a discharging process.
[0078] Charging Process: When charging a battery cell, the positive electrode of an external power source is connected to the positive electrode of the battery cell, and the negative electrode of the external power source is connected to the negative electrode of the battery cell. At this point, the lithium ions inside the battery cell are released from the positive electrode material under the influence of the electric field, pass through the electrolyte, and embed into the negative electrode material (such as graphite). Simultaneously, electrons flow in the external circuit from the negative electrode of the external power source to the negative electrode of the battery cell, and then flow through the conductive structure inside the battery cell to the positive electrode, in the opposite direction of the movement of the lithium ions, thus completing the charging process and converting electrical energy into chemical energy for storage in the battery cell.
[0079] Discharge process: When a battery cell is discharged and connected to an electrical appliance, the lithium ions inside the battery cell are released from the negative electrode material, pass through the electrolyte, and are re-embedded in the positive electrode material. At the same time, electrons flow out of the negative electrode of the battery cell, pass through the electrical appliance, and then flow into the positive electrode of the battery cell, in the opposite direction of the movement of the lithium ions, forming an electric current, converting chemical energy into electrical energy, which provides power to the electrical appliance.
[0080] Alternatively, a charge cycle refers to the process of charging a battery cell from a fully discharged state until it is fully charged again. The charge cycle can be used to measure the battery performance and service life of a battery cell.
[0081] In another optional embodiment, a charging cycle refers to a time period during which a battery cell performs a charging process or a discharging process.
[0082] In another optional embodiment, the charging cycle refers to the time period of the battery cell during the charging process or the discharging process. In this case, the time period is the time period from the time point of starting the charging process or the time point of starting the discharging process to the current time point.
[0083] Optionally, during the charging cycle, the half-peak width integrated area value is the value obtained by integrating the area of the peak based on the width at half the peak height (i.e., half-peak width) in the charging curve (discharge curve). Schematically, when the voltage in the charging process (discharge process) reaches the peak value of a peak, the half-peak width of the peak is recorded, and the charged amount (discharged amount) is continuously integrated within the width range of the half-peak width until the voltage exceeds the half-peak interval range, and then the integral value of the amount is calculated and recorded as the half-peak width integrated area value. The half-peak width integrated area value can reflect comprehensive information such as the intensity and width of the peak, and is used to evaluate the battery health status of the battery cell.
[0084] Optionally, during a charging cycle, the peak voltage refers to the highest voltage that a battery cell can reach during charging (discharging). In another optional embodiment, the peak voltage is determined based on the highest voltage that a battery cell can reach during charging (discharging).
[0085] In an embodiment of the present application, charging data generated by a battery cell during a charging cycle is obtained.
[0086] Charging data includes timestamp information of the charging cycle, voltage, current, state of charge, total mileage information, etc.
[0087] Optionally, when the battery unit is implemented as any battery in a battery pack, the charging data includes the timestamp information of the charging cycle, the total voltage of the battery pack, the total current of the battery pack, the single cell voltage of each battery cell, the single cell current of each battery cell, the state of charge, the total mileage information, etc.
[0088] The charging data is filtered according to the charging status field to obtain the voltage data and current data of the battery unit during the charging cycle.
[0089] The charging status field is a set of parameters or data used to describe the current charging status of the battery cell. The charging status field includes multiple key information used to monitor and manage the charging process of the battery.
[0090] The charging status field includes at least one of the following: charging status, charging mode, charging current, charging voltage, battery temperature, battery capacity, charging time, charging efficiency, charging status indication, charging fault code, charging termination reason, battery health status, battery service life, charging power, battery voltage, battery current, battery charge, battery health, battery internal resistance, battery temperature change rate, battery voltage change rate, battery current change rate, battery balancing status, battery balancing current, battery balancing time, etc.
[0091] The charging status is used to indicate whether the battery cell is currently in a charging state, the charging mode is used to indicate the charging mode currently adopted by the battery cell (including but not limited to at least one of constant current charging, constant voltage charging, trickle charging, and fast charging), the charging current is used to indicate the current value flowing into the battery cell, the charging voltage is used to indicate the voltage value currently applied to the battery cell, the battery temperature is used to indicate the current temperature of the battery cell, the battery capacity is used to indicate the current remaining capacity of the battery cell, the charging time is used to indicate the time from the start of charging of the battery cell to the present, the charging efficiency is used to indicate the energy conversion efficiency of the battery cell during the charging cycle, the charging status indication is used to indicate an intuitive indication of the charging status of the battery cell (usually used in the user interface), the charging fault code is used to indicate the type of fault that occurred during the charging cycle, the charging termination reason is used to indicate the specific reason for the end of the charging process, and the battery health status is used to indicate the battery cell charging status. Indicates the health status of the battery cell. Battery service life is used to indicate the time from the current state of the battery cell to the time when its performance drops to the target threshold. Charging power is used to indicate the power during the current charging process. Battery voltage is used to indicate the current voltage of the battery cell. Battery current is used to indicate the current current of the battery cell. Battery charge is used to indicate the remaining capacity of the battery cell. Battery health is used to indicate the health status of the battery cell. Battery temperature change rate is used to indicate the rate of change of the battery cell temperature. Battery voltage change rate is used to indicate the rate of change of the battery cell voltage. Battery current change rate is used to indicate the rate of change of the battery cell voltage. Battery balancing status is used to indicate the balancing status of each battery cell in the battery pack. Battery internal resistance is used to indicate the internal resistance of the battery cell. Battery balancing current is used to indicate the current during the battery cell balancing process. Battery balancing time is used to indicate the total duration of the battery cell balancing process.
[0092] Schematically, the charging data related to voltage is filtered out from the charging data according to the charging status field to obtain the voltage data of the battery cell during the charging cycle, and the charging data related to current is filtered out from the charging data according to the charging status field to obtain the current data of the battery cell during the charging cycle.
[0093] The capacity increment change formula of the battery cell during the charging cycle is determined based on the voltage data and the current data.
[0094] The capacity increment change formula is used to describe the change in the amount of electricity charged or discharged from the battery cell during a charging cycle. Optionally, the capacity increment change formula can be presented in the form of a digital chart.
[0095] In the embodiment of the present application, the battery cell is implemented as any battery in the battery pack. The above voltage data and current data can be used to obtain the capacity increment change formula of the i-th battery cell according to the following formula 1, where i is a positive integer.
[0096]
[0097] In formula 1, t refers to any time point in the charging cycle, a is a preset value, Used to indicate the current data corresponding to the i-th battery cell at time t, Used to indicate the voltage data corresponding to the i-th battery cell at time t.
[0098] In another optional embodiment, a candidate capacity increment variation formula corresponding to the i-th battery cell during the charging cycle is determined based on the current data and the voltage data, wherein the candidate capacity increment variation formula is expressed as the content of Formula 1.
[0099] The candidate capacity increment change formula corresponding to the i-th battery cell is subjected to Gaussian fitting to obtain the capacity increment change formula. The capacity increment change formula can be found in the following formula 2.
[0100]
[0101] In formula 2, Gaussfit(·) refers to the Gaussian fitting function. Please refer to the above formula 1.
[0102] According to the capacity increment variation formula, the half-peak width integral area value and the peak voltage are determined.
[0103] In an optional embodiment, the process of determining the half-height peak width integrated area value is:
[0104] A capacity increment change curve corresponding to the capacity increment change formula is drawn, where the abscissa of the capacity increment change curve is used to indicate time information, and the ordinate is used to indicate the capacity increment value.
[0105] The area enclosed by the capacity increment change curve and the horizontal axis is determined as the half-peak width integrated area value.
[0106] Indicatively, the half-peak width integrated area value can be calculated using the following formula 3.
[0107]
[0108] In formula 3, x l Starting from the peak point, it expands to the left, and the time point when the capacity increment drops to half of the peak value, x m It is the time point when the capacity increment value drops to half of the peak value starting from the peak point and expanding to the right, where the peak point is the highest point corresponding to the same peak. F(x) can be found in the above formula 2.
[0109] In an optional embodiment, the process of determining the peak voltage is:
[0110] Determine the maximum value among the capacity increments; substitute this maximum value into the capacity increment formula to determine the target capacity; and determine the peak voltage as the reciprocal of the target capacity. The specific calculation process can be found in Formulas 4 and 5 below.
[0111] Formula 4: F max =max(F(x));
[0112]
[0113] In Formula 4 and Formula 5, F(x) can be found in Formula 2 above. max is to take the maximum value in F(x), It refers to the peak voltage corresponding to the i-th battery cell at time t.
[0114] Step 210 , inputting the half-peak width integrated area value and the peak voltage into a battery attenuation prediction model to predict the half-peak width integrated area value and the peak voltage predicted value of the battery cell after the charging cycle.
[0115] Optionally, the battery degradation prediction model is used to predict the half-peak width integrated area value and peak voltage of the battery cell at a future time.
[0116] Illustratively, the half-peak width integrated area value and peak voltage of the battery cell at time t are obtained and input into the battery degradation prediction model. The battery degradation prediction model then outputs a predicted half-peak width integrated area value corresponding to the half-peak width integrated area value at time t+1, as well as a predicted peak voltage value corresponding to the peak voltage. It should be noted that time t+1 is a future time, and t+1 can also be realized as t+g, where g is any value greater than 0.
[0117] Step 220 , inputting the predicted value of the half-peak width integrated area value and the predicted value of the peak voltage into a battery capacity and internal resistance prediction model to predict the battery capacity and internal resistance.
[0118] Optionally, the battery capacity and internal resistance prediction model is used to predict the battery capacity and internal resistance of the battery unit at a future time.
[0119] In an embodiment of the present application, the output value of the battery cell predicted by the battery attenuation prediction model at time t is used as the input value of the battery capacity and internal resistance prediction model to obtain the battery internal resistance prediction value and battery capacity prediction value of the battery cell at time t+1.
[0120] In schematic form, the battery loss characteristic values corresponding to the half-peak width integrated area prediction value and the peak voltage prediction value are determined through the battery capacity internal resistance prediction model.
[0121] The battery loss characteristic value is used to describe the loss of battery cells during a charging cycle. The battery loss characteristic value includes capacity attenuation, internal resistance change, current change, voltage change, etc.
[0122] The battery loss characteristic value is input into the battery capacity and internal resistance prediction model for prediction to obtain the battery capacity prediction value and the battery internal resistance prediction value.
[0123] In the embodiment of the present application, the health factor corresponding to the battery loss characteristic value is determined by the battery capacity internal resistance prediction model. The health factor is used to evaluate the health status of the battery cell;
[0124] The health factor mapping is converted into battery capacity prediction value and battery internal resistance prediction value through the battery capacity internal resistance prediction model.
[0125] In an optional embodiment, the battery degradation prediction model is implemented as a bidirectional long short-term memory network multi-task learning model (Bidirectional Long Short-Term Memory Multi-Task Learning Model, referred to as Bi-LSTM multi-task learning model). The bidirectional long short-term memory network multi-task learning model is a deep learning model that combines a bidirectional LSTM network structure and a multi-task learning framework. The model can process multiple related tasks simultaneously and improve the performance and generalization ability of the model by sharing the underlying feature representation.
[0126] In an optional embodiment, the battery capacity and internal resistance model is implemented as an artificial neural network model (ANN model). The ANN model is a computational model that simulates the structure and function of a biological neural network. In the embodiment of the present application, the ANN is used to predict the output of a continuous value.
[0127] Step 230 : Determine the remaining service life of the battery unit based on the battery capacity prediction value and the battery internal resistance prediction value.
[0128] Optionally, a first time period during which the predicted battery capacity decays to a first preset percentage of the initial capacity is determined, and the first time period is determined as the remaining service life. Schematically, a first time period during which the predicted battery capacity decays to 80% of the initial capacity is determined, and the first time period is determined as the remaining service life.
[0129] Optionally, a second time period during which the predicted value of the battery internal resistance decays to a second preset percentage of the initial internal resistance is determined, and the second time period is determined as the remaining service life.
[0130] In an embodiment of the present application, a first time period in which the battery capacity is predicted to decay to a first preset percentage of the initial capacity value is determined, and a second time period in which the battery internal resistance is predicted to decay to a second preset percentage of the initial internal resistance value is determined, an average time period value of the first time period and the second time period is determined, and the average time period value is determined as the remaining service life; alternatively, the maximum value of the first time period and the second time period is taken as the remaining service life; alternatively, the minimum value of the first time period and the second time period is taken as the remaining service life.
[0131] Step 240 : Based on the battery capacity prediction value and the battery internal resistance prediction value, a particle filter algorithm is used to determine a confidence result of the remaining service life.
[0132] Optionally, the predicted value of the battery capacity and the predicted value of the battery internal resistance are used as observation data, and the particle filter algorithm predicts the remaining service life of the battery based on the observation data.
[0133] As batteries are used, their capacity decreases and their internal resistance increases. The particle filter algorithm accurately predicts the remaining useful life of battery cells and their confidence level by continuously updating the states and weights of the particles.
[0134] Schematically, the specific application process of the particle filter algorithm is:
[0135] S1. Define the state model and observation model.
[0136] The state model is a model that characterizes the change of the health status of the battery cell over time and can be represented by a state vector. The observation model is a model that characterizes the relationship between the predicted battery capacity and internal resistance values and the battery health status.
[0137] In another optional embodiment, the empirical model equation is defined using a double exponential function and / or a cubic polynomial, and the empirical model equation is applied as the state model.
[0138] S2. Initialize particles.
[0139] Generate a set of random particles, each representing a possible battery health state. Assign a weight value to each particle, and all particles have equal weight values.
[0140] S3. Prediction step.
[0141] The state of each particle at the next moment is predicted based on the state model, and the predicted observation data of each particle at the next moment is predicted based on the observation model. The predicted observation data includes the predicted value of the battery capacity and the predicted value of the battery internal resistance.
[0142] S4. Update step.
[0143] According to the observed data, the weight of each particle is updated, and the weight is proportional to the likelihood of the observed data.
[0144] Normalize the weights of all particles so that the sum of the weights of all particles is equal to 1.
[0145] S5, resampling step.
[0146] Resample according to the particle weight to generate a new particle set. The higher the weight value, the higher the probability of the particle being selected.
[0147] S6. Calculate the confidence result.
[0148] The remaining service life is updated according to the state information in the new particle set to obtain an estimated value of the remaining service life.
[0149] In another optional embodiment, the remaining service life is directly calculated based on the state information in the new particle set.
[0150] According to the distribution of the remaining service life in the new particle set, the confidence interval of the remaining service life is calculated.
[0151] The remaining service life with a confidence interval higher than a preset value is taken as the remaining service life of the battery cell.
[0152] In an embodiment of the present application, the battery capacity prediction value and battery internal resistance prediction value of the battery cell after the charging cycle are predicted through nested processing between different prediction models; and the remaining service life of the battery cell is determined based on the battery capacity prediction value and the battery internal resistance prediction value; and then the particle filtering algorithm is used to determine the confidence result of the remaining service life. Based on the confidence result, it is determined whether the predicted remaining service life is usable, which further guarantees the accuracy of the remaining service life of the battery cell. In addition, the particle filtering algorithm can dynamically update the parameters involved in the particle filtering algorithm process based on the battery capacity prediction value and battery internal resistance prediction value and current status of the battery cell, thereby realizing real-time prediction and real-time verification of the remaining service life of the battery, and providing timely and accurate information for the maintenance and management of the battery cell.
[0153] Example 2:
[0154] In combination with the content of the above embodiments, the training process of the battery degradation prediction model and the battery capacity internal resistance model is described in detail. Figure 3 As shown, Figure 3 A flowchart of model training provided by an exemplary embodiment of the present application is shown.
[0155] Step 300: data acquisition.
[0156] An initial battery attenuation prediction model and an initial battery capacity internal resistance prediction model are obtained, and historical charging data generated by the battery cell during historical charging cycles is obtained.
[0157] The historical charging data is determined based on the historical timestamp information, historical voltage, historical current, historical state of charge, historical total mileage information, etc. of the historical charging cycles.
[0158] The historical charging data is filtered according to the charging status field to obtain the historical voltage data and historical current data of the battery unit in the historical charging cycle. For the relevant content of the charging status field, please refer to the above embodiment.
[0159] Optionally, based on the historical voltage data and the historical current data, the historical half-peak width integrated area value and the historical peak voltage of the battery cell in the historical charging cycle are calculated using the above formulas 1 to 5.
[0160] Optionally, historical battery capacity and historical internal resistance degradation data of the battery unit during historical charging cycles are determined based on historical voltage data and historical current data.
[0161] The above historical half-peak peak width integrated area value, historical peak voltage, historical battery capacity and historical internal resistance degradation data are determined as historical charging data.
[0162] Step 310: construct a training data set.
[0163] Based on the historical charging data, a training data set is constructed, which includes the historical half-peak width integrated area value, historical peak voltage, historical battery capacity and historical internal resistance degradation data of the battery cell.
[0164] In an optional embodiment, when the battery cell is used as any battery in a battery pack (including n battery cells, n is a positive integer), the historical half-peak width integrated area value, historical peak voltage, historical battery capacity and historical internal resistance degradation data of each battery cell are constructed into a training data set for training model j. For the i-th battery cell, the training data set can be implemented as the contents of the following formulas 6 to 9.
[0165] It should be noted that the embodiments of the present application involve two models: the initial battery attenuation prediction model and the initial battery capacity internal resistance prediction model. Therefore, the value of j is 2. When j is 1, a training data set for the initial battery attenuation prediction model is constructed. When j is 2, a training data set for the initial battery capacity internal resistance prediction model is constructed.
[0166]
[0167] In formulas 6 to 9, Used to indicate the historical half-peak peak width integrated area value set, Used to indicate the historical peak voltage collection, Used to indicate historical battery capacity collection, It is used to indicate the historical internal resistance degradation data set, i is used to indicate the i-th battery cell, j is used to indicate the j-th model, and k is used to indicate the sliding window input and output length of the model (the model here refers to the initial battery degradation prediction model and the initial battery capacity internal resistance prediction model). For different models, k takes different data, and all satisfy (m+2)k is less than or equal to the remaining service life of the battery cell, and m is an integer greater than or equal to 0.
[0168] In another optional embodiment, the training data of n battery cells are reconstructed and connected to obtain a complete training data set.
[0169] For illustration, an example is given where n is 6. For a specific data set after splicing, refer to the following formulas 10 to 13.
[0170]
[0171] The parameters in Formula 10 to Formula 13 have the same meanings as those in Formula 6 to Formula 9 above and will not be repeated here.
[0172] Step 320: model training.
[0173] Optionally, the complete training data set is used to train the initial battery degradation prediction model, wherein the model input format of the initial battery degradation prediction model is (k, 1).
[0174] In response to the convergence of the first model parameter of the initial battery degradation prediction model, a battery degradation prediction model is obtained; or, in response to the number of iterations of the initial battery degradation prediction model reaching a preset requirement, training of the initial battery degradation prediction model is stopped to obtain a battery degradation prediction model.
[0175] The training data set is used to train the initial battery capacity and internal resistance prediction model, wherein the model input format of the initial battery capacity and internal resistance prediction model is (k, 1).
[0176] In response to the convergence of the second model parameter of the initial battery capacity internal resistance prediction model, a battery capacity internal resistance prediction model is obtained; or, in response to the number of iterations of the initial battery capacity internal resistance prediction model reaching a preset requirement, training the initial battery capacity internal resistance prediction model is stopped to obtain a battery capacity internal resistance prediction model.
[0177] In an embodiment of the present application, the battery capacity prediction value and battery internal resistance prediction value of the battery cell after the charging cycle are predicted through nested processing between different prediction models; and the remaining service life of the battery cell is determined based on the battery capacity prediction value and the battery internal resistance prediction value; and then the particle filtering algorithm is used to determine the confidence result of the remaining service life. Based on the confidence result, it is determined whether the predicted remaining service life is usable, which further guarantees the accuracy of the remaining service life of the battery cell. In addition, the particle filtering algorithm can dynamically update the parameters involved in the particle filtering algorithm process based on the battery capacity prediction value and battery internal resistance prediction value and current status of the battery cell, thereby realizing real-time prediction and real-time verification of the remaining service life of the battery, and providing timely and accurate information for the maintenance and management of the battery cell.
[0178] Figure 4 A structural block diagram of a battery life prediction device provided by an exemplary embodiment of the present application is shown. The device includes: a determination module 400 and an input module 410.
[0179] a determination module 400 configured to determine a half-height peak width integrated area value and a peak voltage of a battery cell during a charging cycle, wherein the half-height peak width integrated area value is used to describe a change in battery capacity of the battery cell during the charging cycle, and the peak voltage refers to a maximum voltage of the battery cell during the charging cycle;
[0180] An input module 410 is configured to input the half-height peak width integrated area value and the peak voltage into a battery degradation prediction model to predict a half-height peak width integrated area value and a peak voltage predicted value of the battery cell after the charging cycle;
[0181] The input module 410 is further configured to input the half-height peak width integrated area prediction value and the peak voltage prediction value into a battery capacity and internal resistance prediction model to predict a battery capacity prediction value and a battery internal resistance prediction value;
[0182] The determination module 400 is further configured to determine the remaining service life of the battery cell based on the battery capacity prediction value and the battery internal resistance prediction value;
[0183] The determination module 400 is further configured to determine a confidence result of the remaining useful life using a particle filter algorithm based on the battery capacity prediction value and the battery internal resistance prediction value, wherein the confidence result is used to evaluate the accuracy of the remaining useful life;
[0184] Among them, the battery attenuation prediction model and the battery capacity internal resistance prediction model are pre-trained machine learning models.
[0185] In an optional embodiment, the determination module 400 is further configured to determine, using the battery capacity internal resistance prediction model, a battery loss characteristic value corresponding to the half-peak width integrated area prediction value and the peak voltage prediction value, wherein the battery loss characteristic value is used to describe the loss of the battery cell during the charging cycle;
[0186] The input module 410 is further configured to input the battery loss characteristic value into the battery capacity and internal resistance prediction model for prediction, thereby obtaining the battery capacity prediction value and the battery internal resistance prediction value.
[0187] In an optional embodiment, if Figure 5 As shown, the determination module 400 is further used to determine the health factor corresponding to the battery loss characteristic value through the battery capacity internal resistance prediction model, and the health factor is used to evaluate the health status of the battery unit;
[0188] The conversion module 420 is configured to convert the health factor mapping into the battery capacity prediction value and the battery internal resistance prediction value through the battery capacity internal resistance prediction model.
[0189] In an optional embodiment, if Figure 5 As shown, the acquisition module 430 is used to obtain the charging data generated by the battery unit during the charging cycle;
[0190] a screening module 440, configured to screen the charging data according to the charging status field to obtain voltage data and current data of the battery cell during the charging cycle;
[0191] The determining module 400 is further configured to determine a capacity increment variation formula of the battery cell during the charging cycle based on the voltage data and the current data;
[0192] The determination module 400 is further configured to determine the half-peak width integrated area value and the peak voltage according to the capacity increment variation formula.
[0193] In an optional embodiment, if Figure 5 As shown, the drawing module 450 is used to draw a capacity increment change curve corresponding to the capacity increment change formula;
[0194] The determining module 400 is further configured to determine the area enclosed by the capacity increment change curve and the horizontal axis as the half-peak width integrated area value.
[0195] In an optional embodiment, if Figure 5 As shown, the determination module 400 is further configured to determine the maximum value in the capacity increment data;
[0196] The determination module 400 is further configured to substitute the maximum value into the capacity increment variation formula to determine a target capacity value;
[0197] The determining module 400 is further configured to determine the inverse of the target capacity value as the peak voltage.
[0198] In an optional embodiment, if Figure 5 As shown, the acquisition module 430 is further used to acquire an initial battery attenuation prediction model and an initial battery capacity internal resistance prediction model;
[0199] The acquisition module 430 is further configured to acquire historical charging data generated by the battery unit during a historical charging cycle;
[0200] A construction module 460 is configured to construct a training data set based on the historical charging data, wherein the training data set includes historical half-peak width integrated area values, historical peak voltages, historical battery capacities, and historical internal resistance degradation data of the battery cells;
[0201] A training module 470 is configured to train the initial battery degradation prediction model using the training data set, and obtain the battery degradation prediction model in response to convergence of a first model parameter of the initial battery degradation prediction model;
[0202] The training module 470 is configured to train the initial battery capacity internal resistance prediction model using the training data set, and obtain the battery capacity internal resistance prediction model in response to the second model parameter of the initial battery capacity internal resistance prediction model converging.
[0203] In an embodiment of the present application, the battery capacity prediction value and battery internal resistance prediction value of the battery cell after the charging cycle are predicted through nested processing between different prediction models; and the remaining service life of the battery cell is determined based on the battery capacity prediction value and the battery internal resistance prediction value; and then the particle filtering algorithm is used to determine the confidence result of the remaining service life. Based on the confidence result, it is determined whether the predicted remaining service life is usable, which further guarantees the accuracy of the remaining service life of the battery cell. In addition, the particle filtering algorithm can dynamically update the parameters involved in the particle filtering algorithm process based on the battery capacity prediction value and battery internal resistance prediction value and current status of the battery cell, thereby realizing real-time prediction and real-time verification of the remaining service life of the battery, and providing timely and accurate information for the maintenance and management of the battery cell.
[0204] Figure 6The following is a block diagram of a computer device 600 provided in accordance with an exemplary embodiment of the present application. The computer device 600 may be a portable mobile terminal, such as a smartphone, a tablet computer, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer, or a desktop computer. The computer device 600 may also be referred to as a user device, a portable terminal, a laptop terminal, a desktop terminal, or other similar names. Optionally, the computer device 600 may also be implemented as a mobile device, such as a mobile smart terminal such as an in-vehicle terminal.
[0205] Typically, the computer device 600 includes a processor 601 and a memory 602 .
[0206] The processor 601 may include one or more processing cores, such as a 4-core processor, a 6-core processor, etc. The processor 601 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), or PLA (Programmable Logic Array). The processor 601 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 601 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 601 may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.
[0207] The memory 602 may include one or more computer-readable storage media, which may be non-transitory. The memory 602 may also include a high-speed random access memory and a non-volatile memory, such as one or more disk storage devices and flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 602 is used to store at least one instruction, which is used to be executed by the processor 601 to implement the model training method or behavior coding method provided in the method embodiment of the present application.
[0208] Those skilled in the art will understand that Figure 6 The structure shown in the figure does not constitute a limitation on the computer device 600, and the computer device 600 may include more or fewer components than shown in the figure, or combine some components, or adopt a different arrangement of components.
[0209] The present application also provides a computer-readable storage medium, which stores at least one instruction, at least one program, code set or instruction set. The at least one instruction, the at least one program, the code set or instruction set is loaded and executed by a processor to implement the battery life prediction method provided by the above method embodiment.
[0210] The present application provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the battery life prediction method provided in the above method embodiment.
[0211] Those skilled in the art will appreciate that all or part of the steps in the above embodiments may be implemented by hardware or by programs instructing the relevant hardware to perform the steps. The programs may be stored in a computer-readable storage medium, such as a read-only memory, a magnetic disk, or an optical disk. The above are merely optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included within the scope of protection of the present application.
Claims
1. A battery life prediction method, characterized in that: The method comprises: Determine a half-height peak width integrated area value and a peak voltage of a battery cell during a charging cycle, wherein the half-height peak width integrated area value is used to describe a change in battery capacity of the battery cell during the charging cycle, and the peak voltage refers to the maximum voltage of the battery cell during the charging cycle; Inputting the half-height peak width integrated area value and the peak voltage into a battery attenuation prediction model to predict a half-height peak width integrated area value prediction value and a peak voltage prediction value of the battery cell after the charging cycle; Inputting the half-height peak width integrated area prediction value and the peak voltage prediction value into a battery capacity and internal resistance prediction model to predict a battery capacity prediction value and a battery internal resistance prediction value; determining a remaining service life of the battery cell based on the battery capacity prediction value and the battery internal resistance prediction value; Based on the battery capacity prediction value and the battery internal resistance prediction value, a particle filter algorithm is used to determine a confidence result of the remaining service life, where the confidence result is used to evaluate the accuracy of the remaining service life; Among them, the battery attenuation prediction model and the battery capacity internal resistance prediction model are pre-trained machine learning models.
2. The method according to claim 1, characterized in that The step of inputting the half-height peak width integrated area prediction value and the peak voltage prediction value into a battery capacity and internal resistance prediction model to predict a battery capacity prediction value and a battery internal resistance prediction value includes: Determine, by using the battery capacity internal resistance prediction model, a battery loss characteristic value corresponding to the half-peak width integrated area prediction value and the peak voltage prediction value, wherein the battery loss characteristic value is used to describe the loss of the battery unit during the charging cycle; The battery loss characteristic value is input into the battery capacity and internal resistance prediction model for prediction to obtain the battery capacity prediction value and the battery internal resistance prediction value.
3. The method according to claim 2, characterized in that The step of inputting the battery loss characteristic value into the battery capacity and internal resistance prediction model for prediction to obtain the battery capacity prediction value and the battery internal resistance prediction value includes: Determine a health factor corresponding to the battery loss characteristic value using the battery capacity internal resistance prediction model, wherein the health factor is used to evaluate the health status of the battery cell; The health factor mapping is converted into the battery capacity prediction value and the battery internal resistance prediction value through the battery capacity internal resistance prediction model.
4. The method according to any one of claims 1 to 3, characterized in that: Determining the half-peak width integrated area value and peak voltage of the battery cell during the charging cycle includes: Acquiring charging data generated by the battery unit during the charging cycle; Filtering the charging data according to the charging status field to obtain voltage data and current data of the battery unit during the charging cycle; determining a capacity increment variation formula of the battery unit during the charging cycle based on the voltage data and the current data; According to the capacity increment variation formula, the half-peak width integrated area value and the peak voltage are determined.
5. The method according to claim 4, characterized in that Determining the half-height peak width integrated area value according to the capacity increment variation formula includes: Draw a capacity increment change curve corresponding to the capacity increment change formula; The area enclosed by the capacity increment change curve and the horizontal axis is determined as the half-peak width integrated area value.
6. The method according to claim 4, characterized in that The determining the peak voltage according to the capacity increment variation formula includes: Determine the maximum value in the capacity increment data; Substituting the maximum value into the capacity increment variation formula to determine a target capacity value; The reciprocal of the target capacity value is determined as the peak voltage.
7. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Obtaining an initial battery attenuation prediction model and an initial battery capacity internal resistance prediction model; Acquiring historical charging data generated by the battery unit during a historical charging cycle; Based on the historical charging data, a training data set is constructed, wherein the training data set includes historical half-peak width integrated area values, historical peak voltages, historical battery capacities, and historical internal resistance degradation data of the battery cells; Training the initial battery degradation prediction model using the training data set, and obtaining the battery degradation prediction model in response to convergence of a first model parameter of the initial battery degradation prediction model; The initial battery capacity internal resistance prediction model is trained using the training data set, and in response to the second model parameter of the initial battery capacity internal resistance prediction model converging, the battery capacity internal resistance prediction model is obtained.
8. A battery life prediction device, characterized in that: The device comprises: a determination module, configured to determine a half-height peak width integrated area value and a peak voltage of a battery cell during a charging cycle, wherein the half-height peak width integrated area value is used to describe a change in battery capacity of the battery cell during the charging cycle, and the peak voltage refers to a maximum voltage of the battery cell during the charging cycle; An input module, configured to input the half-height peak width integrated area value and the peak voltage into a battery attenuation prediction model to predict a predicted value of the half-height peak width integrated area value and a predicted value of the peak voltage of the battery cell after the charging cycle; The input module is further used to input the half-peak width integrated area prediction value and the peak voltage prediction value into the battery capacity and internal resistance prediction model to predict the battery capacity prediction value and the battery internal resistance prediction value; The determination module is further configured to determine the remaining service life of the battery unit based on the battery capacity prediction value and the battery internal resistance prediction value; The determination module is further configured to determine a confidence result of the remaining useful life using a particle filter algorithm based on the battery capacity prediction value and the battery internal resistance prediction value, wherein the confidence result is used to evaluate the accuracy of the remaining useful life; Among them, the battery attenuation prediction model and the battery capacity internal resistance prediction model are pre-trained machine learning models.
9. A computer-readable storage medium, characterized in that The computer device readable storage medium stores a computer program, and the computer program is loaded and executed by a processor to implement the battery life prediction method according to any one of claims 1 to 7.
10. A computer program product or a computer program, characterized in that The computer program product or computer program includes computer instructions, which are stored in a computer-readable storage medium. The processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to implement the battery life prediction method as described in any one of claims 1 to 7.