A data analysis-based energy storage system health assessment and optimization method and system

Abstract

The application relates to the technical field of energy storage measurement and evaluation, and provides an energy storage system health evaluation and optimization method and system based on data analysis. Historical data and working condition data collected by an industrial sensor are acquired by an intelligent sensor, preprocessed to generate first data, a lithium ion concentration diffusion model and a heat conduction model of an energy storage system are constructed, and a coupling feature tensor and stress distribution are jointly generated; a neural network based on a multi-head attention mechanism is trained on the historical data, battery parameters and statistical distribution representing battery health are determined through the neural network; and finally, a health report and a control instruction are generated to control the operation of the energy storage system. The coupling feature tensor is generated based on the ion concentration diffusion model and the heat conduction model, the internal state which cannot be directly measured is converted into quantifiable features, material-level damage diagnosis and early warning are realized, the accuracy of energy storage system pre-diagnosis is improved, and the efficiency of energy storage system health management is enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of energy storage measurement evaluation, in particular to an energy storage system health evaluation and optimization method and system based on data analysis. BACKGROUND

[0002] With the increasing proportion of intermittent energy sources such as wind power and photovoltaic power, the demand for energy storage systems in power grids has surged. In high-intensity application scenarios such as participating in power grid frequency modulation and new energy consumption, traditional methods are difficult to accurately quantify the internal material-level damage of batteries, such as electrode lithium precipitation and particle cracking, and static control strategies cannot balance the contradiction between power grid response benefits and battery life loss, resulting in serious challenges such as safety risks and economic benefits of the system. Therefore, the existing technology has the problem of health misjudgment caused by invisible battery material-level damage, resulting in low accuracy of pre-diagnosis and health management of energy storage systems. SUMMARY

[0003] The present application provides an energy storage system health evaluation and optimization method and system based on data analysis, which can at least partially solve the problem of low accuracy of pre-diagnosis and health management of energy storage systems.

[0004] Other characteristics and advantages of the present application will become apparent from the following detailed description, or will be learned by practice of the present application.

[0005] According to one aspect of the present application, an energy storage system health evaluation and optimization method based on data analysis is provided, comprising: acquiring historical data and working condition data collected by an industrial sensor through an intelligent sensor, and preprocessing the historical data and the working condition data to generate first data; constructing a lithium ion concentration diffusion model and a thermal conduction model of an energy storage system based on the first data, combining the concentration diffusion model and the thermal conduction model to generate a coupling feature tensor and a stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system; training a neural network based on a multi-head attention mechanism through the historical data, inputting the coupling feature tensor into the neural network, and determining battery parameters and statistical distribution representing battery health; generating a health report of the energy storage system based on the stress distribution, the battery parameters and statistical distribution, and generating a control instruction based on the health report to control the operation of the energy storage system; updating the health report of the energy storage system in the process of executing the control instruction.

[0006] In the present application, based on the foregoing scheme, the preprocessing of the historical data and the working condition data to generate first data includes: aligning the historical data and working condition data from different sampling rates through time stamp to generate second data; performing adaptive filtering on the current signal in the second data and information enhancement on the temperature data in the second data to generate first data.

[0007] In the present application, based on the foregoing scheme, the construction of the concentration diffusion model of lithium ions and the heat conduction model of the energy storage system based on the first data, and the generation of the coupling characteristic tensor and the stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system based on the concentration diffusion model and the heat conduction model, includes: constructing a concentration diffusion model of lithium ions in the energy storage system based on the real-time diffusion coefficient of the lithium ion concentration in the first data and the energy storage material; constructing a heat conduction model of the energy storage system based on the reaction data in the first data and the material data of the energy storage material; combining the concentration diffusion model and the heat conduction model to generate a coupling characteristic tensor representing the relationship between the material volume and the electrolyte concentration of the energy storage system; decomposing the coupling characteristic tensor by finite element method to obtain a four-dimensional stress distribution.

[0008] In the present application, based on the foregoing scheme, the training of the neural network based on the multi-head attention mechanism by the historical data, the input of the coupling characteristic tensor into the neural network, and the determination of the battery parameters and statistical distribution representing the battery health, includes: training a neural network based on a multi-head attention mechanism based on the historical data; inputting the coupling characteristic tensor into the neural network to obtain output features and calculating the partial derivative of the output features of the neural network with respect to the input features; determining the battery parameters representing the battery health based on the output features and the partial derivative; performing K times forward propagation of the coupling characteristic tensor in the neural network to generate a statistical distribution of the battery parameters.

[0009] In the present application, based on the foregoing scheme, the generation of the health report of the energy storage system based on the stress distribution, the battery parameters and the statistical distribution, and the generation of the control instruction based on the health report to control the operation of the energy storage system, includes: generating a health report of the energy storage system based on the stress distribution, the battery parameters and the statistical distribution; extracting a risk label from the health report and mapping the risk label to a preset control strategy library to obtain a control instruction corresponding to the risk label; controlling the operation of the energy storage system through the control instruction.

[0010] In the present application, based on the foregoing scheme, the updating the health report of the energy storage system in the process of executing the control instruction comprises: acquiring real-time data based on a preset time period in the process of executing the control instruction; analyzing the real-time data, and updating the health report of the energy storage system according to the analysis result.

[0011] In the present application, based on the foregoing scheme, further comprising: generating a failure strength function of the energy storage system based on stress distribution, temperature and lithium ion concentration; generating a service life probability of the energy storage system based on the failure strength function, taking the service life probability as one of the battery parameters.

[0012] According to one aspect of the present application, a data analysis-based energy storage system health evaluation and optimization system is provided, comprising:

[0013] An acquisition unit is configured to acquire historical data and working condition data collected by an industrial sensor through an intelligent sensor, and pre-process the historical data and the working condition data to generate first data;

[0014] A model unit is configured to construct a lithium ion concentration diffusion model and a thermal conduction model of the energy storage system based on the first data, and generate a coupling characteristic tensor representing the relationship between the material volume of the energy storage system and the electrolyte concentration and a stress distribution by combining the concentration diffusion model and the thermal conduction model;

[0015] A distribution unit is configured to train a neural network based on a multi-head attention mechanism through the historical data, input the coupling characteristic tensor into the neural network, and determine a battery parameter representing battery health and a statistical distribution;

[0016] A control unit is configured to generate a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generate a control instruction based on the health report to control the operation of the energy storage system;

[0017] An execution unit is configured to update the health report of the energy storage system in the process of executing the control instruction.

[0018] According to one aspect of the present application, a computer readable medium having a computer program stored thereon is provided, the computer program being executed by a processor to implement the data analysis-based energy storage system health evaluation and optimization method as described in the above embodiments.

[0019] According to an aspect of the present application, an electronic device is provided, comprising: one or more processors; a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the data analysis-based energy storage system health assessment and optimization method as described in the above embodiments.

[0020] According to an aspect of the present application, a computer program product or computer program is provided, comprising 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 the processor executes the computer instructions, so that the computer device executes the data analysis-based energy storage system health assessment and optimization method provided in the various optional implementations described above.

[0021] The technical scheme of the present application acquires historical data and working condition data collected by an industrial sensor through an intelligent sensor, and pre-processes the historical data and the working condition data to generate first data; constructs a lithium ion concentration diffusion model and a heat conduction model of an energy storage system based on the first data, and generates a coupling feature tensor and a stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system by combining the concentration diffusion model and the heat conduction model; trains a neural network based on a multi-head attention mechanism through the historical data, inputs the coupling feature tensor into the neural network, and determines a battery parameter and a statistical distribution representing battery health; generates a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generates a control instruction based on the health report to control the operation of the energy storage system; and updates the health report of the energy storage system in the process of executing the control instruction. The coupling feature tensor is generated based on the ion concentration diffusion model and the heat conduction model, the internal state that cannot be directly measured is converted into quantifiable features, material-level damage diagnosis and early warning are realized, the accuracy of pre-diagnosis of the energy storage system is improved, and the efficiency of health management of the energy storage system is enhanced.

[0022] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0023] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0024] Figure 1A flowchart of a data analysis based energy storage system health assessment and optimization method in one embodiment of the present application is schematically shown.

[0025] Figure 2 A flowchart of determining coupling feature tensors and stress distribution in one embodiment of the present application is schematically shown.

[0026] Figure 3 A schematic diagram of a data analysis based energy storage system health assessment and optimization system in one embodiment of the present application is schematically shown.

[0027] Figure 4 A structural schematic diagram of a computer system of an electronic device suitable for implementing embodiments of the present application is shown. DETAILED DESCRIPTION

[0028] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations may, however, be implemented in many different forms and should not be construed as limited to the implementations set forth herein; rather, these implementations are provided so that this disclosure will be thorough and complete, and will fully convey the inventive aspects to those skilled in the art. Like reference numerals may refer to like elements throughout.

[0029] Moreover, described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the application. One skilled in the relevant art will recognize, however, that the

[0030] The block diagrams in the drawings show only the functionality of the embodiments and do not imply any particular physical or architectural arrangement of the devices, systems, or methods. No inference should be drawn regarding the implementational aspects of the embodiments from the description of the features or structural elements of the drawings. Further, any described feature, structure, or characteristic can be combined in any suitable manner in one or more embodiments.

[0031] The flow diagrams depicted herein are merely illustrative examples, and are not necessarily meant to imply a fixed order of operations, or that all of the operations or steps are required, or that the order of the operations or steps are the same. For example, as would be recognized by one of ordinary skill in the art, elements of the diagrams and / or flowchart can, in some embodiments, be performed in a different order, or omitted, or combined with other steps or elements into a new step or element, or partitioned into multiple steps or elements.

[0032] The implementation details of the technical solutions of the present application are described in detail as follows:

[0033] Figure 1A flowchart of a data analysis based energy storage system health assessment and optimization method according to an embodiment of the present application is shown. Referring to Figure 1 As shown, the data analysis based energy storage system health assessment and optimization method includes at least steps S110 to S150, which are described in detail as follows:

[0034] S110, obtaining historical data and working condition data collected by an intelligent sensor, and preprocessing the historical data and the working condition data to generate first data.

[0035] In this embodiment, the database integrated with the whole life cycle of the energy storage system and the real-time deployed industrial sensor network obtain multi-dimensional historical data of battery operation in the energy storage system through the database, and obtain current working condition data through the industrial sensor network.

[0036] Specifically, the database composed of historical data includes electrical parameters such as voltage, current and temperature at different stages, in addition to degradation trajectories such as mechanical deformation and internal resistance change. The industrial sensor network obtains electrical signal waveforms, three-dimensional temperature field distribution and structural vibration state of the energy storage system in the charging and discharging process in real time.

[0037] Optionally, after obtaining the historical data and the working condition data, an original data pool is constructed to provide a basis information source covering three dimensions of time, space and state for subsequent physical feature extraction, ensuring that the health assessment of the energy storage system has sufficient physical driving basis.

[0038] Specifically, in an embodiment of the present application, preprocessing the historical data and the working condition data to generate first data includes:

[0039] Aligning the historical data and working condition data from different sampling rates through time stamp to generate second data;

[0040] Adaptive filtering the current signal in the second data and information enhancement on the temperature data in the second data to generate first data.

[0041] In this embodiment, the time labels of the historical data and the working condition data are aligned through a unified clock reference, i.e. time stamp, to solve the data misplacement problem caused by the difference in sensor sampling frequency.

[0042] For example, in the charging and discharging process of the battery of the energy storage system, the high-frequency collected electrical signal and the low-frequency recorded temperature field are re-matched according to the real time sequence of physical events to generate a strictly synchronized second data set. This ensures the accurate correspondence of the cause-effect relationship of key events such as current mutation and temperature rise in subsequent analysis, laying a time sequence foundation for multi-physical field coupling modeling.

[0043] In this embodiment, the adaptive filtering technology is used to dynamically purify the current signal in the second data. Specifically, the pulse interference in the charge-discharge switching stage and the background noise in the steady state operation are identified first, and then the filtering strength parameter is adjusted in real time. The current signal is filtered according to the real-time filtering strength parameter. On the premise of maintaining the characteristics of rapid current change, the high-frequency oscillation and zero drift error introduced by the measurement device are effectively suppressed. The smooth and physically real current signal is output as the first data to improve the numerical stability in the subsequent electrochemical model processing.

[0044] For the temperature data in the second data, in this embodiment, the temperature data collected by the temperature sensor is data augmented and information enhanced based on the heat conduction law. Specifically, the temperature distribution of the region where the sensor is not arranged is deduced through the three-dimensional heat diffusion equation, and the virtual temperature field conforming to the principle of energy conservation is generated by combining the material heat capacity parameters of the energy storage system and the cooling system layout, so as to realize the augmentation and enhancement of the temperature data. In this way, the hardware monitoring blind area is compensated for, and the data collected by the limited industrial sensor is converted into a complete spatial thermal state map.

[0045] The above process synchronously collects historical data and real-time working condition parameters through industrial sensors or intelligent sensors, solves the sampling rate difference problem through time stamp alignment, ensures the time sequence consistency of the electro-thermal force multi-physical field data, eliminates the current signal noise through filtering, and completes the temperature field spatial distribution through the physical information enhancement technology, thereby generating high-fidelity first data. The finally output first data contains high-confidence global temperature gradient information, which provides complete input for thermal stress coupling analysis.

[0046] S120, constructing a concentration diffusion model of lithium ions and a heat conduction model of the energy storage system based on the first data, and generating a coupling characteristic tensor and a stress distribution representing the relationship between the material volume of the energy storage system and the electrolyte concentration by combining the concentration diffusion model and the heat conduction model.

[0047] As shown in Figure 2 In one embodiment of the present application, the concentration diffusion model of lithium ions and the heat conduction model of the energy storage system are constructed based on the first data, and the coupling characteristic tensor and the stress distribution representing the relationship between the material volume of the energy storage system and the electrolyte concentration are generated by combining the concentration diffusion model and the heat conduction model, including:

[0048] S210, constructing a concentration diffusion model of lithium ions in the energy storage system based on the lithium ion concentration in the first data and the real-time diffusion coefficient of the energy storage material;

[0049] S220, constructing a heat conduction model of the energy storage system based on the reaction data in the first data and the material data of the energy storage material;

[0050] S230, generating a coupling characteristic tensor representing the relationship between the material volume of the energy storage system and the electrolyte concentration based on the concentration diffusion model and the heat conduction model;

[0051] S240, performing finite element method decomposition on the coupling characteristic tensor to obtain a four-dimensional stress distribution.

[0052] In practical applications, Joule heat and reaction heat are generated when current passes through the battery, causing temperature rise; temperature rise accelerates ion diffusion, but overheating can trigger side reactions, so the two are bidirectionally coupled through concentration diffusion and heat conduction.

[0053] At the same time, in practical applications, the size of commercial sensors is generally greater than 100 microns, while the ion concentration gradient occurs in the micron-level electrode pores. Moreover, the plug-in sensor will damage the electrolyte sealing environment and trigger side reactions such as electrolyte decomposition. Even if a microelectrode can be locally measured, a laboratory tried to implant a micro ion concentration sensor in the electrolyte, but the result was leakage due to sealing failure, resulting in a 57% decrease in battery cycle life, and the local heating of the sensor made the measured value deviate from the true value by 38%. Therefore, it is difficult to directly measure the ion concentration in practical applications. In this embodiment, the sensor nodes deployed outside the energy storage system are used to obtain data, construct the coupling characteristic tensor and the stress distribution, determine the battery parameters and statistical distribution representing the battery health, and realize the overall pre-diagnosis and health management of the energy storage system.

[0054] In an embodiment of the present application, based on the lithium ion concentration in the first data and the real-time diffusion coefficient of the energy storage material, a concentration diffusion model of lithium ions in the energy storage system is constructed as follows:

[0055]

[0056] wherein, represents the lithium ion concentration of the electrolyte in the energy storage system, t represents time, represents a gradient operation;

[0057] wherein, represents the real-time diffusion coefficient of the energy storage material; specifically, represents the real-time diffusion coefficient of the energy storage material measured by experiment, represents the activation energy of the energy storage material, r represents the gas constant per unit energy conversion, T represents the real-time temperature monitored by the temperature sensor;

[0058] wherein, j represents the local current density, f represents the Faraday constant. In this embodiment, the concentration diffusion model of lithium ions reflects the influence of current on the electrolyte concentration distribution.

[0059] In an embodiment of the present application, based on the reaction data in the first data and the material data of the energy storage material, a heat conduction model of the energy storage system is constructed as:

[0060]

[0061] wherein, represents the material density of the energy storage material, represents the specific heat capacity of the energy storage material, T represents the real-time temperature, t represents time; k represents the thermal conductivity of the energy storage material;

[0062] wherein, I represents the output current of the energy storage system, R represents the real-time resistance of the energy storage system;

[0063] wherein, represents the side reaction heat of the energy storage system, represents the reaction coefficient of the energy storage system, represents the overpotential. The heat conduction model in the embodiment reflects the conduction process of the current heat and the reaction heat.

[0064] In an embodiment of the present application, the concentration diffusion model reflects the influence of the current on the electrolyte concentration distribution, and the heat conduction model reflects the conduction process of the current heat and the reaction heat. The mutual influence of the internal electrochemical reaction and the temperature field of the battery is simulated by the above two models, and key physical characteristics are provided for subsequent health state assessment.

[0065] The concentration diffusion model and the heat conduction model are solved simultaneously, and the equation group composed of the above two equations is solved to obtain the local current density j and the real-time lithium ion concentration of the electrolyte in the energy storage system , which are respectively adapted to the heat flow distribution and the ion concentration field in the temperature control process of the energy storage system, so as to quantify the promotion effect of high temperature on the ion diffusion rate, and predict the hot spot position in the battery of the energy storage system, such as the temperature rise of 5℃ at the edge of the electrode of the battery.

[0066] In actual application, the volume expansion will be caused when the lithium ion is embedded into the electrode material, such as the expansion rate of 8% of the graphite negative electrode. The thermal expansion and contraction of the material caused by temperature change will also be caused. The mechanical stress caused by the superposition of the two will possibly cause the breakage of the electrode particles. In the embodiment, the real-time lithium ion concentration of the electrolyte in the energy storage system is directly inferred to obtain the stress distribution inside the electrode, so as to capture the volume change caused by the embedding or disembedding of the lithium ion into the energy storage system, and based on the lithium ion concentration , a coupling characteristic tensor characterizing the relationship between the material volume and the electrolyte concentration of the energy storage system is generatedF is:

[0067]

[0068] wherein, represents the material expansion coefficient of the energy storage system. Through the above process, the volume corresponding to the coupling characteristic tensor is driven by the lithium ion concentration gradient, and the greater the concentration, the stronger the expansion force. To quantify the mechanical stress change in the battery cycle process, predict the fatigue and aging of the battery caused by stress.

[0069] The finite element decomposition is performed on the coupling characteristic tensor, and the four-dimensional stress distribution is obtained by inverting the stress balance equation as Through the four-dimensional stress distribution, the target area of high stress such as the electrode edge can be located, and the particle crack can also be predicted through the stress distribution, for example, when the stress is greater than 50 MPa, the crack risk increases by 80%.

[0070] The above process quantifies the electrode reaction kinetics based on the ion concentration diffusion model, and analyzes the temperature gradient distribution combined with the heat conduction model; the two field coupling characteristic tensors are decomposed into four-dimensional stress field through the finite element method, and the space-time evolution of the internal stress of the material in the charging and discharging process is revealed. Through multi-physical field coupling modeling, the aging mechanism of the battery is comprehensively described from the three dimensions of electrochemistry, heat and mechanics, and the problem that a single physical field model cannot accurately describe the complex aging process is solved. The mapping from macroscopic operating parameters to microscopic material damage is realized, such as locating the particle crack risk area of the electrode, so that the health assessment has the material-level failure warning capability.

[0071] S130, train a neural network based on a multi-head attention mechanism through the historical data, input the coupling characteristic tensor into the neural network, and determine the battery parameter and statistical distribution representing the battery health.

[0072] In an embodiment of the present application, the neural network based on the multi-head attention mechanism is trained through the historical data, the coupling characteristic tensor is input into the neural network, and the battery parameter and statistical distribution representing the battery health are determined, including:

[0073] Based on the historical data, train a neural network based on a multi-head attention mechanism;

[0074] Input the coupling characteristic tensor into the neural network to obtain output features, and calculate the partial derivative of the output features of the neural network with respect to the input features;

[0075] Based on the output features and the partial derivative, determine the battery parameter representing the battery health;

[0076] In the neural network, the coupling characteristic tensor is KThe next forward propagation generates the statistical distribution of battery parameters.

[0077] In this embodiment, based on historical data, a neural network based on a multi-head attention mechanism is trained using a convolutional neural network; the coupled feature tensor is input into the neural network to obtain the output features, and the partial derivative of the output features of the neural network with respect to the input features is calculated. Based on the neural network and the partial derivatives, battery parameters characterizing battery health are determined. S for:

[0078]

[0079] in, This refers to a neural network based on a multi-head self-attention mechanism. Indicates output features, This represents the regularization strength coefficient. i and N These represent the identifier and total number of input features, respectively. In this embodiment, the calculated battery parameters are used to characterize the difference between the current battery capacity and the initial capacity, thus reflecting the battery's health status.

[0080] In the trained neural network The middle retains a random deactivation layer, and performs operations on the input features corresponding to the same coupled feature tensor. K The first forward propagation (e.g., K=100) determines the statistical distribution of the battery parameters. Specifically, the statistical distribution includes the first statistical parameter. for:

[0081]

[0082] in, Indicates the first k Battery parameters generated by the second forward propagation k Indicates the number of forward propagations. K This indicates the total number of forward propagations.

[0083] Then, based on the calculated first statistical parameter, the second statistical parameter belonging to the statistical distribution is determined. for:

[0084]

[0085] Then, the statistical distribution corresponding to the battery parameters is determined based on the first statistical parameter and the second statistical parameter. for:

[0086]

[0087] After the statistical distribution is calculated, if the confidence interval corresponding to the statistical distribution is narrow, it indicates that the neural network performs stably on the battery data; if the confidence interval is wide, it indicates that the neural network encounters unknown working conditions, and a warning is triggered to switch the energy storage system back to a safe mode, such as reduced operation, or trigger manual inspection when the threshold is reached at the interval boundary.

[0088] In addition, in the embodiments of the present application, the failure intensity function of the energy storage system is generated based on stress distribution, temperature and lithium ion concentration is:

[0089]

[0090] wherein, respectively represent function factors trained according to historical data;

[0091] wherein, represents the non-uniformity of temperature T , represents the two-dimensional physical coordinates of the battery hardware in the energy storage system; x, y wherein,

[0092] represents the stress distribution of the battery hardware in the three-dimensional coordinates at time t , represents the lithium ion concentration of the electrolyte in the energy storage system. In the embodiments, the failure intensity function is used to associate the physical state of the energy storage system with the failure risk of the battery, and after the failure intensity function is determined, the use life probability of the energy storage system within a time

[0093] is predicted based on the failure intensity function.

[0094]

[0095] wherein, represents a time integral variable.

[0096] By constructing the failure intensity function based on the stress distribution, the temperature gradient and the ion concentration change rate, the use life probability of the energy storage system is obtained. The use life probability is used as one of the battery parameters to dynamically quantify the material damage accumulation speed and update the driving life probability distribution in real time.

[0097] The above process uses a multi-head attention neural network to mine the association between coupled features and aging modes, constrains the model to comply with electrochemical laws through output feature partial derivatives, generates a probability distribution of battery parameters through forward propagation, and quantitatively evaluates uncertainty. By fusing physical mechanism and data-driven, the aging degree and risk level of the battery are directly reflected.​​​

[0098] S140, generating a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generating a control instruction based on the health report to control the operation of the energy storage system.

[0099] In an embodiment of the present application, generating a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generating a control instruction based on the health report to control the operation of the energy storage system comprises:

[0100] Generating a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution;

[0101] Extracting a risk label from the health report, mapping the risk label to a preset control strategy library to obtain a control instruction corresponding to the risk label;

[0102] Controlling the operation of the energy storage system through the control instruction.

[0103] Based on the stress distribution, the battery parameter and the statistical distribution output by the above steps, a structured health report is automatically generated. Specifically, the health report can include electrode stress overrun area, statistical layer life prediction and economic indicators, such as failure probability greater than 30% after 180 days. In addition, it can also include residual value assessment, forming a quantitative analysis including safety level, life stage and maintenance recommendations.

[0104] Extracting a risk label from the health report, mapping the risk label to a preset control strategy library to obtain a control instruction corresponding to the risk label. Specifically, the risk label in the health report is directly mapped to the control strategy library, for example, if the width of the statistical distribution is greater than 0.15, it is a high-risk scenario, then a derating instruction is generated to compress the maximum charge and discharge current to a safety threshold of 50%, and simultaneously increase the liquid cooling flow by 200% to counter local hot spots; if the battery parameter is less than 0.7, it is a life-end scenario, triggering a long-life mode, reducing the cutoff voltage by 0.1V and turning off the high-stress area cooling fan to delay structural failure.

[0105] Compiling the control instruction into a hardware executable signal and issuing it to the energy storage system. Specifically, in the power control layer, current switching is achieved through switch timing adjustment to ensure that the action accurately matches the optimization strategy; in the thermal management execution layer, the liquid cooling system is controlled based on the stress distribution, and the air cooling system is started and stopped based on the mechanical vibration model; in the safety monitoring layer, the embedded electrochemical solver continuously monitors the ion concentration change rate, and executes power-off protection when it is overrun.

[0106] S150, updating the health report of the energy storage system in the process of executing the control instruction.

[0107] In an embodiment of the present application, in the process of executing the control instruction, the health report of the energy storage system is updated, comprising:

[0108] In the process of executing the control instruction, real-time data is obtained based on a preset time period;

[0109] The real-time data is analyzed, and the health report of the energy storage system is updated according to the analysis result.

[0110] In the embodiment of the present application, while executing the optimization control instruction, the real-time running state of the energy storage unit is automatically captured at a preset time period to obtain real-time data, which can include core parameters such as electrical response, temperature distribution and mechanical vibration. This time period acquisition mechanism takes into account the monitoring needs of high-frequency key indicators (such as current mutation) and low-frequency slow variables (such as aging accumulation), ensuring continuous data flow reflecting the true state of the battery during the dynamic adjustment of the charging and discharging strategy. By synchronously recording the control action and the change of physical state, timeliness basis is provided for the iterative update of the health state.

[0111] Based on the newly added real-time data stream, multi-dimensional fusion analysis is triggered, first to identify abnormal trend characteristics by associating with the historical degradation mode library, then to deduce material-level damage progression combined with physical models, and finally to update the health state assessment by integrating life prediction algorithms. Each update not only corrects quantitative indicators such as capacity attenuation rate, but also reveals implicit risks such as electrode stress concentration and accelerated interfacial side reactions, making the health report continuously approach the true degradation state. This closed-loop process ensures that the operation and maintenance decisions always adapt to the latest working conditions of the battery, forming a self-evolution guarantee system of monitoring-evaluation-optimization, thereby achieving fault prediction and health management of the energy storage system.

[0112] In the embodiment, the health report and the control instruction form a dynamic closed loop. In the short-term control process, the high attenuation rate of the battery is actively suppressed during grid frequency regulation, which improves the cluster revenue by 12%; in the long-term control process, based on the distribution of battery service life, spare parts resources are pre-scheduled to reduce downtime by 18%; in the safety dimension, through stress-temperature coupling coefficient early warning, 7 times of dendrite growth risk are blocked. Ultimately, the composite benefits of 1.4 times life extension and 18% reduction in degree of electricity cost are achieved, thereby achieving pre-diagnosis and health management of the energy storage system.

[0113] The above process maps the risk label of the health report to the control strategy library to generate optimization instructions for current, voltage or temperature, periodically obtains real-time data during execution, updates the health state through online analysis, forms an evaluation, control and calibration closed loop, extends the service life of the battery by 53%, and reduces the probability of dangerous events.

[0114] The technical scheme of the application obtains historical data and working condition data collected by an industrial sensor through an intelligent sensor, pre-processes the historical data and the working condition data, and generates first data; constructs a concentration diffusion model of lithium ions and a heat conduction model of an energy storage system based on the first data, generates a coupling characteristic tensor and stress distribution representing the relationship between the material volume of the energy storage system and the electrolyte concentration by combining the concentration diffusion model and the heat conduction model; trains a neural network based on a multi-head attention mechanism through the historical data, inputs the coupling characteristic tensor into the neural network, and determines battery parameters and statistical distribution representing battery health; generates a health report of the energy storage system based on the stress distribution, the battery parameters and statistical distribution, and generates a control instruction based on the health report to control the operation of the energy storage system; and updates the health report of the energy storage system in the process of executing the control instruction. The coupling characteristic tensor is generated based on the ion concentration diffusion model and the heat conduction model, the internal state that cannot be directly measured is converted into quantifiable features, material-level damage diagnosis and early warning are realized, the accuracy of pre-diagnosis of the energy storage system is improved, and the efficiency of health management of the energy storage system is enhanced.

[0115] The following introduces a system embodiment of the application, which can be used to execute the data analysis-based energy storage system health evaluation and optimization method in the above embodiments of the application. It can be understood that the data analysis-based energy storage system health evaluation and optimization system can be a computer program (including program code) running in a computer device, for example, the data analysis-based energy storage system health evaluation and optimization system is an application software; the data analysis-based energy storage system health evaluation and optimization system can be used to execute the corresponding steps in the method provided in the embodiments of the application. For details not disclosed in the system embodiments of the application, please refer to the above embodiments of the data analysis-based energy storage system health evaluation and optimization method of the application.

[0116] Figure 3 A block diagram of a data analysis-based energy storage system health evaluation and optimization system according to an embodiment of the application is shown.

[0117] Referring to Figure 3 The data analysis-based energy storage system health evaluation and optimization system according to an embodiment of the application includes:

[0118] The acquisition unit 310 is configured to obtain historical data and working condition data collected by an industrial sensor through an intelligent sensor, pre-process the historical data and the working condition data, and generate first data;

[0119] The model unit 320 is configured to construct a concentration diffusion model of lithium ions and a heat conduction model of the energy storage system based on the first data, combine the concentration diffusion model and the heat conduction model, and generate a coupling characteristic tensor and a stress distribution representing a relationship between a material volume and an electrolyte concentration of the energy storage system.

[0120] The distribution unit 330 is configured to train a neural network based on a multi-head attention mechanism by using the historical data, input the coupling characteristic tensor into the neural network, and determine a battery parameter and a statistical distribution representing battery health.

[0121] The control unit 340 is configured to generate a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generate a control instruction based on the health report to control operation of the energy storage system.

[0122] The execution unit 350 is configured to update the health report of the energy storage system in a process of executing the control instruction.

[0123] In the present application, based on the foregoing scheme, the preprocessing of the historical data and the working condition data to generate the first data includes: aligning the historical data and the working condition data from different sampling rates by using a time stamp to generate second data; performing adaptive filtering on a current signal in the second data and information enhancement on temperature data in the second data to generate the first data.

[0124] In the present application, based on the foregoing scheme, the construction of the concentration diffusion model of lithium ions and the heat conduction model of the energy storage system based on the first data, the combination of the concentration diffusion model and the heat conduction model, and the generation of the coupling characteristic tensor and the stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system include: constructing a concentration diffusion model of lithium ions in the energy storage system based on lithium ion concentration and a real-time diffusion coefficient of energy storage material in the first data; constructing a heat conduction model of the energy storage system based on reaction data in the first data and material data of the energy storage material; combining the concentration diffusion model and the heat conduction model to generate a coupling characteristic tensor representing the relationship between the material volume and the electrolyte concentration of the energy storage system; and performing finite element method decomposition on the coupling characteristic tensor to obtain a four-dimensional stress distribution.

[0125] In the present application, based on the foregoing scheme, the training of the neural network based on the multi-head attention mechanism through the historical data, inputting the coupling feature tensor into the neural network, determining the battery parameter and the statistical distribution representing the battery health, comprises: training the neural network based on the multi-head attention mechanism based on the historical data; inputting the coupling feature tensor into the neural network to obtain output features, and calculating the partial derivative of the output features of the neural network to the input features; determining the battery parameter representing the battery health based on the output features and the partial derivative; performing K times of forward propagation on the coupling feature tensor in the neural network to generate a statistical distribution of the battery parameter.

[0126] In the present application, based on the foregoing scheme, the generation of the health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and the generation of the control instruction based on the health report to control the operation of the energy storage system, comprises: generating the health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution; extracting a risk label from the health report, mapping the risk label to a preset control strategy library to obtain a control instruction corresponding to the risk label; and controlling the operation of the energy storage system through the control instruction.

[0127] In the present application, based on the foregoing scheme, the updating of the health report of the energy storage system in the process of executing the control instruction, comprises: acquiring real-time data based on a preset time period in the process of executing the control instruction; analyzing the real-time data and updating the health report of the energy storage system according to the analysis result.

[0128] In the present application, based on the foregoing scheme, further comprising: generating a failure intensity function of the energy storage system based on the stress distribution, the temperature and the lithium ion concentration; generating a service life probability of the energy storage system based on the failure intensity function, and taking the service life probability as one of the battery parameters.

[0129] The technical scheme of the application obtains historical data and working condition data collected by an industrial sensor through an intelligent sensor, pre-processes the historical data and the working condition data, and generates first data; constructs a concentration diffusion model of lithium ions and a heat conduction model of an energy storage system based on the first data, generates a coupling characteristic tensor and stress distribution representing the relationship between the material volume of the energy storage system and the electrolyte concentration by combining the concentration diffusion model and the heat conduction model, trains a neural network based on a multi-head attention mechanism through the historical data, inputs the coupling characteristic tensor into the neural network, determines battery parameters and statistical distribution representing battery health, generates a health report of the energy storage system based on the stress distribution, the battery parameters and statistical distribution, generates a control instruction based on the health report to control the operation of the energy storage system, and updates the health report of the energy storage system in the process of executing the control instruction. The coupling characteristic tensor is generated based on the ion concentration diffusion model and the heat conduction model, the internal state that cannot be directly measured is converted into quantifiable features, material-level damage diagnosis and early warning are realized, the accuracy of pre-diagnosis of the energy storage system is improved, and the efficiency of health management of the energy storage system is enhanced.

[0130] Figure 4 A structural schematic diagram of a computer system of an electronic device suitable for implementing the embodiments of the application is shown.

[0131] It should be noted that the computer system of the electronic device in this embodiment is only an example, and should not impose any limitation on the functions and use range of the embodiments of the application.

[0132] The computer system in this embodiment includes a central processing unit 401, which can perform various appropriate actions and processes according to programs stored in a read-only memory 402 or loaded into a random access memory 403 from a storage part 408, such as the data analysis-based energy storage system health evaluation and optimization method described in the above embodiments. Various programs and data required for system operation are also stored in the random access memory 403. The central processing unit 401, the read-only memory 402, and the random access memory 403 are connected to each other through a bus 404. An input / output interface 405 is also connected to the bus 404.

[0133] The following components are connected to the input / output interface 405: an input portion 406 including input devices such as a keyboard and a mouse; an output portion 407 including output devices such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and a speaker; a storage portion 408 including a hard disk; and a communication portion 409 including a network interface card such as a LAN (Local Area Network) card, a modem, and the like. The communication portion 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as necessary. A removable media 411 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, and the like is attached to the drive 410 as necessary, so that a computer program read therefrom is installed in the storage portion 408 as necessary.

[0134] In particular, the processes described above with reference to the flowcharts can be implemented as a computer software program according to embodiments of the present application. For example, embodiments of the present application include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing computer programs for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network by the communication portion 409, and / or installed from the removable media 411. When the computer program is executed by the central processing unit 401, various functions defined in the system of the present application are executed.

[0135] It should be noted that the computer-readable medium in the embodiments of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination thereof. The computer-readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination thereof. More specific examples of the computer-readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (Compact Disc Read-Only Memory, CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer-readable signal medium can include a data signal carrying computer-readable computer programs in a baseband or as a part of a carrier wave. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, which can send, propagate or transmit programs for use by or in conjunction with an instruction execution system, device or apparatus. The computer programs contained in the computer-readable medium can be transmitted by any suitable medium, including but not limited to wireless, wired, or the like, or any suitable combination thereof.

[0136] The flowcharts and block diagrams in the drawings illustrate the possible implementation architectures, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In the flowcharts or block diagrams, each block can represent a module, a program segment or a part of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks can occur in different orders than that shown in the drawings. For example, two blocks represented in succession can actually be executed substantially in parallel, and sometimes in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams or flowcharts, and the combination of blocks in the block diagrams or flowcharts, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0137] The units described in the embodiments of the present application can be implemented by software, or can be implemented by hardware, and the units described can also be arranged in a processor. In some cases, the names of the units do not constitute a limitation on the units themselves.

[0138] According to an aspect of the present application, a computer program product or computer program is provided, 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 the processor executes the computer instructions, so that the computer device executes the method provided in the various optional implementation manners described above.

[0139] As another aspect, the present application also provides a computer readable medium, which can be included in the electronic device described in the above embodiments, or can exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to implement the data analysis based energy storage system health assessment and optimization method described in the above embodiments.

[0140] It should be noted that although several modules or units of the device for action execution are mentioned in the above detailed description, such division is not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into several modules or units.

[0141] From the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, U disk, mobile hard disk, etc.) or network, and includes several instructions to make a computing device (which can be a personal computer, server, touch terminal, or network device, etc.) execute the methods according to the embodiments of the present application.

[0142] Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the embodiments disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the application following the general principles thereof and including such departures from the present disclosure as come within known use or custom in the art.

[0143] It is to be understood that the application is not limited to the precise construction already described above and shown in the drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application should only be limited by the claims appended hereto.

Claims

1. A data analysis based energy storage system health assessment and optimization method, characterized in that, The method comprises: obtaining historical data and working condition data collected by an intelligent sensor, and preprocessing the historical data and the working condition data to generate first data; constructing a lithium ion concentration diffusion model and a heat conduction model of the energy storage system based on the first data, combining the concentration diffusion model and the heat conduction model to generate a coupling characteristic tensor and a stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system; training a neural network based on a multi-head attention mechanism based on the historical data, inputting the coupling characteristic tensor into the neural network, and determining a battery parameter and a statistical distribution representing the battery health; generating a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generating a control instruction based on the health report to control the operation of the energy storage system; updating the health report of the energy storage system in the process of executing the control instruction.

2. The data analytics based health assessment and optimization method of energy storage system according to claim 1, wherein, The preprocessing of the historical data and the working condition data to generate first data comprises: aligning the historical data and the working condition data from different sampling rates through time stamping to generate second data; performing adaptive filtering on the current signal in the second data and information enhancement on the temperature data in the second data to generate first data.

3. The data analytics based energy storage system health assessment and optimization method of claim 1, wherein, The construction of the lithium ion concentration diffusion model and the heat conduction model of the energy storage system based on the first data, and the combination of the concentration diffusion model and the heat conduction model to generate a coupling characteristic tensor and a stress distribution representing the relationship between the material volume and the electrolyte concentration of the energy storage system comprises: constructing a lithium ion concentration diffusion model in the energy storage system based on the lithium ion concentration and the real-time diffusion coefficient of the energy storage material in the first data; constructing a heat conduction model of the energy storage system based on the reaction data in the first data and the material data of the energy storage material; combining the concentration diffusion model and the heat conduction model to generate a coupling characteristic tensor representing the relationship between the material volume and the electrolyte concentration of the energy storage system; performing finite element method decomposition on the coupling characteristic tensor to obtain a four-dimensional stress distribution.

4. The data analytics based health assessment and optimization method of energy storage system according to claim 1, wherein, The training of a neural network based on a multi-head attention mechanism based on the historical data, the input of the coupling characteristic tensor into the neural network, and the determination of a battery parameter and a statistical distribution representing the battery health comprise: training a neural network based on a multi-head attention mechanism based on the historical data; inputting the coupling characteristic tensor into the neural network to obtain output features, and calculating the partial derivative of the output features of the neural network with respect to the input features; determining a battery parameter representing the battery health based on the output features and the partial derivative. performing a forward pass on the coupling feature tensor in the neural network K a second forward pass, generating a statistical distribution of battery parameters.

5. The data analytics based energy storage system health assessment and optimization method of claim 4, wherein, Further comprising: generating a failure intensity function of the energy storage system based on the stress distribution, the temperature and the lithium ion concentration; generating a service life probability of the energy storage system based on the failure intensity function, and taking the service life probability as one of the battery parameters.

6. The data analytics based energy storage system health assessment and optimization method of claim 1, wherein, The generation of a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and the generation of a control instruction based on the health report to control the operation of the energy storage system comprise: generate a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution; extract a risk label from the health report, map the risk label to a preset control strategy library, and obtain a control instruction corresponding to the risk label; control the energy storage system to operate through the control instruction.

7. The data analytics based energy storage system health assessment and optimization method of claim 1, wherein, In the process of executing the control instruction, the health report of the energy storage system is updated, including: In the process of executing the control instruction, real-time data is obtained based on a preset time period; analyze the real-time data, and update the health report of the energy storage system according to the analysis result.

8. A data analytics based energy storage system health assessment and optimization system, characterized in that, including: an acquisition unit configured to acquire historical data and working condition data collected by an industrial sensor through an intelligent sensor, and to preprocess the historical data and the working condition data to generate first data; a model unit configured to construct a concentration diffusion model of lithium ions and a heat conduction model of the energy storage system based on the first data, combine the concentration diffusion model and the heat conduction model, and generate a coupling feature tensor and a stress distribution representing a relationship between a material volume of the energy storage system and a concentration of electrolyte; a distribution unit configured to train a neural network based on a multi-head attention mechanism through the historical data, input the coupling feature tensor into the neural network, and determine a battery parameter and a statistical distribution representing battery health; a control unit configured to generate a health report of the energy storage system based on the stress distribution, the battery parameter and the statistical distribution, and generate a control instruction based on the health report to control the energy storage system to operate; an execution unit configured to update the health report of the energy storage system in the process of executing the control instruction.

9. The data analytics based energy storage system health assessment and optimization system, as claimed in claim 8, wherein, The preprocessing of the historical data and the working condition data to generate first data includes: aligning the historical data and the working condition data from different sampling rates through timestamps to generate second data; performing adaptive filtering on current signals in the second data and information enhancement on temperature data in the second data to generate first data.

10. The data analytics based energy storage system health assessment and optimization system, as claimed in claim 8, wherein, Based on the first data, a concentration diffusion model of lithium ions and a heat conduction model of the energy storage system are constructed, and the concentration diffusion model and the heat conduction model are combined to generate a coupling feature tensor and a stress distribution representing a relationship between a material volume of the energy storage system and a concentration of electrolyte, including: based on the lithium ion concentration in the first data and the real-time diffusion coefficient of the energy storage material, a concentration diffusion model of lithium ions in the energy storage system is constructed; based on the reaction data in the first data and the material data of the energy storage material, a heat conduction model of the energy storage system is constructed; combine the concentration diffusion model and the heat conduction model to generate a coupling feature tensor representing the relationship between the material volume of the energy storage system and the concentration of the electrolyte; perform finite element method decomposition on the coupling feature tensor to obtain a four-dimensional stress distribution.

Images