Method, apparatus and server for determining battery health status

Abstract

This specification provides a method, apparatus, and server for determining the state of health of a battery. Before specific implementation, a corresponding preset battery state-of-health prediction model is trained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model. During specific implementation, a target ultrasonic transducer group is deployed on the working surface of the target battery; among them, the target ultrasonic transducer group includes three ultrasonic transducers, and the connecting edges between the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; then, waveform signals about the target battery are collected by using the target ultrasonic transducer group; according to the waveform signals, corresponding time-domain data and frequency-domain data are obtained and utilized to determine target features about the target battery; the state of health of the target battery is determined by using the preset battery state-of-health prediction model to process the target features. Thus, the state of health of the target battery can be monitored and determined more efficiently and accurately.

Description

Technical Field

[0001] This specification belongs to the technical field of electrical data processing, and particularly relates to a method, device, and server for determining the state of health of a battery. Background Art

[0002] With the development and popularization of new energy technologies, more and more new energy batteries (e.g., sodium-ion batteries) have gradually been applied to scenarios such as electric vehicles, energy storage industries, and communication base stations.

[0003] During the use of new energy batteries, it is often necessary to monitor the state of health of the batteries. However, based on existing methods, when determining the state of health of a battery, there are often problems such as poor accuracy, low reference value, and easy occurrence of errors.

[0004] In response to the above problems, no effective solution has been proposed yet. Summary of the Invention

[0005] This specification provides a method, device, and server for determining the state of health of a battery, which can efficiently and accurately monitor and determine the state of health of a target battery, and effectively reduce detection errors.

[0006] This specification provides a method for determining the state of health of a battery, including:

[0007] Using a target ultrasonic transducer array to emit ultrasonic signals on the working surface of a target battery; and collecting corresponding waveform signals; wherein, the target ultrasonic transducer array includes 3 ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery;

[0008] According to the waveform signals, obtaining corresponding time-domain data and frequency-domain data;

[0009] According to the time-domain data and frequency-domain data, determining target characteristics regarding the target battery;

[0010] Using a preset battery state-of-health prediction model to process the target characteristics to obtain a corresponding target prediction result; wherein, the preset battery state-of-health prediction model is trained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model;

[0011] According to the target prediction result, determining the state of health of the target battery.

[0012] In one embodiment, the target characteristics include at least one of the following: average intensity attenuation ratio, average flight time, average main frequency offset ratio.

[0013] In one embodiment, when the target feature at least includes the average intensity attenuation ratio, based on the time-domain data and the frequency-domain data, determining the target feature regarding the target battery includes:

[0014] Determining the average intensity attenuation ratio of the target battery according to the following formula:

[0015]

[0016] where is the average intensity attenuation ratio, A i1 is the signal intensity of the signal ranked first in the signal intensity values in the signal data received by the ultrasonic transducer numbered i, A i2 is the signal intensity of the signal ranked second in the signal intensity values in the signal data received by the ultrasonic transducer numbered i, A i3 is the signal intensity of the signal ranked third in the signal intensity values in the signal data received by the ultrasonic transducer numbered i, and A is the signal intensity of the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0017] In one embodiment, when the target feature at least includes the average flight time, based on the time-domain data and the frequency-domain data, determining the target feature regarding the target battery includes:

[0018] Determining the average flight time of the target battery according to the following formula:

[0019]

[0020] where is the average flight time, T i1 is the flight time of the signal ranked first in the signal intensity values in the signal data received by the ultrasonic transducer numbered i, T i2 is the flight time of the signal ranked second in the signal intensity values in the signal data received by the ultrasonic transducer numbered i, T i3 is the flight time of the signal ranked third in the signal intensity values in the signal data received by the ultrasonic transducer numbered i.

[0021] In one embodiment, when the target feature at least includes the average main frequency offset ratio, based on the time-domain data and the frequency-domain data, determining the target feature regarding the target battery includes:

[0022] Determining the average main frequency offset ratio of the target battery according to the following formula:

[0023]

[0024] where is the average main frequency offset ratio, f i1The frequency of the signal with the highest signal intensity value among the signal data received by the ultrasonic transducer numbered i, f i2 The frequency of the signal with the second highest signal intensity value among the signal data received by the ultrasonic transducer numbered i, f i3 The frequency of the signal with the third highest signal intensity value among the signal data received by the ultrasonic transducer numbered i, f is the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0025] In one embodiment, the method further includes:

[0026] Select a plurality of batteries in different health states as sample batteries;

[0027] Deploy the corresponding target ultrasonic transducer group on the working plane of the sample battery according to the preset deployment rule, and use the target ultrasonic transducer group to obtain the sample data corresponding to the sample battery through a test experiment;

[0028] Use the sample data to construct the corresponding sample data set;

[0029] Determine a plurality of computing particles;

[0030] Randomly extract a plurality of sample data from the sample data set to generate a plurality of sub-sample data sets; and construct an initial prediction model based on the random forest regression prediction model;

[0031] Call a plurality of computing particles to perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data set until the training end condition is met, and obtain a preset battery health state prediction model that meets the requirements.

[0032] In one embodiment, calling a plurality of computing particles to perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data set includes:

[0033] Call a plurality of computing particles to respectively determine the current round's velocity parameters corresponding to the computing particles based on the local solution particle parameter group of the previous round and the global solution particle parameter group of the previous round; wherein, the particle parameter group at least includes: pruning threshold, number of decision trees, proportion of test samples;

[0034] Call a plurality of computing particles to respectively update the prediction model of the previous round based on the current round's velocity parameters and the sub-sample data set to obtain the prediction model of the current round for each;

[0035] Call a plurality of computing particles to respectively update the local solution particle parameter group of the current round according to the prediction model of the current round for each; and update the global solution particle parameter group of the current round according to the local solution particle parameter group of the current round;

[0036] Detect whether the current condition for ending the training is satisfied;

[0037] When it is determined that the condition for ending the training is satisfied, based on the global solution particle parameter group of the current round, determine a preset battery health state prediction model that meets the requirements.

[0038] This specification also provides a device for determining the battery health state, including:

[0039] An acquisition module, configured to use a target ultrasonic transducer group to transmit ultrasonic signals on the working surface of a target battery; and acquire corresponding waveform signals; wherein, the target ultrasonic transducer group includes 3 ultrasonic transducers, the connection sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery;

[0040] An acquisition module, configured to obtain corresponding time-domain data and frequency-domain data according to the waveform signals;

[0041] A first determination module, configured to determine a target feature regarding the target battery according to the time-domain data and the frequency-domain data;

[0042] A processing module, configured to process the target feature by using a preset battery health state prediction model to obtain a corresponding target prediction result; wherein, the preset battery health state prediction model is trained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model;

[0043] A second determination module, configured to determine the health state of the target battery according to the target prediction result.

[0044] This specification also provides a server, including a processor and a memory for storing processor-executable instructions, and when the processor executes the instructions, the steps of the method for determining the battery health state are implemented.

[0045] This specification also provides a computer-readable storage medium, on which computer instructions are stored, and when the instructions are executed by a processor, the steps of the method for determining the battery health state are implemented.

[0046] Based on the method, device, and server for determining the state of health of a battery provided in this specification, before specific implementation, by jointly using the particle swarm optimization algorithm and the random forest regression prediction model, it is possible to efficiently train a preset battery state of health prediction model with relatively high accuracy and good effect. During specific implementation, first, according to the preset deployment rules, deploy the corresponding target ultrasonic transducer group on the working surface of the target battery of concern; among them, the target ultrasonic transducer group includes 3 ultrasonic transducers, the 3 ultrasonic transducers are equidistant from each other, the connecting edges between the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery. In this way, the above-mentioned target ultrasonic transducer group can be used to simultaneously transmit ultrasonic signals to collect waveform signals that are relatively comprehensive, complete, and of high reference value regarding the target battery. Then, based on the waveform signals, obtain and utilize the corresponding time-domain data and frequency-domain data to determine the target features regarding the target battery; use the preset battery state of health prediction model to process the target features to obtain the corresponding target prediction results; and based on this target prediction result, determine the state of health of the target battery. Thus, it is possible to monitor and determine the state of health of the target battery more efficiently and accurately, effectively reduce detection errors, and ensure the stable and safe operation of the target battery. Description of the Drawings

[0047] To more clearly illustrate the embodiments of this specification, the following will briefly introduce the drawings required for use in the embodiments. The drawings described below are only some embodiments recorded in this specification. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0048] Figure 1 It is a flowchart showing the method for determining the state of health of a battery provided by an embodiment of this specification;

[0049] Figure 2 It is a schematic diagram of an embodiment applying the method for determining the state of health of a battery provided by an embodiment of this specification in a scenario example;

[0050] Figure 3 It is a schematic diagram of an embodiment applying the method for determining the state of health of a battery provided by an embodiment of this specification in a scenario example;

[0051] Figure 4 It is a schematic diagram of an embodiment applying the method for determining the state of health of a battery provided by an embodiment of this specification in a scenario example;

[0052] Figure 5 It is a schematic diagram showing the structural composition of a server provided by an embodiment of this specification;

[0053] Figure 6 It is a schematic structural diagram of a device for determining the battery health state provided by an embodiment of this specification;

[0054] Figure 7 It is a schematic diagram of an embodiment of applying the method for determining the battery health state provided by the embodiment of this specification in a scenario example;

[0055] Figure 8 It is a schematic diagram of an embodiment of applying the method for determining the battery health state provided by the embodiment of this specification in a scenario example;

[0056] Figure 9 It is a schematic diagram of an embodiment of applying the method for determining the battery health state provided by the embodiment of this specification in a scenario example. Detailed implementation manners

[0057] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this specification. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this specification.

[0058] It should be noted that the information and data related to users involved in the embodiments of this specification are all information and data authorized by the users or fully authorized by the relevant parties. Moreover, the processing of relevant data, such as collection, storage, use, processing, transmission, provision, disclosure, and application, all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users or relevant parties to choose to authorize or refuse.

[0059] It should also be noted that in the embodiments of this specification, some industry-existing solutions such as certain software, components, models, etc. may be mentioned. They should be regarded as exemplary, and their purpose is only to illustrate the feasibility in the implementation of the technical solutions of this application, but it does not mean that the applicant has already or necessarily used this solution.

[0060] Refer to Figure 1 As shown, the embodiments of this specification provide a method for determining the battery health state, where this method is specifically applied to the server side. Specifically in implementation, this method may include the following content:

[0061] S101: Transmit ultrasonic signals on the working surface of the target battery using the target ultrasonic transducer array; and collect the corresponding waveform signals. Among them, the target ultrasonic transducer array includes 3 ultrasonic transducers, the connecting edges of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery;

[0062] S102: Obtain the corresponding time-domain data and frequency-domain data according to the waveform signals;

[0063] S103: Determine the target characteristics of the target battery according to the time-domain data and frequency-domain data;

[0064] S104: Process the target characteristics using a preset battery state of health prediction model to obtain the corresponding target prediction result. Among them, the preset battery state of health prediction model is trained by jointly using the particle swarm optimization algorithm and the random forest regression prediction model;

[0065] S105: Determine the state of health of the target battery according to the target prediction result.

[0066] Among them, the above state of health (SOH) can be specifically understood as an embodiment of the battery health life status, and can specifically be the specific embodiment of the battery's power, energy, charge and discharge power and other states.

[0067] The above target battery can specifically be a sodium-ion battery, which has good low-temperature performance, fast charging performance, energy density, etc. Of course, in addition to sodium-ion batteries, the above target battery can also be other types of batteries such as lead-acid batteries.

[0068] Specifically, the above target battery can be a battery used for storing electric energy deployed in a new energy power station (for example, a wind-solar hybrid power station).

[0069] Before specific implementation, the target ultrasonic transducer array can be deployed on the working surface of the target battery according to a preset deployment rule. For details, please refer to Figure 2 As shown in the figure. The tab can specifically refer to the metal conductor that leads out the positive and negative electrodes from the battery core of the target battery, and can be understood as the contact point of the positive and negative electrodes when the target battery is charged and discharged.

[0070] The above target ultrasonic transducer array can specifically include 3 identical ultrasonic transducers; and the distances between the ultrasonic transducers are equal (for example, the distance d between each other 1 =d 2 =d 3(= 1.5 cm), the connecting edges between different ultrasonic transducers form an equilateral triangle. Further, the center of the equilateral triangle formed by the three ultrasonic transducers coincides with the center of the working surface of the target battery. Among them, the working surface of the above-mentioned target battery can specifically be the surface with the largest internal area of the target battery (for example, 10 cm * 7 cm). When specifically deploying the ultrasonic transducer sensors, try to make the three ultrasonic transducer sensors lie on the same horizontal plane.

[0071] Among them, the above ultrasonic transducer can specifically be understood as a device that can convert electrical energy into sound energy.

[0072] Specifically, the above target ultrasonic transducer group is also connected to the server by wired or wireless means. Correspondingly, the server can control the start of the target ultrasonic transducer group to collect and upload the corresponding waveform signals, and then determine the health status of the target battery according to the above waveform signals.

[0073] It should be noted that here three ultrasonic transducers are selected to construct the target ultrasonic transducer group, considering that the equilateral triangle formed by the three ultrasonic transducers can better ensure the uniform distribution of the emitted ultrasonic signals; and, using three ultrasonic transducers can effectively avoid redundant signals, reduce complex calculations, and at the same time can provide sufficient signal coverage, balancing signal reception and computational efficiency compared to using a larger number of ultrasonic transducers. Specifically, using three ultrasonic transducers to form an equilateral triangle can ensure the uniform distribution of signals, making the propagation of ultrasonic signals in the battery more comprehensive, avoiding signal blind spots, and having better balance. In addition, based on the symmetry of the above equilateral triangle, the signal capture ability can also be effectively enhanced, helping the ultrasonic transducer to collect relatively more refined signals, so as to better detect potential problems inside the battery (such as defects, gas accumulation, etc.).

[0074] In this way, the ultrasonic signals emitted by the target ultrasonic transducer group deployed according to the preset deployment rules can more completely and comprehensively cover the entire area of the target battery, avoiding omissions, and can more accurately achieve the detection of the overall situation of the target battery.

[0075] Further, the distance between the ultrasonic transducer and the center of the working surface is greater than the first distance threshold and less than or equal to the second distance threshold. Among them, the first distance threshold and the second distance threshold can be determined according to the size parameters of the working surface of the target battery and the performance parameters of the ultrasonic transducer. In this way, on the one hand, the deployed target ultrasonic transducer group can comprehensively and completely cover the entire area of the target battery; on the other hand, it can also reduce the attenuation of signals during signal acquisition, ensuring that the signals collected by the ultrasonic transducer have a high degree of credibility.

[0076] In specific implementation, it is possible to control the 3 ultrasonic transducers included in the target ultrasonic transducer group to simultaneously emit the same ultrasonic signal; then, simultaneously control the 3 ultrasonic transducers to collect the reflected signals of the ultrasonic signals they emit themselves and the ultrasonic signals emitted by other ultrasonic transducers, so as to obtain corresponding waveform signals.

[0077] Among them, the transmission center frequencies of the above ultrasonic transducers can include: 300KHz - 5MHz. Specifically, according to the battery attribute characteristics of the target battery, a matching frequency can be selected as the transmission center frequency of the ultrasonic transducer.

[0078] For example, in the case of determining that the positive electrode material of the target battery is a sodium-ion soft-pack battery with Prussian blue, the transmission center frequency of the ultrasonic transducer can be determined to be 1MHz; in the case of determining that the positive electrode material of the target battery is a sodium-ion soft-pack battery with layered transition metal oxide, the transmission center frequency of the ultrasonic transducer can be determined to be 2.5MHz; in the case of determining that the positive electrode material of the target battery is sodium iron phosphate, the transmission center frequency of the ultrasonic transducer can be determined to be 800KHz.

[0079] In this way, according to different battery attribute characteristics, the target ultrasonic transducer group can be controlled to emit ultrasonic signals at different transmission center frequencies, so as to obtain waveform signals with relatively better effects and higher precision.

[0080] In specific implementation, first, corresponding time-domain data and frequency-domain data, two different-dimensional data, can be obtained according to the waveform signal; then, the time-domain data and frequency-domain data, these two different-dimensional data, are jointly used to calculate the required signal characteristic parameters of the target battery based on the ultrasonic signal as the target characteristics; then, the preset battery health state prediction model that has been pre-trained is used to process the above target characteristics to obtain the corresponding target prediction result; and according to this target prediction result, the health state of the target battery is finally determined.

[0081] Among them, the above preset battery health state prediction model is obtained by pre-jointly using the particle swarm optimization algorithm and the random forest regression prediction model, and it is a random forest regression prediction model that can automatically predict and output the battery health state based on the relevant target characteristics input into the model. The specific training method of this model will be described separately later.

[0082] Based on the above embodiments, before specific implementation, by jointly using the particle swarm optimization algorithm and the random forest regression prediction model, a preset battery health state prediction model with relatively high accuracy and good effect can be trained more efficiently. During specific implementation, a target ultrasonic transducer group can be deployed on the working surface of the target battery to be concerned; among them, the target ultrasonic transducer group includes 3 ultrasonic transducers, the 3 ultrasonic transducers are equidistant from each other, and the connecting edges form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery. In this way, the above target ultrasonic transducer group can be used to collect a relatively comprehensive and complete waveform signal about the target battery by simultaneously emitting ultrasonic signals. Then, according to the waveform signal, the corresponding time-domain data and frequency-domain data are obtained and utilized to determine the target features about the target battery; the preset battery health state prediction model is used to process the target features to obtain the corresponding target prediction result; and according to the target prediction result, the health state of the target battery is determined. Thus, the health state of the target battery can be monitored and determined more efficiently and accurately, the error can be effectively reduced, and the stable and safe operation of the target battery can be ensured.

[0083] During specific implementation, the target ultrasonic transducer group can be pre-deployed on the working surface of the target battery, and then according to each specified time interval, the target ultrasonic transducer group is triggered to start collecting and utilizing the corresponding waveform signal to perform regular monitoring on the health state of the target battery to ensure the safe and stable operation of the target battery.

[0084] In some embodiments, the above-mentioned obtaining the corresponding time-domain data and frequency-domain data according to the waveform signal may specifically include the following contents during specific implementation:

[0085] S1: By performing filtering and amplification processing on the waveform signal, the corresponding time-domain signal is obtained;

[0086] S2: By performing Fourier transform on the time-domain signal, the corresponding frequency-domain signal is obtained.

[0087] Based on the above embodiments, the waveform signal can be used to process and obtain signal data about the target battery in different dimensions such as time domain and frequency domain, so as to comprehensively and accurately predict the health state of the target battery based on the above signal data in the follow-up.

[0088] In some embodiments, the target features include at least one of the following: average intensity attenuation ratio, average flight time, average main frequency offset ratio, etc.

[0089] During specific implementation, one or a combination of the above-listed features can be used as the target features.

[0090] In some embodiments, when the target feature at least includes the average intensity attenuation ratio, based on the time-domain data and the frequency-domain data, the target feature regarding the target battery is determined. Specifically in implementation, it may include:

[0091] According to the following formula, determine the average intensity attenuation ratio of the target battery:

[0092]

[0093] Wherein, is the average intensity attenuation ratio, A i1 is the signal intensity of the signal ranked first in the signal intensity values of the signal data received by the ultrasonic transducer numbered i, A i2 is the signal intensity of the signal ranked second in the signal intensity values of the signal data received by the ultrasonic transducer numbered i, A i3 is the signal intensity of the signal ranked third in the signal intensity values of the signal data received by the ultrasonic transducer numbered i, and A is the signal intensity of the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0094] In some embodiments, when the target feature at least includes the average flight time, based on the time-domain data and the frequency-domain data, the target feature regarding the target battery is determined. Specifically in implementation, it may include:

[0095] According to the following formula, determine the average flight time of the target battery:

[0096]

[0097] Wherein, is the average flight time, T i1 is the flight time of the signal ranked first in the signal intensity values of the signal data received by the ultrasonic transducer numbered i, T i2 is the flight time of the signal ranked second in the signal intensity values of the signal data received by the ultrasonic transducer numbered i, T i3 is the flight time of the signal ranked third in the signal intensity values of the signal data received by the ultrasonic transducer numbered i.

[0098] In some embodiments, when the target feature at least includes the average main frequency offset ratio, based on the time-domain data and the frequency-domain data, the target feature regarding the target battery is determined. Specifically in implementation, it may include:

[0099] According to the following formula, determine the average main frequency offset ratio of the target battery:

[0100]

[0101] Wherein, is the average main frequency offset ratio, f i1 is the frequency of the signal with the first signal intensity value in the signal data received by the ultrasonic transducer numbered i, f i2 is the frequency of the signal with the second signal intensity value in the signal data received by the ultrasonic transducer numbered i, f i3 is the frequency of the signal with the third signal intensity value in the signal data received by the ultrasonic transducer numbered i, and f is the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0102] In some embodiments, during specific implementation, the signal data received by each ultrasonic transducer can be extracted based on the time-domain data and frequency-domain data first; among them, the above signal data can include multiple groups of signal data corresponding to different frequencies, and each group of signal data includes at least attribute information such as signal intensity, flight signal, and frequency. Then, according to the signal intensity, the signal data received by each ultrasonic transducer is sorted in descending order of signal intensity; then, the attribute information of the top three sorted signal data is obtained; and according to the attribute information of the top three sorted signal data, the corresponding target characteristics of the target battery are calculated.

[0103] In some embodiments, after determining the health status of the target battery according to the target prediction result, when the method is specifically implemented, the following content can also be included:

[0104] S1: When it is determined that the health status of the target battery is abnormal, determine the abnormal risk level of the target battery;

[0105] S2: Detect whether the abnormal risk level of the target battery is less than a preset level threshold;

[0106] S3: When it is determined that the abnormal risk level is less than the preset level threshold, determine a matching target maintenance strategy from a preset set of maintenance strategies according to the abnormal risk level;

[0107] S4: Automatically maintain the target battery according to the target maintenance strategy.

[0108] Among them, the preset set of maintenance strategies contains multiple preset maintenance strategies; among them, each preset maintenance strategy corresponds to at least one abnormal risk level.

[0109] Before specific implementation, a large number of historical battery maintenance records can be collected; then, the historical battery maintenance records are clustered to obtain multiple maintenance operation data groups; among them, each maintenance operation data group corresponds to at least one abnormal risk level and contains the common maintenance operations corresponding to the corresponding abnormal risk level; according to each maintenance operation data group, multiple preset maintenance strategies corresponding to the abnormal risk level are constructed; and multiple preset maintenance strategies are combined to obtain a preset maintenance strategy set.

[0110] Based on the above embodiments, when the abnormal risk level of the target battery is relatively small and the risk degree is relatively low, the corresponding target maintenance strategy can be determined and utilized to automatically maintain the target battery, so as to eliminate in time the risk factors affecting the health state of the target battery and continue to operate stably and safely.

[0111] During specific implementation, after detecting whether the abnormal risk level of the target battery is less than a preset level threshold, the method may further include the following when specifically implemented:

[0112] S1: When it is determined that the abnormal risk level is greater than or equal to the preset level threshold, according to the abnormal risk level, a corresponding target maintenance strategy is determined from the preset maintenance strategy set;

[0113] S2: Invoke a preset abnormal cause analysis model to obtain multiple abnormal causes (such as gas accumulation, etc.) with probability values greater than or equal to a preset probability threshold by processing target features; where the preset abnormal cause analysis model is an algorithm model obtained by pre-using big data analysis and machine learning training;

[0114] S3: Combine the target maintenance strategy and the abnormal cause to generate an abnormal analysis and processing report for the target battery;

[0115] S4: Send an abnormal risk handling prompt message about the target battery to the operation and maintenance end; where the abnormal risk handling prompt message carries at least the abnormal analysis and processing report.

[0116] Correspondingly, the operation and maintenance end can receive and respond to the abnormal risk handling prompt message and display the abnormal analysis and processing report to the operation and maintenance personnel, so that the operation and maintenance personnel can use the abnormal analysis and processing report as a reference to manually maintain the target battery to ensure the safe and stable operation of the target battery and extend the service life of the target battery.

[0117] In some embodiments, referring to Figure 3 as shown, the method may further include the following when specifically implemented:

[0118] S1: Screen multiple batteries with different health states as sample batteries;

[0119] S2: Deploy the corresponding target ultrasonic transducer group on the working plane of the sample battery according to the preset deployment rules, and use the target ultrasonic transducer group to conduct test experiments to obtain sample data corresponding to the sample battery;

[0120] S3: Use the sample data to construct a corresponding sample data set;

[0121] S4: Determine multiple computing particles;

[0122] S5: Randomly extract multiple sample data from the sample data set to generate multiple sub-sample data sets; and construct an initial prediction model based on the random forest regression prediction model;

[0123] S6: Invoke multiple computing particles to perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data set until the training end condition is met, and obtain a preset battery health state prediction model that meets the requirements.

[0124] Among them, the above computing particles can specifically be relatively independent computing nodes (for example, computer nodes) in a distributed cluster. Each computing particle can be regarded as an individual, and is deployed with corresponding local iterative training algorithm rules, and is connected to different computing particles. Among them, the above distributed cluster can specifically be understood as a system composed of multiple computer nodes. The computer nodes communicate and cooperate with each other through a network to jointly complete a task or provide a service, and has advantages such as high availability, high performance, and scalability. In this way, the performance advantages of the distributed system can be utilized to efficiently complete relevant training.

[0125] The above sample battery is specifically a battery with a known health state. Specifically, the above sample battery can at least include a sample battery with a normal health state and a sample battery with an abnormal health state.

[0126] When specifically conducting test experiments, ultrasonic signals can be emitted on the working surface of the sample battery by using the target ultrasonic transducer group; and the corresponding waveform signals can be collected; then the waveform signals are combined with the health state label of the sample battery to obtain sample data corresponding to the sample battery.

[0127] During specific implementation, for the same sample battery, test experiments can be conducted in the above manner multiple times to obtain multiple waveform signals; then according to the multiple waveform signals, through statistical analysis, a credible waveform signal is selected to construct the corresponding sample data.

[0128] During specific implementation, multiple sample data can be combined to construct a corresponding sample data set (or sample database); further, according to the data scale of the sample data set, combined with the state performance of a single computing particle, a matching numerical value can be determined, and then multiple computing particles can be determined.

[0129] Based on the above embodiments, the target ultrasonic transducer group can be used to collect relatively comprehensive and valuable sample data; furthermore, the particle swarm optimization algorithm and the random forest regression prediction model can be jointly used to efficiently train a preset battery health state prediction model with high model accuracy and good application effect by using the above sample data.

[0130] In specific implementation, ultrasonic signals can be emitted on the working plane of the sample battery by using the target ultrasonic transducer group; and waveform signals are collected at preset time intervals (for example, every 1 second) as sample data corresponding to the sample battery.

[0131] In specific implementation, according to the number of computing particles, multiple sample data can be randomly drawn from the sample data set with replacement to generate multiple sub-sample data sets. Different sub-sample data sets can be allocated for different computing particles.

[0132] The above sub-sample data set can be specifically expressed as: D (b) ={(x i ,y i )|i∈S b}. Wherein, D (b) is the sub-sample data set numbered b, S b is the sample data index set of the sub-sample data set, (x i ,y i ) represents a sample data numbered i, x i represents the sample feature numbered i of the sample battery, and y i represents the sample health state corresponding to the sample feature numbered i of the sample battery.

[0133] In specific implementation, an initial prediction model based on the random forest regression prediction model can be constructed according to the number of features included in the sample data in the sample data set.

[0134] Specifically, the formula [log 2 (M + 1)] can be used to determine the initial value m of the number of features involved in a single decision tree in the initial prediction model (i.e., the decision attribute feature). For example, when M takes the value of 3; correspondingly, m takes the value of 2. Furthermore, an initial prediction model based on the random forest regression prediction model can be constructed according to the initial value of the number of features.

[0135] In specific implementation, the initial prediction model can be assigned to each computing particle; then the computing particles are called to perform multiple rounds of iterative training respectively based on the initial prediction model and the corresponding sub-sample data set, and the prediction model is continuously optimized and updated until a preset battery health state prediction model that meets the requirements is obtained.

[0136] In some embodiments, referring to Figure 4 as shown, the above-mentioned calling of multiple computing particles for multiple rounds of iterative training based on the initial prediction model and the sub-sample data set may specifically include the following content when implemented:

[0137] S1: Call multiple computing particles to respectively determine the velocity parameters of the current round corresponding to the computing particles based on the local solution particle parameter group of the previous round and the global solution particle parameter group of the previous round; wherein, the particle parameter group at least includes: pruning threshold, number of decision trees, and proportion of test samples;

[0138] S2: Call multiple computing particles to respectively update the prediction model of the previous round based on the velocity parameters of the current round and the sub-sample data set to obtain the prediction model of the current round for each;

[0139] S3: Call multiple computing particles to respectively update the local solution particle parameter group of the current round according to the prediction model of the current round for each; and update the global solution particle parameter group of the current round according to the local solution particle parameter group of the current round;

[0140] S4: Detect whether the training end condition is satisfied currently;

[0141] S5: In the case of determining that the training end condition is satisfied, determine a preset battery health state prediction model that meets the requirements according to the global solution particle parameter group of the current round.

[0142] Based on the above embodiments, multiple computing particles can be used to continuously adjust and utilize the latest particle parameter groups through multiple rounds of iterative training, adjust and update the prediction models of each computing particle, and efficiently train a preset battery health state prediction model with better globality.

[0143] When specifically implemented, a pruning threshold (which can be denoted as ), the number of decision trees (which can be denoted as L), and the proportion of test samples (which can be denoted as X) can be combined as a particle parameter group, expressed as .

[0144] Among them, the above-mentioned pruning threshold and the number of decision trees can be used to optimize and update the prediction model during the model training process. The above-mentioned proportion of test samples can be used to process the sub-sample data set allocated to the computing particles to obtain specific training sample sets and test sample sets during the model training process, so as to perform training updates and test updates on the prediction model respectively.

[0145] When specifically implemented, the above-mentioned determination of the velocity parameters of the current round corresponding to the computing particles based on the local solution particle parameter group of the previous round and the global solution particle parameter group of the previous round may specifically include the following content when implemented:

[0146] Determine the velocity parameter of the current round of the calculation particle according to the following formula:

[0147] v id = wv id + c 1 * rand() * (p id - x id ) + c 2 * rand() * (p gd - x id )

[0148] Wherein, v id is the velocity parameter of the current round, w is the inertia weight, c 1 is the first acceleration coefficient, c 2 is the second acceleration coefficient, rand() represents a random function that varies within the range [0, 1], p id is the local solution particle parameter group (local optimal solution) of the previous round, p gd is the global solution particle parameter group (global optimal solution) of the previous round, x id is the sample feature in the sample data.

[0149] Specifically, the values of the above first acceleration coefficient and second acceleration coefficient can be 2, and the value of the inertia weight can be 0.8. The specific values of the above first acceleration coefficient, second acceleration coefficient, and inertia weight can be determined through big data analysis and learning of a large number of sample data.

[0150] In specific implementation, each calculation particle can be called to update the sample feature respectively using the velocity parameter of the current round and the sub-sample data set (for example, update the sample feature x according to the following formula id = x id + v id ), to obtain the updated sample data; then based on the global solution particle parameter group of the previous round, split the updated sample data into the training sample set and test sample set of the current round, and determine the model training parameters (including: pruning threshold, number of decision trees, etc.). Then use the model training parameters and the training sample set to train and update the prediction model of the previous round to obtain an updated and optimized prediction model; then use the test sample set to test and update the updated and optimized model until the obtained prediction model meets the preset training requirements, and end the update process of the current round of calculation particles to obtain the prediction model of the current round of this calculation particle.

[0151] Specifically, when updating the model using the training sample set and the test sample set for each computational particle respectively, for each decision tree, the training sample set and the test sample set can be split using the feature selection formula; then the decision tree is updated using the split training sample set and the split test sample set, and finally the corresponding regression tree is obtained, and then the prediction model for the current round of this computational particle is obtained.

[0152] Among them, the above feature selection formula can be expressed in the following form:

[0153]

[0154] Among them, j is the selected feature index, m is the number of sample data in the current node, y i is the true health status of the sample data numbered i, is the predicted value of the health status obtained after splitting according to feature j.

[0155] In specific implementation, each computational particle can calculate and detect the fitting effect of the prediction model according to the evaluation indexes mean squared error (MSE) and R-squared parameter of the prediction model, and then judge whether the prediction model meets the preset training requirements. If the evaluation indexes mean squared error and R-squared parameter calculated based on the prediction model are respectively less than the reference value of the preset evaluation index mean squared error and the reference value of the preset R-squared parameter, it can be determined that the prediction model meets the preset training requirements, and the prediction model is determined as the prediction model for the current round of this computational particle. On the contrary, if the evaluation index mean squared error calculated based on the prediction model is greater than or equal to the reference value of the preset evaluation index mean squared error, and / or, the R-squared parameter is greater than or equal to the reference value of the preset R-squared parameter, it can be determined that the prediction model does not meet the preset training requirements, and then the model can be continuously updated according to the training sample set and the test sample set.

[0156] Specifically, the evaluation index mean squared error and R-squared parameter can be calculated according to the following formulas:

[0157]

[0158] Among them, MSE is the evaluation index mean squared error, and R 2 is the squared parameter. The reference value of the preset evaluation index mean squared error and the reference value of the preset R-squared parameter can be specifically determined through statistical analysis according to historical data.

[0159] During specific implementation, each computing particle can determine the corresponding particle parameter group according to its prediction model in the current round; and compare the fitting effect determined by the particle parameter group based on the evaluation indexes of mean square error and R-squared parameter with the current local solution particle parameter group of the computing particle (i.e., the current local optimal solution). If the particle parameter group is better than the current local solution particle parameter group, replace the current local solution particle parameter group of the computing particle with the particle parameter group; otherwise, keep the current local solution particle parameter group unchanged, so as to update the local particle parameter group in the current round.

[0160] After updating the current local solution particle parameter group of the computing particle to the particle parameter group, the computing particle will be triggered to compare the current local solution particle parameter group with the current global solution particle parameter group (i.e., the current global optimal solution). If the current local solution particle parameter group is better than the current global solution particle parameter group, replace the current global solution particle parameter group with the current local solution particle parameter group of the computing particle; otherwise, keep the current global solution particle parameter group unchanged, so as to update the global solution particle parameter group.

[0161] During specific implementation, it can be detected whether the training end condition is satisfied currently. For example, by detecting whether the number of rounds in the current round exceeds the specified number of rounds, and / or by judging whether the model error of the optimal prediction model trained based on the global solution particle parameter group in the current round is less than the preset model error threshold, it is determined whether the training end condition is satisfied.

[0162] When it is determined that the training end condition is satisfied currently, a model that meets the requirements can be selected from the prediction models of multiple computing particles in the current round according to the global solution example parameter group in the current round as the battery health state prediction model that meets the requirements; and the training is ended.

[0163] On the contrary, when it is determined that the training end condition is not satisfied currently, the above process can be repeated to continue the iterative training in the next round until the training end condition is satisfied.

[0164] As can be seen from the above, before the specific implementation of the method for determining the battery health state provided in the embodiments of this specification, by jointly using the particle swarm optimization algorithm and the random forest regression prediction model, a preset battery health state prediction model with relatively high efficiency and high accuracy and good effect can be trained. During the specific implementation, a target ultrasonic transducer group can be deployed on the working surface of the target battery to be concerned; among them, the target ultrasonic transducer group includes 3 ultrasonic transducers, the distances between the 3 ultrasonic transducers are equal to each other, and the connecting sides form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery. In this way, the above-mentioned target ultrasonic transducer group can be used to simultaneously transmit ultrasonic signals to collect relatively comprehensive and complete waveform signals about the target battery. Then, according to the waveform signals, the corresponding time-domain data and frequency-domain data are obtained and utilized to determine the target features about the target battery; the preset battery health state prediction model is used to process the target features to obtain the corresponding target prediction results; and according to the target prediction results, the health state of the target battery is determined. Thus, the health state of the target battery can be monitored and determined more efficiently and accurately, the error can be effectively reduced, and the stable and safe operation of the target battery can be ensured.

[0165] Embodiments of this specification provide a server, refer to Figure 5 as shown. Among them, the server includes a network communication port 501, a processor 502, and a memory 503, and the above structures are connected by internal cables so that each structure can perform specific data interactions.

[0166] Among them, the network communication port 501 can specifically be used to receive monitoring instructions about the target battery.

[0167] The processor 502 can specifically be used to respond to the monitoring instructions, use the target ultrasonic transducer group to transmit ultrasonic signals on the working surface of the target battery; and collect the corresponding waveform signals; among them, the target ultrasonic transducer group includes 3 ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; according to the waveform signals, the corresponding time-domain data and frequency-domain data are obtained; according to the time-domain data and frequency-domain data, the target features about the target battery are determined; the preset battery health state prediction model is used to process the target features to obtain the corresponding target prediction results; among them, the preset battery health state prediction model is trained by jointly using the particle swarm optimization algorithm and the random forest regression prediction model; according to the target prediction results, the health state of the target battery is determined.

[0168] The memory 503 can specifically be used to store the corresponding instruction programs, as well as related data such as target features and the preset battery health state prediction model.

[0169] Based on the above method, the relevant structural performance of the server can be effectively utilized to improve the data processing speed of the electronic device and efficiently implement the data processing for determining the battery health state.

[0170] In this embodiment, the network communication port 501 can be bound to different communication protocols, so as to send or receive different data virtual ports. For example, the network communication port can be a port responsible for web data communication, or a port responsible for FTP data communication, or a port responsible for mail data communication. In addition, the network communication port can also be a physical communication interface or communication chip. For example, it can be a wireless mobile network communication chip, such as GSM, CDMA, etc.; it can also be a Wifi chip; it can also be a Bluetooth chip.

[0171] In this embodiment, the processor 502 can be implemented in any suitable manner. For example, the processor can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, application specific integrated circuit (ASIC), programmable logic controller, and embedded microcontroller, etc. This specification does not make a limitation.

[0172] In this embodiment, the memory 503 can include multiple levels. In a digital system, as long as it can store binary data, it can be a memory; in an integrated circuit, a circuit without a physical form but with a storage function is also called a memory, such as RAM, FIFO, etc.; in a system, a storage device with a physical form is also called a memory, such as a memory stick, TF card, etc.

[0173] This embodiment of the specification also provides a computer-readable storage medium based on the above method for determining the battery health state. The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed, it realizes: using a target ultrasonic transducer group to emit ultrasonic signals on the working surface of a target battery; and collecting corresponding waveform signals; wherein, the target ultrasonic transducer group includes 3 ultrasonic transducers, the connection sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; according to the waveform signals, obtaining corresponding time-domain data and frequency-domain data; according to the time-domain data and frequency-domain data, determining target features regarding the target battery; using a preset battery health state prediction model to process the target features to obtain a corresponding target prediction result; wherein, the preset battery health state prediction model is obtained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model; according to the target prediction result, determining the health state of the target battery.

[0174] In this embodiment, the above storage medium includes, but is not limited to, Random Access Memory (RAM), Read-Only Memory (ROM), Cache, Hard Disk Drive (HDD), or Memory Card. The memory can be used to store computer program instructions. The network communication unit can be set according to the standards specified by the communication protocol and is used as an interface for network connection communication.

[0175] In this embodiment, the functions and effects specifically implemented by the program instructions stored in the computer-readable storage medium can be explained by comparison with other embodiments and will not be elaborated here.

[0176] This embodiment of the specification also provides a computer program product, which at least includes a computer program. When the computer program is executed by a processor, the following method steps are implemented: transmitting an ultrasonic signal on the working surface of a target battery by using a target ultrasonic transducer group; and collecting corresponding waveform signals; wherein, the target ultrasonic transducer group includes three ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; obtaining corresponding time-domain data and frequency-domain data according to the waveform signals; determining target features about the target battery according to the time-domain data and the frequency-domain data; processing the target features by using a preset battery health state prediction model to obtain a corresponding target prediction result; wherein, the preset battery health state prediction model is trained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model; and determining the health state of the target battery according to the target prediction result.

[0177] Refer to Figure 6 As shown, this embodiment of the specification also provides a device for determining the health state of a battery. The device can specifically include the following structural modules:

[0178] An acquisition module 601, which can specifically be used to transmit an ultrasonic signal on the working surface of a target battery by using a target ultrasonic transducer group; and collect corresponding waveform signals; wherein, the target ultrasonic transducer group includes three ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery;

[0179] An acquisition module 602, which can specifically be used to obtain corresponding time-domain data and frequency-domain data according to the waveform signals;

[0180] A first determination module 603, which can specifically be used to determine target features about the target battery according to the time-domain data and the frequency-domain data;

[0181] A processing module 604, which can be specifically used to process target features by using a preset battery health state prediction model to obtain corresponding target prediction results; wherein, the preset battery health state prediction model is trained by jointly using a particle swarm optimization algorithm and a random forest regression prediction model;

[0182] A second determination module 605, which can be specifically used to determine the health state of the target battery according to the target prediction result.

[0183] In some embodiments, the target features may specifically include at least one of the following: average intensity attenuation ratio, average flight time, average main frequency offset ratio, etc.

[0184] In some embodiments, when the target features at least include the average intensity attenuation ratio, when the above-mentioned first determination module 603 is specifically implemented, the average intensity attenuation ratio of the target battery can be determined according to the following formula:

[0185]

[0186] Wherein, is the average intensity attenuation ratio, A i1 is the signal intensity of the signal with the first signal intensity value in the signal data received by the ultrasonic transducer numbered i, A i2 is the signal intensity of the signal with the second signal intensity value in the signal data received by the ultrasonic transducer numbered i, A i3 is the signal intensity of the signal with the third signal intensity value in the signal data received by the ultrasonic transducer numbered i, and A is the signal intensity of the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0187] In some embodiments, when the target features at least include the average flight time, according to the time-domain data and frequency-domain data, when the above-mentioned first determination module 603 is specifically implemented, the average flight time of the target battery can be determined according to the following formula:

[0188]

[0189] Wherein, is the average flight time, T i1 is the flight time of the signal with the first signal intensity value in the signal data received by the ultrasonic transducer numbered i, T i2 is the flight time of the signal with the second signal intensity value in the signal data received by the ultrasonic transducer numbered i, T i3 is the flight time of the signal with the third signal intensity value in the signal data received by the ultrasonic transducer numbered i.

[0190] In some embodiments, when the target feature at least includes the average main frequency offset ratio, when the above-mentioned first determination module 603 is specifically implemented, the average main frequency offset ratio of the target battery can be determined according to the following formula:

[0191]

[0192] Wherein, is the average main frequency offset ratio, f i1 is the frequency of the signal with the first signal intensity value in the signal data received by the ultrasonic transducer numbered i, f i2 is the frequency of the signal with the second signal intensity value in the signal data received by the ultrasonic transducer numbered i, f i3 is the frequency of the signal with the third signal intensity value in the signal data received by the ultrasonic transducer numbered i, and f is the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

[0193] In some embodiments, when the device is specifically implemented, it can also be used to: screen batteries in multiple different health states as sample batteries; deploy corresponding target ultrasonic transducer groups on the working plane of the sample batteries according to preset deployment rules, and use the target ultrasonic transducer groups to obtain sample data corresponding to the sample batteries through test experiments; use the sample data to construct corresponding sample data sets; determine a plurality of computing particles; randomly extract a plurality of sample data from the sample data set to generate a plurality of sub-sample data sets; and construct an initial prediction model based on a random forest regression prediction model; call a plurality of computing particles, and perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data sets until the training end condition is met, and obtain a preset battery health state prediction model that meets the requirements.

[0194] In some embodiments, when the device is specifically implemented, multiple computing particles can be called in the following manner to perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data set: Call multiple computing particles to respectively determine the velocity parameters of the current round corresponding to the computing particles based on the local solution particle parameter group of the previous round and the global solution particle parameter group of the previous round; wherein, the particle parameter group at least includes: pruning threshold, number of decision trees, and proportion of test samples; Call multiple computing particles to respectively update the prediction model of the previous round based on the velocity parameters of the current round and the sub-sample data set to obtain the prediction model of the current round for each; Call multiple computing particles to respectively update the local solution particle parameter group of the current round according to the prediction model of the current round for each; and update the global solution particle parameter group of the current round according to the local solution particle parameter group of the current round; Detect whether the training end condition is satisfied currently; In the case of determining that the training end condition is satisfied, determine a preset battery health state prediction model that meets the requirements according to the global solution particle parameter group of the current round.

[0195] It should be noted that the units, devices, or modules etc. illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. For the convenience of description, when describing the above devices, they are divided into various modules according to functions and described separately. Of course, when implementing this specification, the functions of each module can be realized in the same or multiple software and / or hardware, or the modules realizing the same function can be realized by a combination of multiple sub-modules or sub-units, etc. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0196] As can be seen from the above, based on the battery health state determination device provided in the embodiments of this specification, it is possible to monitor and determine the health state of the target battery more efficiently and accurately, effectively reduce errors, and ensure the stable and safe operation of the target battery.

[0197] In a specific scenario example, the battery health state determination method provided in this specification can be applied to implement the health state monitoring of sodium-ion batteries. The specific implementation process can include the following content.

[0198] In this scenario example, considering that ultrasonic immersion testing is usually performed on the battery to obtain the reflection coefficient angle spectrum of the lithium-ion battery at different states of charge (SOC), and then establishing the mapping relationship between the angle spectrum and the state of charge of the lithium-ion battery, the state of charge (SOC) of the lithium-ion battery is characterized by the peak distance between the two peaks of the angle spectrum, and then the health state of the battery is determined. However, when building an ultrasonic immersion testing platform using conventional methods, an angle fixture is mostly used to fix the angle between two ultrasonic probes to complete the layout of the ultrasonic transducers. This layout of the ultrasonic transducers will cause the ultrasonic signal to only pass through part of the battery area and cannot completely cover the entire area of the battery, resulting in only local measurement of the battery and unable to explore the overall situation of the battery, affecting the accuracy of the determined battery health state. In addition, the characteristic values collected based on the ultrasonic probes have great contingency, and different characteristic values affect each other and often show non-linear coupling, increasing the difficulty of regression prediction. Furthermore, the regression prediction model trained based on conventional methods has low accuracy and precision in evaluating the battery health status.

[0199] To address the above problems (limitations in the signal action area and contingency of characteristic values) and the root causes of these problems, the applicant proposes a non-destructive, efficient, low-cost, and high-precision method for monitoring the health state of sodium-ion batteries to achieve real-time and accurate assessment of the health state of sodium-ion batteries. Refer to Figure 7 as shown, it may include the following steps.

[0200] Step S1: Select sodium-ion batteries with different battery health conditions as the data sources for the training set and the test set for ultrasonic testing, and arrange three equally spaced ultrasonic transducers at the center area of the working surface of the battery to be tested (i.e., deploy the target ultrasonic transducer group); among them, the geometric center of the transducer coincides with the geometric center of the surface to be tested, ensuring that the sensors are on the same horizontal plane and parallel to the bottom surface of the working surface, and the transducer is connected to the computer.

[0201] Step S2: The three ultrasonic transducers simultaneously emit ultrasonic signals with a specific frequency and intensity, and simultaneously receive the signals emitted by themselves and the other two transducers. The waveform signals received by the ultrasonic transducers are filtered and amplified to obtain time-domain signals, and then Fourier-transformed to obtain frequency-domain signals. Calculate the characteristic values (such as target characteristics) of the processed time-domain signals and frequency-domain signals: average intensity attenuation ratio, average flight time, average main frequency offset ratio, and combine the battery health condition data as the test data set (such as the sample data set).

[0202] Step S3: Establish a random forest regression prediction model, and optimize the random forest parameters through the particle swarm algorithm to reduce the model training time and improve the model accuracy. Use the training set to train the random forest model (obtain the preset battery health state prediction model), and verify the model accuracy through the test set.

[0203] Step S4: Input the real-time ultrasonic signal eigenvalue of the sodium-ion battery to be measured into the random forest regression prediction model optimized by the particle swarm optimization algorithm to realize the real-time SOH prediction of the sodium-ion battery to be measured, and then determine the battery health state.

[0204] Specifically, the battery is a soft-pack sodium-ion battery.

[0205] Further, for the soft-pack sodium-ion battery, the working surface is the surface with the largest area of the battery, and its size is 10 cm * 7 cm.

[0206] Further, for the sodium-ion soft-pack battery in this embodiment, its positive electrode material is Prussian blue.

[0207] Further, in step S2, the calculation method of each eigenvalue is as follows:

[0208] Average intensity attenuation ratio . Wherein, A i1 , A i2 and A i3 are the intensities corresponding to the three highest frequencies of the i-th ultrasonic transducer, and A is the main frequency intensity emitted by the ultrasonic transducer. Since each ultrasonic transducer receives signals emitted by three transducers, each transducer receives signals with three main frequencies.

[0209] Average flight time . Wherein, T i1 , T i2 and T i3 are the signal flight times corresponding to the three highest frequencies of the i-th ultrasonic transducer.

[0210] Average main frequency offset ratio . Wherein, f i1 , f i2 and f i3 are the three highest frequencies of the i-th ultrasonic transducer, and f is the main frequency of the signal emitted by the ultrasonic transducer.

[0211] Further, in step S2, the signals received by the ultrasonic transducers are collected once every 1 s.

[0212] Further, the empirical function of the initial value m of the number of random feature attributes of the random forest regression prediction model optimized by the particle swarm optimization algorithm is [log 2 (M + 1)], where M is the total number of attributes, and the value is 3, so the initial value of m is 2.

[0213] During specific implementation, refer to Figure 8As shown, the training process of the random forest regression prediction model optimized by the particle swarm algorithm may include the following:

[0214] (1) Sample extraction: Randomly extract multiple samples from the original dataset with replacement to generate multiple sub-datasets. The sample extraction formula is:

[0215] D (b) ={(x i ,y i )|i∈S b}

[0216] where D (b) is the b-th sub-dataset. S b is the sample index set of the b-th sub-dataset.

[0217] (2) Training of decision trees: For each decision tree, use the randomly selected feature set for splitting to generate a regression tree. The feature selection formula is:

[0218]

[0219] where j is the feature index, m is the number of samples in the current node, y i is the true value of the i-th sample, is the predicted value after splitting according to feature j.

[0220] (3) Optimizing parameters by the particle swarm algorithm: Combine the pruning threshold , the number of decision trees L, and the pre-test sample rate X to form a spatial vector as a particle in the particle swarm algorithm. Iteratively optimize through the particle swarm optimization algorithm to select a relatively better particle, so as to select appropriate parameters to improve the accuracy of the final classification of the weighted random forest model. Among them, the velocity and vector update formulas of the particle are as follows:

[0221] v id =wv id +c 1 *rand()*(p id -x id )+c 2 *rand()*(p gd -x id )

[0222] x id =x id +v id

[0223] where w is the inertia weight, c 1 , c 2 are acceleration constants, and the value here is c 1= c 2 = 2. rand() is a random function that varies within the range [0, 1], and p id is the individual optimal position, and p gd is the global optimal position, x id is the position of the particle, and v id is the velocity of the particle.

[0224] Preferably, take w = 0.8, c 1 = c 2 = 2.

[0225] (4) Model prediction: Use the random forest model parameters obtained by the particle swarm optimization algorithm in (3) to generate multiple decision trees for combination, and its prediction function is expressed as:

[0226]

[0227] where k represents the k-th decision tree, x represents the input sample, and J k represents the number of leaf nodes of the -th decision tree, and c kj represents the prediction value of the j-th leaf node of the -th decision tree, and R kj represents the sample set of the -th leaf node of the -th decision tree, and K represents the number of decision trees.

[0228] (5) Model evaluation: Jointly use the mean squared error (MSE) and R-squared (R 2 ) metrics to detect the fitting effect of the model, and its calculation formula is as follows:

[0229]

[0230] where n represents the number of samples, y i represents the true value of the i-th sample, represents the predicted value of the i-th sample, represents the average value of all sample true values.

[0231] Specifically, the pairwise distance between the three ultrasonic transducers described in step S1 is 1.5 cm, and the central frequency of the ultrasonic transducer emission is 1 Mhz.

[0232] Specifically, the prediction results of the test set of the random forest regression prediction model optimized by the particle swarm algorithm can be referred to Figure 9 as shown.

[0233] Among them, the mean squared error (MSE) of the model evaluation index is 0.63, and R 2 is 0.9851. The predicted SOH is very close to the actual SOH within a large range, indicating that the model has high precision and high stability.

[0234] Through the above scenario example, the determination method of the battery health state provided in this specification is verified. First, the deployment method of the ultrasonic transducer is improved: three ultrasonic transducers are arranged at equal distances from each other in the central area of the working surface of the battery to be tested, and the geometric centers of the transducers coincide with the geometric center of the surface to be tested, ensuring that the sensors are on the same horizontal plane and parallel to the bottom surface of the working surface. This arrangement enables each ultrasonic transducer to simultaneously receive the signals emitted by itself and the other two transducers, expanding the signal coverage range, better identifying the internal information of the battery, and improving the stability and reliability of the collected signal characteristic values. Second, the used characteristic values are improved: the average intensity attenuation ratio, average flight time, and average main frequency offset ratio are calculated and jointly used for prediction. Furthermore, the model and algorithm are improved: the particle swarm optimization algorithm is introduced and used to optimize the random forest model. Specifically, multiple samples are randomly drawn from the original dataset with replacement to generate multiple sub-datasets; for each decision tree, the randomly drawn feature set is used for splitting to generate regression trees; then the pruning threshold , the number of decision trees L, and the pre-test sample rate X form a spatial vector as a particle in the particle swarm optimization algorithm. The relatively optimal particle is selected through iterative optimization of the particle swarm optimization algorithm, so as to select appropriate parameters to improve the accuracy of the final classification of the weighted random forest model.

[0235] Therefore, by adopting multiple ultrasonic characteristic values in the time domain and frequency domain, including the average intensity attenuation ratio, average flight time, and average main frequency offset ratio, and using the particle swarm optimization algorithm to iteratively optimize the parameters of the random forest algorithm model, the disadvantages of the random forest algorithm model with few characteristic quantities and unoptimized model parameters in the conventional method are overcome, and the prediction of the battery SOH can be more accurately realized; at the same time, by arranging three ultrasonic transducers at equal distances from each other at the geometric center of the battery surface to be tested and having both the function of transmitting and receiving signals, the signal covers the entire battery surface to be tested, overcoming the defect that the conventional method can only measure the local signals of the battery, and realizing the accurate prediction of the overall SOH of the battery.

[0236] Although this specification provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-creative means. The order of steps listed in the embodiments is only one way among the execution orders of numerous steps and does not represent the only execution order. When the actual device or client product is executed, it can be executed in the order of the method shown in the embodiments or the drawings or executed in parallel (for example, in a parallel processor or multi-threaded processing environment, or even in a distributed data processing environment). The terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, product or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, product or device. Without further limitation, there is no exclusion of additional identical or equivalent elements in the process, method, product or device comprising the said elements. The terms such as first, second, etc. are used to denote names and do not denote any particular order.

[0237] Those skilled in the art also know that in addition to implementing the controller in the form of pure computer-readable program code, the method steps can be logically programmed to enable the controller to implement the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Therefore, such a controller can be regarded as a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or the structures within the hardware component.

[0238] This specification can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, classes, etc. that perform specific tasks or implement specific abstract data types. This specification can also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer-readable storage media including storage devices.

[0239] As can be seen from the description of the above embodiments, those skilled in the art can clearly understand that this specification can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of this specification can essentially be embodied in the form of a software product, which can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, mobile terminal, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments of this specification.

[0240] The embodiments in this specification are described in a progressive manner. For the same or similar parts between the embodiments, reference can be made to each other. The key point of each embodiment is to illustrate the differences from other embodiments. This specification can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, and so on.

[0241] Although this specification is depicted through embodiments, those of ordinary skill in the art know that this specification has many variations and changes without departing from the spirit of this specification. It is hoped that the appended claims will cover these variations and changes without departing from the spirit of this specification.

Claims

1. A method for determining a battery health status, characterized in that: include: Using a target ultrasonic transducer group to transmit an ultrasonic signal on a working surface of a target battery; and collecting a corresponding waveform signal; wherein the target ultrasonic transducer group includes three ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; According to the waveform signal, obtaining corresponding time domain data and frequency domain data; Determine a target feature of the target battery according to the time domain data and the frequency domain data; the target feature includes at least one of the following: average intensity attenuation ratio, average flight time, and average main frequency deviation ratio; The target features are processed using a preset battery health state prediction model to obtain a corresponding target prediction result; wherein the preset battery health state prediction model is obtained by training by jointly using a particle swarm optimization algorithm and a random forest regression prediction model; According to the target prediction results, the health status of the target battery is determined.

2. The method according to claim 1, characterized in that: In the case where the target feature includes at least the average intensity attenuation ratio, the target feature of the target battery is determined according to the time domain data and the frequency domain data, including: Determine the average intensity attenuation ratio of the target battery according to the following formula: in, is the average intensity attenuation ratio, A i1 is the signal strength of the signal with the highest signal strength value in the signal data received by the ultrasonic transducer numbered i, A i2 is the signal strength of the signal whose signal strength value ranks second in the signal data received by the ultrasonic transducer numbered i, A i3 is the signal strength of the signal ranked third in signal strength value in the signal data received by the ultrasonic transducer numbered i, and A is the signal strength of the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

3. The method according to claim 1, characterized in that When the target feature includes at least the average flight time, the target feature of the target battery is determined based on the time domain data and the frequency domain data, including: Determine the average flight time of the target battery using the following formula: in, is the average flight time, T i1 T is the flight time of the signal with the highest signal strength value in the signal data received by the ultrasonic transducer numbered i, i2 T is the flight time of the signal with the second highest signal strength value in the signal data received by the ultrasonic transducer numbered i, i3 It is the flight time of the signal with the third highest signal strength value in the signal data received by the ultrasonic transducer numbered i.

4. The method according to claim 1, characterized in that: When the target feature includes at least the average main frequency offset ratio, the target feature of the target battery is determined according to the time domain data and the frequency domain data, including: Determine the average main frequency deviation ratio of the target battery according to the following formula: in, is the average main frequency deviation ratio, f i1 is the frequency of the signal with the highest signal strength value in the signal data received by the ultrasonic transducer numbered i, f i2 is the frequency of the signal whose signal strength value ranks second in the signal data received by the ultrasonic transducer numbered i, f i3 is the frequency of the signal with the third highest signal strength value in the signal data received by the ultrasonic transducer numbered i, and f is the frequency of the main frequency signal of the ultrasonic signal emitted by the ultrasonic transducer.

5. The method according to claim 1, characterized in that The method further comprises: Screening multiple batteries in different health states as sample batteries; According to the preset deployment rules, a corresponding target ultrasonic transducer group is deployed on the working plane of the sample battery, and the target ultrasonic transducer group is used to perform a test experiment to obtain sample data corresponding to the sample battery; Using the sample data, construct a corresponding sample data set; A plurality of computational particles are determined; A plurality of sample data are randomly extracted from the sample data set to generate a plurality of sub-sample data sets; and an initial prediction model based on the random forest regression prediction model is constructed; Multiple computing particles are called to perform multiple rounds of iterative training based on the initial prediction model and the sub-sample data set until the training end conditions are met, and a preset battery health status prediction model that meets the requirements is obtained.

6. The method according to claim 5, characterized in that Call multiple computing particles to perform multiple rounds of iterative training based on the initial prediction model and sub-sample datasets, including: Calling multiple computing particles to determine the speed parameters of the current round corresponding to the computing particles based on the local solution particle parameter group of the previous round and the global solution particle parameter group of the previous round; wherein the particle parameter group includes at least: pruning threshold, number of decision trees, and test sample ratio; Call multiple computing particles to update their respective prediction models of the previous round based on the speed parameters and sub-sample data sets of the current round, and obtain their respective prediction models of the current round; Call multiple computing particles to update the local solution particle parameter group of the current round according to their respective prediction models of the current round; and update the global solution particle parameter group of the current round according to the local solution particle parameter group of the current round; Check whether the training end conditions are met; When it is determined that the training end conditions are met, a preset battery health status prediction model that meets the requirements is determined based on the global solution particle parameter group of the current round.

7. A device for determining a battery health status, characterized in that: include: A collection module, used to transmit ultrasonic signals on the working surface of the target battery using a target ultrasonic transducer group; and collect corresponding waveform signals; wherein the target ultrasonic transducer group includes three ultrasonic transducers, the connecting sides of the ultrasonic transducers form an equilateral triangle, and the center of the equilateral triangle coincides with the center of the working surface of the target battery; An acquisition module, used for acquiring corresponding time domain data and frequency domain data according to the waveform signal; A first determination module is used to determine a target feature of a target battery according to the time domain data and the frequency domain data; the target feature includes at least one of the following: an average intensity attenuation ratio, an average flight time, and an average main frequency offset ratio; A processing module, used to process target features using a preset battery health state prediction model to obtain a corresponding target prediction result; wherein the preset battery health state prediction model is obtained by training by jointly using a particle swarm optimization algorithm and a random forest regression prediction model; The second determination module is used to determine the health status of the target battery according to the target prediction result.

8. A server, characterized in that: The method comprises a processor and a memory for storing processor-executable instructions, wherein the processor implements the steps of the method according to any one of claims 1 to 6 when executing the instructions.

9. A computer-readable storage medium, characterized in that: Computer instructions are stored thereon, and when the instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images