5G Internet of Things data acquisition optimization method and system

Abstract

The invention relates to the technical field of storage battery pack data acquisition, in particular to a 5G Internet of Things data acquisition optimization method and system, and the method comprises the steps: obtaining an abnormal region of an abnormal storage battery pack through the analysis of the working condition data of a storage battery; analyzing the working condition data of the single storage battery in the abnormal region to obtain the fault probability of the single storage battery; in addition, the influence coefficient between the storage battery packs of the same scale is analyzed; and finally, the sampling frequency is adjusted in combination with the fault probability and the influence coefficient. According to the method and the system, on one hand, the data acquisition quantity can be reduced, and the uploading of invalid data is reduced, so that the network bandwidth pressure is effectively reduced, and the transmission pressure is improved; on the other hand, effectiveness and reliability of abnormal area data can be ensured, and the abnormal area data can be timely and accurately recognized and fed back.

Description

Technical Field

[0001] The present invention relates to the technical field of battery pack data collection, and in particular to a 5G Internet of Things data collection optimization method and system. Background Art

[0002] With the continuous development of 5G technology, the Internet of Things (IoT) has entered a new stage, providing users with higher network speeds, lower latency, and larger-capacity communication services. The larger-capacity communication services enable the Internet of Things to connect to more network devices, further improving the convenience of users in work and life.

[0003] A battery is a device that can store electrical energy and release it when needed. Multiple batteries form a battery pack. For a battery pack, due to long-term charging and discharging, and the battery performance is greatly affected by the environment, the health status of each battery usually varies. In order to ensure the safety and reliability of the battery pack during use, it is necessary to collect battery data for each battery in each battery pack in a factory or system, and upload it to the central server through the Internet of Things to determine the health status of each battery.

[0004] The existing data collection method (edge ​​computing) can perform preliminary processing on the collected battery data at a location close to the data source (battery), such as data screening, compression, aggregation, etc., thereby reducing the amount of data collection transmission while sharing part of the data processing pressure of the IoT central server, and achieving more efficient data collection. However, in the actual application of this data collection method, in order to upload more abundant data to the central server, data collection is usually carried out on a large scale, and there is a large amount of redundant information or invalid information in the uploaded data, which will consume a large amount of computing resources of the central server to identify the data, resulting in reduced data processing efficiency. In addition, when the battery fails, it cannot be quickly discovered and located, resulting in a slow response speed when an accident occurs. At the same time, excessive data volume may cause data corruption and data loss when compressed and uploaded. Summary of the invention

[0005] In order to solve the above technical problems, the purpose of the present invention is to provide a 5G Internet of Things data collection optimization method and system.

[0006] According to a first aspect of an embodiment of the present invention, a 5G Internet of Things data collection optimization method is provided, and the technical solution adopted is as follows: Setting an initial sampling frequency to collect multi-dimensional operating data of the battery pack, wherein the multi-dimensional operating data includes operating voltage and operating current; Analyzing the first state abnormality index of the battery pack according to the operating voltage and the operating current to obtain an abnormal battery pack; Analyze the voltage change trend and current change trend of the abnormal battery group to obtain a local abnormality judgment coefficient and determine a possible abnormal area; Analyze the second state abnormality index of the possible abnormal area to determine the abnormal area; For the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient; Clustering the battery groups to obtain clusters, analyzing the discreteness between all the battery groups in the clusters, and combining the failure probability to obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group; According to the influence coefficient and the failure probability, the initial sampling frequency is adjusted to obtain a corrected sampling frequency.

[0007] In some embodiments of the present invention, analyzing the first state abnormality index of the battery pack according to the operating voltage and the operating current to obtain the abnormal battery pack includes: According to the working voltage and the working current of the battery pack at the current moment, combined with the theoretical voltage and the theoretical current of the battery pack, a first state abnormality index of the battery pack at the current moment is obtained; Presetting a first abnormality index threshold; Determine whether the first state abnormality index is greater than the first abnormality index threshold; If yes, the battery pack corresponding to the first state abnormality index is an abnormal battery pack.

[0008] In some embodiments of the present invention, analyzing the voltage change trend and the current change trend of the abnormal battery group to obtain a local abnormality judgment coefficient and determine a possible abnormal area includes: Based on the working voltage and the working current of the battery pack at the current moment and the previous moment, analyzing the voltage fitting slope and the current fitting slope at the current moment, and obtaining the voltage change trend and the current change trend of the abnormal battery pack; According to the voltage variation trend, combined with the working current and theoretical current of the battery pack at the current moment, a series abnormality coefficient is obtained; According to the voltage variation trend and the current variation trend, a parallel abnormality coefficient is obtained; Obtaining a local abnormality determination coefficient according to the series abnormality coefficient and the parallel abnormality coefficient; According to the local abnormality judgment coefficient, a possible abnormal area is determined.

[0009] In some embodiments of the present invention, analyzing the second state abnormality index of the possible abnormal area to determine the abnormal area includes: According to the working voltage and the working current of the possible abnormal area at the current moment, combined with the theoretical voltage and the theoretical current of the possible abnormal area, a second state abnormality index of the possible abnormal area at the current moment is obtained; Presetting a second abnormality index threshold; Determine whether the second state abnormality index is greater than the second abnormality index threshold; If yes, the possible abnormal area corresponding to the second state abnormality index is the abnormal area.

[0010] In some embodiments of the present invention, for the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the method further includes: According to the second state abnormality index corresponding to the abnormal area, combined with the initial sampling frequency, a temporary sampling frequency of the abnormal area is obtained; According to the temporary sampling frequency, the latest multi-dimensional working data of the abnormal area is re-acquired.

[0011] In some embodiments of the present invention, for the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient, including: For the single storage battery in the abnormal area, calculate the mean of the working data in each dimension, analyze the difference in the working data in each dimension between the current single storage battery and other single storage batteries in the abnormal area, and obtain the relative deviation of the working data in each dimension; According to the relative deviation of the working data of each dimension, the total deviation of the single battery is obtained; The failure probability of the single storage battery is obtained according to the total deviation and in combination with the local abnormality judgment coefficient of the local area where the single storage battery is located.

[0012] In some embodiments of the present invention, the battery groups are clustered to obtain clusters, the discreteness between all the battery groups in the clusters is analyzed, and the influence coefficient of other battery groups in the clusters on the current abnormal battery group is obtained in combination with the failure probability, including: According to the scale of the battery group, cluster analysis is performed on the battery group to obtain cluster clusters; Calculate the mean value of the working data of each battery group in each dimension in the cluster where the current abnormal battery group is located; Calculate the variance of the mean value of each dimension of working data among all battery groups in the cluster where the current abnormal battery group is located, obtain the degree of discreteness of the dimensional data of each dimension of working data among all battery groups in the cluster, and then obtain the degree of discreteness among all battery groups in the cluster; According to the discrete degree and in combination with the fault probability, the influence coefficient of other storage battery groups in the cluster on the current abnormal storage battery group is obtained.

[0013] In some embodiments of the present invention, the influence coefficient of other battery groups in the cluster on the current abnormal battery group is obtained according to the discrete degree and in combination with the fault probability, including: According to the failure probability of the single battery, the failure probability of the current abnormal battery group is calculated, and the average failure probability of all battery groups in the cluster where the current abnormal battery group is located is calculated to obtain the abnormal degree of the failure probability of the current abnormal battery group; According to the discrete degree and in combination with the abnormal degree of the fault probability, the influence coefficient of other storage battery groups in the cluster on the current abnormal storage battery group is obtained.

[0014] According to a second aspect of an embodiment of the present invention, a 5G Internet of Things data collection optimization system is provided, including: a memory and a processor, wherein: The memory is used to store program codes; The processor is used to read the program code stored in the memory and execute the method described in the first aspect of the embodiment of the present invention.

[0015] In some embodiments of the present invention, the processor comprises: A working data acquisition module, used to set an initial sampling frequency and collect multi-dimensional working data of the battery pack, wherein the multi-dimensional working data includes working voltage and working current; an abnormal region determination module, configured to analyze a first state abnormality index of the battery pack according to the operating voltage and the operating current, and obtain an abnormal battery pack; A fault probability analysis module is used to analyze the voltage change trend and current change trend of the abnormal battery group, obtain the local abnormality judgment coefficient, and determine the possible abnormal area; and analyze the second state abnormality index of the possible abnormal area to determine the abnormal area; finally, for the single battery in the abnormal area, analyze the relative deviation of the working data of each dimension, and combine the local abnormality judgment coefficient to obtain the fault probability of the single battery; An influence coefficient analysis module is used to cluster the battery groups to obtain clusters, analyze the discreteness between all the battery groups in the clusters, and obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group in combination with the failure probability; The sampling frequency adjustment module is used to adjust the initial sampling frequency according to the influence coefficient and the fault probability to obtain a modified sampling frequency.

[0016] Compared with the prior art, the 5G Internet of Things data collection optimization method and system provided by the present invention has the following beneficial effects: 1. The present invention obtains an abnormal battery pack by analyzing the abnormal index of the working voltage and the working current, and determines the possible abnormal area by analyzing the voltage change trend and the current change trend. The abnormal area is further determined, and the abnormal area (certain series area or parallel area) of the battery pack can be accurately located; the data sampling frequency of the sensor is increased only for the abnormal area, on the one hand, the amount of data collected can be reduced, and the upload of invalid data can be reduced, thereby effectively reducing the pressure on the network bandwidth and increasing the transmission pressure; on the other hand, the validity and reliability of the data in the abnormal area can be ensured, so that the data in the abnormal area can be timely and accurately identified and fed back.

[0017] 2. The present invention analyzes the failure probability of single batteries in abnormal areas and the influence relationship between battery groups of the same size, and then obtains the adjustment coefficient of the data sampling frequency of the sensor to ensure the validity and reliability of the collected data samples.

[0018] 3. The edge computing node adjusts the data acquisition parameters of the sensor according to the real-time working conditions of the battery pack, and uploads the acquired battery data according to the situation; for battery packs with a low probability of failure, only the overall data is uploaded, and the data of the individual batteries is stored under the edge computing node; for battery packs with a high probability of failure, the data of each individual battery is uploaded, which can reduce the upload of invalid data and ensure timely and accurate identification and feedback of abnormal battery packs and abnormal batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A basic flow chart of a 5G Internet of Things data collection optimization method provided by an embodiment of the present invention; Figure 2 A schematic diagram of the basic composition of a 5G Internet of Things data acquisition and optimization system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a 5G Internet of Things data collection optimization method and system proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. Terms such as "comprises", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such article or device. In the absence of further restrictions, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the article or device including the element.

[0023] The specific scenario targeted by the present invention is: a battery is a device that can store electrical energy and release it when needed. It has very wide applications in the fields of industry, transportation, and power grids, and usually plays a key role in ensuring energy supply, improving energy efficiency, and supporting the use of renewable energy. In some places where there is a large demand for use, multiple batteries are connected in series to obtain a battery pack, thereby increasing the battery's electrical energy storage capacity.

[0024] Since batteries or battery packs are charged and discharged for a long time, and the performance of batteries is greatly affected by the environment, there are usually some differences in the health status of batteries. In order to ensure the safety and reliability of batteries during use, it is necessary to collect battery data for each battery in each battery pack in a factory or system, and upload it to the central server through the Internet of Things to determine the health status of each battery.

[0025] The data dimensions of battery health status include voltage, current, temperature, battery capacity, etc. Therefore, multi-dimensional data and a large number of batteries will result in a large amount of data being uploaded to each node. When these data are aggregated to the central server, it will take a lot of time to process. When a battery fails, it cannot be discovered and located quickly, resulting in a slow response speed when an accident occurs. At the same time, excessive data volume may cause data corruption or loss when it is compressed and uploaded.

[0026] The purpose of the present invention is to monitor the health status and predict faults of battery packs by analyzing the working conditions of batteries in edge computing nodes. At the same time, it adjusts the amount of data uploaded by nodes to the central server, reduces the pressure of data processing and uploading, and improves the overall data collection and processing efficiency.

[0027] The following is a detailed description of a 5G Internet of Things data collection optimization method provided by the present invention in conjunction with the accompanying drawings.

[0028] See also Figure 1 , which shows the basic process of a 5G Internet of Things data collection optimization method provided by an embodiment of the present invention.

[0029] like Figure 1 As shown, a 5G Internet of Things data collection optimization method provided by an embodiment of the present invention specifically includes: S100: setting an initial sampling frequency to collect multi-dimensional operating data of the battery pack, where the multi-dimensional operating data includes operating voltage and operating current.

[0030] In actual applications, battery packs are usually more common than single batteries, especially in high-power and large-capacity applications, while single batteries are usually used as backup power sources for certain functions in the system, ensuring that when the battery pack fails, the single battery can ensure that some functions can still be used for a short time. Therefore, in the embodiments of the present invention, the battery pack is used as the object for analysis.

[0031] Set the battery data of the battery or battery group to be collected once every 1 minute, that is, the data collection frequency times / hour. Select sensors to collect the required multi-dimensional working data of the battery or battery pack under each edge computing node through the sensors. The multi-dimensional working data includes voltage data and current data, as well as temperature data and internal resistance data. Among them: 1) Voltage sensor: installed at both ends (positive and negative) of each battery, collects voltage data of the entire battery pack and each battery in the battery pack to obtain the group voltage and single cell voltage; 2) Current sensor: installed at the charging end and the discharging end, respectively collects the current data flowing into and out of the entire battery pack and each single battery in the battery pack to obtain the pack current and single battery current; 3) Infrared temperature sensor: Installed directly opposite the battery or directly above the battery pack, it acquires the surface temperature data of the entire battery pack and each single battery in the battery pack to obtain the pack temperature and single battery temperature; 4) Internal resistance sensor: installed at both ends (positive and negative) of the battery cell, it acquires the internal resistance data of the entire battery pack and each single battery in the battery pack to obtain the group internal resistance and single battery internal resistance.

[0032] Based on the obtained voltage data, current data, temperature data and internal resistance data, the capacity of the entire battery pack and each single battery in the battery pack is estimated according to existing methods (such as an integration method based on the charging / discharging process).

[0033] Then all the collected multi-dimensional working data of single cells or battery groups are stored in the designated storage area of ​​the edge computing node according to the collection order.

[0034] In addition, the size of the battery group corresponding to each edge computing node is also obtained, that is, the number of single batteries contained in the battery group.

[0035] At this point, the collection of multi-dimensional working data of the battery pack has been completed.

[0036] The battery pack is composed of multiple single cells connected in series and parallel. When a single cell in the battery pack is abnormal, the single cells connected in series or parallel with it will also have abnormal working data. By analyzing the abnormal value and the connection mode between two or more single cells, the real abnormal point (single cell) in the battery pack is located. The specific steps include S200 to S400.

[0037] S200: Analyze the first state abnormality index of the battery pack according to the operating voltage and the operating current to obtain an abnormal battery pack.

[0038] According to the working voltage and the working current, the abnormal index of the first state of the battery pack is analyzed to obtain the abnormal battery pack. The specific implementation method is as follows: First, according to the working voltage and working current of the battery pack at the current moment, that is, the group voltage and group current of the battery pack at the current moment, combined with the theoretical voltage and theoretical current of the battery pack, the first state abnormality index of the battery pack at the current moment is obtained, wherein the calculation formula for constructing the first state abnormality index of the battery pack at the current moment is:

[0039] In the formula, Indicates The current time of the battery pack The first state abnormality index; Indicates The current time of the battery pack The group current; Indicates The current time of the battery pack The group voltage; Indicates Theoretical current of a battery pack; Indicates Theoretical voltage of a battery pack; represents the linear normalization function.

[0040] Then, a first abnormal index threshold is preset, and the value of the first abnormal index threshold may be 0.4; and, it is determined whether the first state abnormal index is greater than the first abnormal index threshold; if so, that is, , indicating that the working state of the battery pack corresponding to the first state abnormality index is abnormal, then the current A battery pack is an abnormal battery pack.

[0041] S300: Analyze the voltage change trend and current change trend of the abnormal battery group, obtain the local abnormality judgment coefficient, and determine the possible abnormal area.

[0042] The series-parallel combination design achieves the required voltage by connecting multiple single batteries in series, and then achieves the required capacity and current by connecting multiple series battery groups in parallel. When the group voltage drops and the group current is low, it means that the series part of the battery group is abnormal; when the group voltage is relatively stable and the group current changes greatly, it means that the parallel part of the battery group is abnormal.

[0043] Therefore, by analyzing the voltage change trend and current change trend of the abnormal battery group, the local abnormality judgment coefficient is obtained to determine the possible abnormal area. First, based on the working voltage and working current of the battery pack at the current moment and the previous moment, the voltage fitting slope and current fitting slope at the current moment are analyzed, that is, the current moment of the battery pack is converted into The group voltage and group current obtained at the previous moment respectively construct the time-voltage curve and the time-current curve, and obtain the current moment on the time-voltage curve and the time-current curve The voltage fitting slope and current fitting slope at the position are as follows, wherein the voltage fitting slope is the voltage variation trend of the abnormal battery group, and the current fitting slope is the current variation trend of the abnormal battery group.

[0044] Then, according to the voltage change trend, combined with the working current and theoretical current of the battery pack at the current moment, the series abnormal coefficient is obtained to construct the first Abnormal battery pack at the current time The calculation formula of the series anomaly coefficient is:

[0045] In the formula, Indicates Abnormal battery pack at the current time The series anomaly coefficient of Indicates Abnormal battery pack at the current time The voltage fitting slope of Indicates Theoretical current of an abnormal battery pack; Indicates Abnormal battery pack at the current time The group current.

[0046] At the same time, according to the voltage change trend and current change trend, the parallel abnormal coefficient is obtained to construct the first Abnormal battery pack at the current time The calculation formula of the parallel abnormal coefficient is:

[0047] In the formula, Indicates Abnormal battery pack at the current time The parallel anomaly coefficient of Indicates Abnormal battery pack at the current time The voltage fitting slope of Indicates Abnormal battery pack at the current time The current fitting slope of represents the first correction parameter, which is used to prevent the denominator from being 0. In this embodiment, .

[0048] For the series anomaly coefficient , when the value is negative and the data is smaller, the probability of series anomaly is greater; for the parallel anomaly coefficient , when the value is larger, the probability of parallel abnormality is greater.

[0049] Then, according to the series anomaly coefficient and the parallel anomaly coefficient, the local anomaly judgment coefficient is obtained to construct the first Abnormal battery pack at the current time The calculation formula of the local abnormality judgment coefficient is:

[0050] In the formula, Indicates Abnormal battery pack at the current time The local abnormality judgment coefficient; Indicates Abnormal battery pack at the current time The series anomaly coefficient of Indicates Abnormal battery pack at the current time The parallel anomaly coefficient of is a normalization function, and the normalized value range is [-1,1].

[0051] Finally, the possible abnormal area is determined according to the local abnormality judgment coefficient, that is, the magnitude relationship between the local abnormality judgment coefficient and 0 is determined to determine whether the series area or the parallel area is abnormal. When the abnormal battery group shows abnormal series connection, the local abnormality judgment coefficient of the abnormal battery group is When the abnormal battery group shows abnormal parallel connection; when the local abnormality judgment coefficient of the abnormal battery group , the abnormal battery pack behaves as normal.

[0052] S400: Analyze the second state abnormality index of the possible abnormal area to determine the abnormal area.

[0053] According to the local abnormality judgment coefficient of the abnormal battery group The possible abnormal area of ​​the abnormal battery pack can be determined, that is, the parallel area is abnormal or the series area is abnormal. However, for an abnormal battery pack, there may be multiple parallel areas or multiple series areas. The above cannot determine which specific parallel area or which specific series area, and further analysis is required.

[0054] Therefore, the abnormal area is determined by analyzing the second state abnormal index of the possible abnormal area. First, according to the working voltage and working current of the possible abnormal area at the current moment, that is, the local voltage and local current of the possible abnormal area at the current moment, combined with the theoretical voltage and theoretical current of the possible abnormal area, the second state abnormal index of the possible abnormal area at the current moment is obtained, wherein the calculation formula for constructing the second state abnormal index of the possible abnormal area at the current moment is:

[0055] In the formula, Indicates Possible abnormal area at the current time The second state abnormality index; Indicates Possible abnormal area at the current time The group current; Indicates Possible abnormal area at the current time The group voltage; Indicates Theoretical current in possible abnormal areas; Indicates Theoretical voltage of possible abnormal areas; represents the linear normalization function.

[0056] Then, a second abnormal index threshold is preset, and the value of the second abnormal index threshold may be 0.4; and, whether the second state abnormal index is greater than the second abnormal index threshold is determined; if so, that is, , indicating that the working state of the possible abnormal area corresponding to the second state abnormality index is abnormal, then the current The possible abnormal area is the abnormal area.

[0057] S500: For the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient.

[0058] The abnormal battery group is determined through the above steps, and the specific abnormal area in the abnormal battery group is determined, but the abnormal area is a parallel area or a series area, that is, the abnormal area includes multiple single batteries, and it may be that one or several single batteries are abnormal, causing the entire parallel area or series area to show abnormality. Therefore, it is necessary to further analyze the failure probability of each single battery in the abnormal area.

[0059] In the embodiment of the present invention, for the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient.

[0060] For the single battery in the abnormal area, since the initial sampling frequency is 60 times / hour, the sampling frequency is relatively low. It is possible that the single battery did not show abnormality in the previous minute, but just showed abnormality in the current minute. If the existing multi-dimensional working data is directly used for analysis, there may be a large deviation. Therefore, in some embodiments of the present invention, for the single battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, which also includes: according to the second state abnormality index corresponding to the abnormal area, combined with the initial sampling frequency, the temporary sampling frequency of the abnormal area is obtained, and the first The calculation formula for the temporary sampling frequency of an abnormal area is:

[0061] In the formula, Indicates Temporary sampling frequency of abnormal areas; Indicates The current time of the abnormal area The second state abnormality index; Indicates the initial sampling frequency. That is, the second state abnormality index The larger it is, the larger the corresponding temporary sampling frequency is.

[0062] According to the temporary sampling frequency, the latest multi-dimensional working data of the abnormal area and each single battery in the abnormal area are re-acquired. The time for data acquisition according to the temporary sampling frequency can be reasonably set as needed, such as 10 minutes.

[0063] After obtaining the latest multi-dimensional working data, for the single battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single battery is obtained by combining the local abnormality judgment coefficient. The specific implementation method is as follows: First, for the single battery in the abnormal area, the mean of the latest working data in each dimension is calculated, and the difference in working data in each dimension between the current single battery and other single batteries in the abnormal area is analyzed to obtain the relative deviation of working data in each dimension. In the abnormal area The first Dimension work data at the current moment The relative deviation calculation formula is:

[0064] In the formula, Indicates In the abnormal area The first Dimension work data at the current moment The relative deviation of Indicates The number of single batteries in the abnormal area; Indicates In the abnormal area Single battery The mean of the latest work data in each dimension; Indicates In the abnormal area, except Other than single battery Single battery The average of the latest work data in each dimension.

[0065] By analyzing the difference in the latest working data of each dimension between the current single battery and other single batteries in the abnormal area, the current working data of each dimension of the current single battery can be obtained. relative deviation.

[0066] It should be noted that when a certain dimensional data (such as voltage, current, temperature, etc.) of a single battery is abnormal, other dimensional data may not necessarily show abnormalities, that is, different dimensional data of a single battery are usually independent, so each dimensional data is analyzed as independent dimensional data.

[0067] According to the relative deviation of the working data of each dimension, the total deviation of the single battery is obtained, that is, the relative deviation mean of the working data of all dimensions of the single battery.

[0068] Then, according to the total deviation, combined with the local abnormality judgment coefficient of the local area where the single battery is located, the failure probability of the single battery is obtained, and the first In the abnormal area The current time of each single battery The failure probability calculation formula is:

[0069] In the formula, Indicates In the abnormal area The current time of each single battery The probability of failure; Indicates In the abnormal area The current time of each single battery Total deviation of Indicates In the abnormal area The current time of the abnormal battery group where the single battery is located The local abnormality judgment coefficient; represents the linear normalization function.

[0070] S600: clustering the battery groups to obtain clusters, analyzing the discreteness between all the battery groups in the clusters, and combining the failure probability to obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group.

[0071] If the size and quantity of battery groups between multiple edge computing nodes are quite different, the mutual influence between them is small and can be ignored. However, if the size and quantity of battery groups are similar, the mutual influence between them is large. And if the working conditions of the battery groups corresponding to each edge computing node are similar, the mutual influence between them will also be large. When the influence between the battery groups is large, it is necessary to increase the sampling frequency and obtain more working data samples to improve the accuracy of data analysis.

[0072] Based on the above analysis, in an embodiment of the present invention, the battery groups are clustered to obtain clusters, the discreteness between all battery groups in the clusters is analyzed, and the influence coefficient of other battery groups in the clusters on the current abnormal battery group is obtained in combination with the failure probability. The specific implementation method is as follows: First, according to the size of the battery group, the battery group is clustered and analyzed to obtain clusters. Then, the mean of the working data of each dimension of each battery group in the cluster where the current abnormal battery group is located is calculated; and the variance of the mean of the working data of each dimension between all battery groups in the cluster where the current abnormal battery group is located is calculated to obtain the dimensional data dispersion degree of each dimensional working data between all battery groups in the cluster, and then the dispersion degree between all battery groups in the cluster is obtained, which is the mean of the dimensional data dispersion degree of all dimensional working data. Finally, according to the degree of discreteness and the failure probability, the influence coefficient of other battery groups in the cluster on the current abnormal battery group is obtained; specifically, according to the failure probability of the single battery, the failure probability of the current abnormal battery group is calculated, which is the mean failure probability of all single batteries in the current abnormal battery group. It should be noted that the above only calculates the failure probability of the single battery in the abnormal area. For the failure probability of the single battery in the non-abnormal area, 0 is used for calculation here, and the average failure probability of all battery groups in the cluster where the current abnormal battery group is located is calculated to obtain the abnormal degree of failure probability of the current abnormal battery group; according to the degree of discreteness and the failure probability abnormality, the influence coefficient of other battery groups in the cluster on the current abnormal battery group is obtained. Construct the current The other battery groups in the cluster where the abnormal battery group is located have an impact on the current The calculation formula for the influence coefficient of an abnormal battery group is:

[0073] In the formula, Indicates the current The other battery groups in the cluster where the abnormal battery group is located have an impact on the current The influence coefficient of abnormal battery group; Indicates the current The mean of the failure probability of all battery packs in the cluster where the abnormal battery pack is located; Indicates the current Failure probability of an abnormal battery pack; Indicates the current The degree of dispersion among all battery groups in the cluster where the abnormal battery group is located; and Respectively represent the second correction parameter and the third correction parameter, both used to prevent the denominator from being 0. In this embodiment, , .

[0074] Indicates the abnormal degree of the fault probability of the current abnormal battery group. The smaller the value, the smaller the difference in fault probability between the current abnormal battery group and other battery groups in the cluster where it is located, indicating that the operating conditions between the battery groups are more similar and the influence between the battery groups is greater; It indicates the degree of dispersion among all battery groups in the cluster where the current abnormal battery group is located. The smaller the value, the smaller the difference in operating condition data among all battery groups in the cluster where the current abnormal battery group is located, indicating that the operating condition similarity among battery groups is greater and the influence among battery groups is greater.

[0075] S700: According to the influence coefficient and the failure probability, the initial sampling frequency is adjusted to obtain a corrected sampling frequency.

[0076] According to the influence coefficient and the failure probability, the initial sampling frequency is adjusted to obtain the corrected sampling frequency and construct the The calculation formula for correcting the sampling frequency of an abnormal battery group is:

[0077] In the formula, Indicates the current Corrected sampling frequency for each abnormal battery group; Indicates the current Failure probability of an abnormal battery pack; Current The other battery groups in the cluster where the abnormal battery group is located have an impact on the current The influence coefficient of abnormal battery group; Indicates the initial sampling frequency.

[0078] Failure probability The larger the value is, the more abnormal the current abnormal battery group is, and the more it is necessary to increase the sampling frequency of the current abnormal battery group; at the same time, the influence coefficient The larger the value is, the greater the interference of other battery groups on the current abnormal battery group is. Therefore, it is more necessary to increase the sampling frequency of the current abnormal battery group.

[0079] The edge computing node adjusts the data acquisition parameters of the sensor according to the real-time working conditions of the battery pack, and then uploads the acquired battery working data according to the situation. For battery packs with a low probability of failure, only the overall data is uploaded, and the data of the single battery is stored under the edge computing node. For battery packs with a high probability of failure, the data of each single battery needs to be uploaded. After the required data is uploaded to the central server, the health status of the battery pack under each edge computing node is evaluated respectively, and the evaluated data is displayed through the Internet of Things terminal (mobile phone, computer, etc.).

[0080] Based on the same inventive concept as the above method, this embodiment also provides a 5G Internet of Things data collection optimization system.

[0081] See also Figure 2 , which shows the basic composition of a 5G Internet of Things data collection and optimization system provided by an embodiment of the present invention.

[0082] like Figure 2 As shown, a 5G Internet of Things data collection optimization system includes: a memory 10 and a processor 20, wherein: A memory 10, used for storing program codes; The processor 20 is used to read the program code stored in the memory 10, and execute the setting of the initial sampling frequency, collect the multi-dimensional working data of the battery group, and the multi-dimensional working data includes the working voltage and the working current; analyze the first state abnormal index of the battery group according to the working voltage and the working current to obtain the abnormal battery group; analyze the voltage change trend and the current change trend of the abnormal battery group to obtain the local abnormal judgment coefficient and determine the possible abnormal area; analyze the second state abnormal index of the possible abnormal area to determine the abnormal area; for the single battery in the abnormal area, analyze the relative deviation of the working data of each dimension, and combine the local abnormal judgment coefficient to obtain the failure probability of the single battery; cluster the battery groups to obtain cluster clusters, analyze the discrete degree between all the battery groups in the cluster cluster, and combine the failure probability to obtain the influence coefficient of other battery groups in the cluster cluster on the current abnormal battery group; adjust the initial sampling frequency according to the influence coefficient and the failure probability to obtain the corrected sampling frequency.

[0083] Furthermore, the processor includes: a working data acquisition module 21, an abnormal area determination module 22, a fault probability analysis module 23, an influence coefficient analysis module 24 and a sampling frequency adjustment module 25. Among them: The working data acquisition module 21 is used to set the initial sampling frequency and collect multi-dimensional working data of the battery pack, where the multi-dimensional working data includes working voltage and working current; The abnormal area determination module 22 is used to analyze the first state abnormality index of the battery pack according to the working voltage and the working current to obtain the abnormal battery pack; The fault probability analysis module 23 is used to analyze the voltage change trend and current change trend of the abnormal battery group, obtain the local abnormality judgment coefficient, and determine the possible abnormal area; and analyze the second state abnormality index of the possible abnormal area to determine the abnormal area; finally, for the single battery in the abnormal area, analyze the relative deviation of the working data of each dimension, and combine the local abnormality judgment coefficient to obtain the fault probability of the single battery; The influence coefficient analysis module 24 is used to cluster the battery groups to obtain clusters, analyze the discreteness between all the battery groups in the clusters, and obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group in combination with the failure probability; The sampling frequency adjustment module 25 is used to adjust the initial sampling frequency according to the influence coefficient and the failure probability to obtain a modified sampling frequency.

[0084] It should be noted that the sequence of the above embodiments of the present invention is for description only and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A 5G Internet of Things data collection optimization method, characterized in that: The method comprises: Setting an initial sampling frequency to collect multi-dimensional operating data of the battery pack, wherein the multi-dimensional operating data includes operating voltage and operating current; Analyzing the first state abnormality index of the battery pack according to the operating voltage and the operating current to obtain an abnormal battery pack; Analyze the voltage change trend and current change trend of the abnormal battery group to obtain a local abnormality judgment coefficient and determine a possible abnormal area; Analyze the second state abnormality index of the possible abnormal area to determine the abnormal area; For the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient; Clustering the battery groups to obtain clusters, analyzing the discreteness between all the battery groups in the clusters, and combining the failure probability to obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group; According to the influence coefficient and the failure probability, the initial sampling frequency is adjusted to obtain a corrected sampling frequency.

2. The 5G Internet of Things data collection optimization method according to claim 1 is characterized in that: Analyzing the first state abnormality index of the battery pack according to the operating voltage and the operating current to obtain an abnormal battery pack includes: According to the working voltage and the working current of the battery pack at the current moment, combined with the theoretical voltage and the theoretical current of the battery pack, a first state abnormality index of the battery pack at the current moment is obtained; Presetting a first abnormality index threshold; Determine whether the first state abnormality index is greater than the first abnormality index threshold; If yes, the battery pack corresponding to the first state abnormality index is an abnormal battery pack.

3. The 5G Internet of Things data collection optimization method according to claim 1 is characterized in that: Analyze the voltage change trend and current change trend of the abnormal battery group to obtain the local abnormality judgment coefficient and determine the possible abnormal area, including: Based on the working voltage and the working current of the battery pack at the current moment and the previous moment, analyzing the voltage fitting slope and the current fitting slope at the current moment, and obtaining the voltage change trend and the current change trend of the abnormal battery pack; According to the voltage variation trend, combined with the working current and theoretical current of the battery pack at the current moment, a series abnormality coefficient is obtained; According to the voltage variation trend and the current variation trend, a parallel abnormality coefficient is obtained; Obtaining a local abnormality determination coefficient according to the series abnormality coefficient and the parallel abnormality coefficient; According to the local abnormality judgment coefficient, a possible abnormal area is determined.

4. The 5G Internet of Things data collection optimization method according to claim 1 is characterized in that: Analyzing the second state abnormality index of the possible abnormal area to determine the abnormal area includes: According to the working voltage and the working current of the possible abnormal area at the current moment, combined with the theoretical voltage and the theoretical current of the possible abnormal area, a second state abnormality index of the possible abnormal area at the current moment is obtained; Presetting a second abnormality index threshold; Determine whether the second state abnormality index is greater than the second abnormality index threshold; If yes, the possible abnormal area corresponding to the second state abnormality index is the abnormal area.

5. The 5G Internet of Things data collection optimization method according to claim 1 is characterized in that: For the single storage battery in the abnormal area, the relative deviation of the working data in each dimension is analyzed, which also includes: According to the second state abnormality index corresponding to the abnormal area, combined with the initial sampling frequency, a temporary sampling frequency of the abnormal area is obtained; According to the temporary sampling frequency, the latest multi-dimensional working data of the abnormal area is re-acquired.

6. The 5G Internet of Things data collection optimization method according to claim 5 is characterized in that: For the single storage battery in the abnormal area, the relative deviation of the working data of each dimension is analyzed, and the failure probability of the single storage battery is obtained by combining the local abnormality judgment coefficient, including: For the single storage battery in the abnormal area, calculate the mean of the working data in each dimension, analyze the difference in the working data in each dimension between the current single storage battery and other single storage batteries in the abnormal area, and obtain the relative deviation of the working data in each dimension; According to the relative deviation of the working data of each dimension, the total deviation of the single battery is obtained; The failure probability of the single storage battery is obtained according to the total deviation and in combination with the local abnormality judgment coefficient of the local area where the single storage battery is located.

7. The 5G Internet of Things data collection optimization method according to claim 1 is characterized in that: The battery groups are clustered to obtain clusters, the discreteness between all the battery groups in the clusters is analyzed, and the influence coefficient of other battery groups in the clusters on the current abnormal battery group is obtained in combination with the failure probability, including: According to the scale of the battery group, cluster analysis is performed on the battery group to obtain cluster clusters; Calculate the mean value of the working data of each battery group in each dimension in the cluster where the current abnormal battery group is located; Calculate the variance of the mean value of each dimension of working data among all battery groups in the cluster where the current abnormal battery group is located, obtain the degree of discreteness of the dimensional data of each dimension of working data among all battery groups in the cluster, and then obtain the degree of discreteness among all battery groups in the cluster; According to the discrete degree and in combination with the fault probability, the influence coefficient of other storage battery groups in the cluster on the current abnormal storage battery group is obtained.

8. The 5G Internet of Things data collection optimization method according to claim 7 is characterized in that: According to the discrete degree and in combination with the fault probability, the influence coefficient of other battery groups in the cluster on the current abnormal battery group is obtained, including: According to the failure probability of the single battery, the failure probability of the current abnormal battery group is calculated, and the average failure probability of all battery groups in the cluster where the current abnormal battery group is located is calculated to obtain the abnormal degree of the failure probability of the current abnormal battery group; According to the discrete degree and in combination with the abnormal degree of the fault probability, the influence coefficient of other storage battery groups in the cluster on the current abnormal storage battery group is obtained.

9. A 5G Internet of Things data collection optimization system, characterized in that: The system comprises: a memory and a processor, wherein: The memory is used to store program codes; The processor is configured to read the program code stored in the memory and execute the method according to any one of claims 1 to 8.

10. The 5G Internet of Things data acquisition optimization system according to claim 9, characterized in that: The processor comprises: A working data acquisition module, used to set an initial sampling frequency and collect multi-dimensional working data of the battery pack, wherein the multi-dimensional working data includes working voltage and working current; an abnormal region determination module, configured to analyze a first state abnormality index of the battery pack according to the operating voltage and the operating current, and obtain an abnormal battery pack; A fault probability analysis module is used to analyze the voltage change trend and current change trend of the abnormal battery group, obtain the local abnormality judgment coefficient, and determine the possible abnormal area; and analyze the second state abnormality index of the possible abnormal area to determine the abnormal area; finally, for the single battery in the abnormal area, analyze the relative deviation of the working data of each dimension, and combine the local abnormality judgment coefficient to obtain the fault probability of the single battery; An influence coefficient analysis module is used to cluster the battery groups to obtain clusters, analyze the discreteness between all the battery groups in the clusters, and obtain the influence coefficient of other battery groups in the clusters on the current abnormal battery group in combination with the failure probability; The sampling frequency adjustment module is used to adjust the initial sampling frequency according to the influence coefficient and the fault probability to obtain a modified sampling frequency.

Images