A temperature abnormality warning method and system for lithium battery stereoscopic warehouse

Abstract

The present invention relates to the field of data processing technology, and more specifically, to a temperature anomaly warning method and system for a lithium battery stereoscopic warehouse, the method comprising: segmenting a temperature sequence at different positions in the stereoscopic warehouse to obtain a plurality of data segments of the same length, randomly selecting a fixed number of data points from each data segment, obtaining initial particles of a particle filter algorithm, calculating the weight of each initial particle, using the weight as the weight of the initial particle initialized by the particle filter algorithm, and using the particle filter algorithm after completing the initialization step to perform a state evaluation on the temperature at each sampling moment, obtaining fused data at each sampling moment, and warning the temperature anomaly of the stereoscopic warehouse based on the anomaly detection results of each fused data. The present invention can improve the accuracy of temperature anomaly detection.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and more specifically, to a temperature abnormality early warning method and system for a lithium battery stereoscopic warehouse. Background Art

[0002] A lithium battery warehouse is a facility specifically used to store lithium batteries. It needs to have efficient space utilization and advanced temperature control systems to ensure that the batteries are stored under optimal environmental conditions. Abnormal temperature in the warehouse may affect the performance and safety of lithium batteries, and even cause fire or explosion. The temperature in the warehouse is usually unevenly distributed and exhibits complex nonlinear characteristics. Therefore, there is an urgent need for a method that can effectively process complex data. The particle filter algorithm is a recursive estimation method based on Bayesian theory, which can effectively integrate the unevenly distributed and complex nonlinear characteristics of temperature caused by various mechanisms such as heat conduction, convection and radiation.

[0003] In the related technology, a filtering-based lithium battery core and surface temperature estimation method is disclosed in the patent application document with publication number CN113076681A. This method uses polyhedrons to regionally limit the diffusion range of particles, making the distribution of particles closer to the true value and denser, avoiding replacement during the resampling process and retaining the diversity of particles.

[0004] However, in the process of state estimation using the particle filter algorithm, the above scheme does not take into account the impact of differences in the importance of information represented by each data. It only improves the resampling step of the algorithm and retains the diversity of particles, but may lead to the loss of key information, thereby affecting the judgment of the overall state, making the accuracy of the state estimation value determined at each moment low, resulting in the inability to accurately judge the temperature abnormality based on the obtained state estimation value. Summary of the invention

[0005] In order to solve the problem that the temperature abnormality cannot be accurately judged due to the low accuracy of the state estimation value determined by the particle filtering algorithm, the present invention provides a temperature abnormality warning method and system for a lithium battery stereoscopic warehouse.

[0006] According to a first aspect of the present invention, a temperature abnormality warning method for a lithium battery stereoscopic warehouse is provided, comprising:

[0007] The temperature series at different positions in the stereo library are segmented to obtain multiple data segments of the same length, and a fixed number of data points are randomly selected from each data segment to obtain the initial particles of the particle filter algorithm;

[0008] Calculate the weight of each initial particle, use the weight as the weight of the initial particle initialized by the particle filter algorithm, and use the particle filter algorithm after the initialization step to evaluate the temperature at each sampling time, and obtain the fusion data at each sampling time, so as to warn the temperature anomaly of the stereoscopic warehouse based on the abnormal detection results of each fusion data;

[0009] The process of obtaining the weight of each initial particle is as follows: taking the normalized value of the stability of the data segment where any initial particle is located as the initial weight of the initial particle, taking the temperature other than the initial particle collected at the sampling time corresponding to the initial particle as the associated data of the initial particle, determining the correlation between the initial particle and the associated data, normalizing the sum of the correlation and the initial weight, and obtaining the weight of the initial particle;

[0010] Relevance for: , For the The data segment where the initial particle is located is the same as the The variance of the difference between the corresponding data points in the data segment where the associated data is located, For the The average value of the variance of the difference between the data segment where the initial particles are located and the corresponding data points in the data segment where the associated data are located, is the data volume of the temperature series, is the absolute value symbol.

[0011] The present invention can set weights for each initial particle differently, and the weights of initial particles with less important characteristic information are lower, while the weights of initial particles with more important characteristic information are higher, thereby ensuring that important temperature characteristic information affecting the safety of the three-dimensional warehouse is not diluted, thereby improving the accuracy of state estimation of the temperature at each sampling moment, obtaining fused data with higher accuracy, and then based on the abnormality detection results of the fused data with higher accuracy, realizing accurate early warning of temperature abnormalities in the three-dimensional warehouse.

[0012] Preferably, the weight of each initial particle satisfies the following relationship:

[0013] ;

[0014] In the formula, For the The weight of the initial particles; For the The initial weight of the initial particles; For the The correlation between the initial particles and the associated data; is the normalization function.

[0015] The calculation formula for the weight of each initial particle provided by the present invention has a low amount of calculation and can reduce the difficulty of calculation while ensuring the accuracy of the determined weight.

[0016] Preferably, the initial weight of the initial particle satisfies the following relationship:

[0017] ;

[0018] In the formula, For the The initial weight of the initial particles; For the The stability of the data segment where the initial particles are located; A natural constant An exponential function with base .

[0019] The present invention limits the range of the initial weight of the initial particle to 0-1, which can reduce the difficulty of calculating the initial weight.

[0020] Preferably, the method for obtaining the stability of a data segment includes:

[0021] Get the average value of the data in a data segment composed of the first data point in any data segment and any subsequent data point, and obtain the mean sequence of any data segment;

[0022] Calculate the average of the rate of change of each data in the mean sequence to obtain the stability of any data segment.

[0023] The method of determining the stability of a data segment of the present invention can effectively smooth the noise data in the data segment, thereby ensuring the accuracy of the stability.

[0024] Preferably, Satisfies the following relationship:

[0025] ;

[0026] In the formula, For the The data segment where the initial particle is located is the same as the The variance of the difference between the corresponding data points in the data segment where the associated data are located; For the The sampling time of the initial particle in the data segment is The data points of The sampling time of the associated data in the data segment is The difference between the data points For the The data segment where the initial particle is located is the same as the The average value of the differences between the corresponding data points in the data segment where the associated data are located; For the The amount of data in the data segment where the initial particle is located.

[0027] Preferably, the abnormal temperature detection result of each fused data is used to warn the abnormal temperature of the stereoscopic warehouse, including:

[0028] Obtain a preset anomaly score threshold, use the local outlier factor algorithm to perform anomaly detection on the fused data at any sampling time, and obtain the anomaly score of the fused data at any sampling time;

[0029] When the abnormality score is greater than the abnormality score threshold, an alarm is issued to remind the operator that the temperature of the stereoscopic warehouse is abnormal at any sampling time.

[0030] The present invention can accurately give early warning of abnormal temperature conditions in the stereoscopic warehouse, thereby ensuring the safety of the stereoscopic warehouse.

[0031] Preferably, the method for acquiring fusion data includes:

[0032] The particle filter algorithm after the initialization step is obtained to perform state evaluation on the temperature at each sampling moment, and the state estimation of the temperature at each sampling moment is obtained, and the state estimation of the temperature at each sampling moment is used as the fusion data of the corresponding sampling moment.

[0033] According to a second aspect of the present invention, a temperature anomaly warning system for a lithium battery warehouse is provided. The temperature anomaly warning system for a lithium battery warehouse comprises a memory and a processor. A computer program is stored in the memory. The processor executes the computer program to implement the steps of the first aspect of the present invention.

[0034] The present invention has the following effects:

[0035] The present invention determines the weight of each initial particle by comprehensively considering the correlation between the initial particles and the associated data, as well as the stability of the data segment where the initial particles are located, so that the set weight can more accurately reflect the importance of the characteristic information represented by different temperature sequences, and effectively avoids the key information dilution problem caused by setting the same weight in the initialization stage in the particle filter algorithm, so that the temperature state at each sampling moment can be accurately estimated, and the accuracy of the fused data is improved; and through the abnormal detection result of the fused data, the temperature anomaly of the stereoscopic library is warned, so that potential temperature problems can be discovered and handled in time, and the safety risks caused by the temperature anomaly are reduced, thereby improving the safety of the stereoscopic library. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0037] Figure 1 This is a schematic diagram of the steps of a temperature abnormality warning method for a lithium battery stereoscopic warehouse according to an embodiment of the present invention;

[0038] Figure 2 1 is a schematic flow chart of the steps of calculating the weight of each initial particle in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0040] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0041] Reference Figure 1 , a temperature abnormality warning method for a lithium battery stereoscopic warehouse, comprising steps S1 to S3, specifically as follows:

[0042] S1: The temperature sequence at different positions in the stereo library is segmented to obtain multiple data segments of the same length, and a fixed number of data points are randomly selected from each data segment to obtain the initial particles of the particle filter algorithm.

[0043] Specifically, temperature sensors can be installed at multiple different locations in the lithium battery stereoscopic warehouse, and the data collection frequency can be set, such as 1Hz, to collect temperature, and then the temperatures collected by all temperature sensors within the same time period, such as within 12 hours, are used as the temperature sequence of the lithium battery stereoscopic warehouse. It should be noted that the installation positions of the temperature sensors need to fully cover the lithium battery stereoscopic warehouse, and this embodiment does not specifically limit the number of installed temperature sensors and the specific installation positions.

[0044] Optionally, each temperature sequence may be divided into multiple data segments of the same length according to a certain length, such as the length of the temperature data segment collected within 1 hour, and a fixed number, such as 1 data point, may be randomly selected from the multiple data segments corresponding to each temperature sequence to obtain initial particles of the particle filter algorithm. This embodiment does not specifically limit the length of the divided data segments and the number of selected initial particles.

[0045] It should be noted that the initial particles determined by the present invention are particles selected when the particle filter algorithm is initialized.

[0046] S2: Calculate the weight of each initial particle and use the weight as the weight of the initial particle initialized by the particle filter algorithm.

[0047] It should be noted that when the particle filter algorithm initializes the weight of the initial particles, the same weight is generally set for each initial particle. However, due to the different locations of the temperature sensors, the importance of the collected temperatures also varies to a certain extent. If the same weight is set for each initial particle, it may lead to the loss of key information and affect the overall judgment. Therefore, the present invention uses the calculated weight of each initial particle as the weight of each initial particle initialized by the particle filter algorithm, thereby improving the way in which the particle filter algorithm initializes the weight of each initial particle.

[0048] It should be further explained that the present invention only improves the process of initializing the weights of the initial particles in the particle filter algorithm, and does not improve other parameters of the algorithm and the steps after initialization.

[0049] Specifically, refer to Figure 2 As shown in FIG. 1 , the process of obtaining the weight of each initial particle includes the following steps:

[0050] S201: taking the normalized value of the stability of the data segment where any initial particle is located as the initial weight of the initial particle;

[0051] Among them, stability refers to data that can reflect the overall change level of a data segment. For example, when the overall change level of any data segment is relatively stable, the stability of the data segment is relatively strong. On the contrary, when the overall change of any data segment fluctuates greatly, the stability of the data segment is relatively poor.

[0052] In an exemplary embodiment of the present invention, the stability of a data segment may be determined by the following steps:

[0053] Step 1: Get the average value of the data in a data segment composed of the first data point in any data segment and any subsequent data point, and obtain the mean sequence of any data segment;

[0054] It should be noted that due to the existence of problems such as sensor accuracy and service life, there may be certain noise data in the collected temperature series, that is, there is noise data in the segmented data segment. The stability of the data segment is determined by the mean method, which can effectively smooth the impact of noise data on the overall change level of the data segment, thereby ensuring that the initial weights of each initial particle can reflect the actual temperature changes in the data segment.

[0055] Optionally, the mean sequence of any data segment can be recorded as , among which, when When is the average of the first and second data points in the data segment. When It is the average value of the data in the data segment composed of the first data point and the third data point in the data segment, that is, the average value of the first data point, the second data point and the third data point. And so on, the average value of the data in the data segment composed of the first data point in the data segment and any subsequent data point is obtained, and then the mean sequence of the data segment is obtained.

[0056] Step 2: Calculate the average value of the rate of change of each data in the mean sequence to obtain the stability of any data segment.

[0057] First, we can calculate the mean sequence corresponding to any data segment The change rate of each data in the equation satisfies the following relationship:

[0058] ;

[0059] In the formula, is the mean value in the sequence The rate of change of individual data; The first data point in any data segment and the The average value of the data in the data segment consisting of data points; The first data point in the data segment and the The average value of the data in a data segment consisting of data points.

[0060] Then, the average value of the rate of change of each data in the mean sequence can be calculated to obtain the stability of the corresponding data segment.

[0061] Furthermore, after determining the stability of the data segment where any initial particle is located, the stability can be normalized to obtain the initial weight of the initial particle. Specifically, the initial weight of any initial particle satisfies the following relationship:

[0062] ;

[0063] In the formula, For the The initial weight of the initial particles; For the The stability of the data segment where the initial particles are located; A natural constant An exponential function with base .

[0064] In another embodiment, other functions may be used to normalize the stability of the data segment, such as by The function is normalized. This embodiment does not specifically limit the normalization method.

[0065] S202: taking the temperature of the initial particle except the initial particle collected at the sampling time corresponding to the initial particle as the associated data of the initial particle, and determining the correlation between the initial particle and the associated data;

[0066] Specifically, the correlation between any initial particle and the associated data satisfies the following relationship:

[0067] ;

[0068] In the formula, For the The correlation between the initial particles and the associated data; For the The data segment where the initial particle is located is the same as the The variance of the difference between the corresponding data points in the data segment where the associated data are located; For the The average value of the variance of the difference between the corresponding data points in the data segment where the initial particles are located and the data segment where the associated data are located; is the data volume of the temperature series; is the absolute value symbol.

[0069] in, , where , For the The sampling time of the initial particle in the data segment is The data points of The sampling time of the associated data in the data segment is The difference between the data points For the The sampling time of the initial particle in the data segment is data points; For the The sampling time of the associated data in the data segment is data points; is the absolute value symbol; For the The data segment where the initial particle is located is the same as the The average value of the differences between the corresponding data points in the data segment where the associated data are located; For the The amount of data in the data segment where the initial particle is located.

[0070] S203: normalizing the sum of the correlation and the initial weight to obtain the weight of the initial particle.

[0071] Specifically, the weight of any initial particle satisfies the following relationship:

[0072] ;

[0073] In the formula, For the The weight of the initial particles; For the The initial weight of the initial particles; For the The correlation between the initial particles and the associated data; is the normalization function.

[0074] Among them, using The correlation between an initial particle and the associated data can be used to correct the initial weight of the initial particle. When the correlation between the initial particle and the associated data is large, the initial weight of the initial particle can be strengthened, so that when the particle filter algorithm is used to evaluate the state of each temperature series, the initial particle with a large weight can be paid attention to, so that the obtained fusion data can reflect the key information represented by the initial particle.

[0075] S3: After completing the initialization step, the particle filter algorithm is used to evaluate the temperature at each sampling moment, and the fusion data at each sampling moment is obtained, so as to warn the temperature anomaly of the three-dimensional warehouse based on the abnormal detection results of each fusion data.

[0076] Specifically, after the weight of each initial particle is determined through step S2, the weight of each initial particle can be used as the weight of each initial particle in the particle filter algorithm to initialize the weight of each initial particle, and then the unimproved steps of the algorithm are continued to be executed to perform state evaluation on the temperature at each sampling moment, thereby determining the fusion data at each sampling moment.

[0077] In an exemplary embodiment of the present invention, the determination of fusion data at each sampling moment can be achieved by the following steps:

[0078] The particle filter algorithm after the initialization step is obtained to perform state evaluation on the temperature at each sampling moment, and the state estimation of the temperature at each sampling moment is obtained, and the state estimation of the temperature at each sampling moment is used as the fusion data of the corresponding sampling moment.

[0079] Optionally, the fused data at any sampling moment can fuse the temperatures collected by all temperature sensors at the sampling moment, and the temperature representing the greater importance of information has a greater weight in the fused data, and the temperature representing the less important information has a smaller weight in the fused data, so that the fused data at each sampling moment can be analyzed to accurately detect temperature anomalies in the stereoscopic library. It should be noted that the process of determining the state estimation of the system using the particle filter algorithm is a prior art, and state estimation is a professional term in the algorithm, which will not be described in detail in this embodiment.

[0080] In an exemplary embodiment of the present invention, an early warning of abnormal temperature in a stereoscopic warehouse can be achieved by the following steps:

[0081] Obtain a preset anomaly score threshold, use the local outlier factor algorithm to perform anomaly detection on the fused data at any sampling time, and obtain the anomaly score of the fused data at any sampling time; when the anomaly score is greater than the anomaly score threshold, an alarm is issued to remind the operator that the temperature of the stereoscopic warehouse is abnormal at any sampling time.

[0082] It should be noted that the process of performing anomaly detection on data using a local outlier factor (LOF) algorithm to obtain anomaly scores of the data is a prior art and will not be described in detail in this embodiment.

[0083] Optionally, the anomaly score threshold can be set to 1. When the anomaly score of the fused data at any sampling moment determined by the algorithm is greater than 1, the early warning mechanism is triggered, so that the operator can be promptly reminded that the temperature of the stereoscopic warehouse is abnormal at the sampling moment, so that corresponding processing can be carried out in time to ensure the safety of the stereoscopic warehouse for storing lithium batteries.

[0084] The present invention also provides a temperature anomaly warning system for a lithium battery warehouse. The system includes a memory and a processor, and a computer program is stored in the memory. The computer program integrates the function of a temperature anomaly warning method for a lithium battery warehouse. When the computer program is executed, the accuracy of temperature anomaly detection can be improved through a temperature anomaly warning method for a lithium battery warehouse.

[0085] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0086] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A temperature abnormality warning method for a lithium battery stereoscopic warehouse, characterized in that: include: The temperature series at different positions in the stereo library are segmented to obtain multiple data segments of the same length, and a fixed number of data points are randomly selected from each data segment to obtain the initial particles of the particle filter algorithm; The weight of each initial particle is calculated, and the weight is used as the weight of the initial particle to initialize the particle filter algorithm, and the particle filter algorithm after the initialization step is used to perform a state evaluation on the temperature at each sampling moment to obtain the fusion data at each sampling moment, so as to give an early warning of the temperature anomaly of the stereoscopic warehouse based on the abnormality detection result of each fusion data, including: obtaining a preset abnormality score threshold, using a local outlier factor algorithm to perform anomaly detection on the fusion data at any sampling moment, and obtaining the abnormality score of the fusion data at any sampling moment; when the abnormality score is greater than the abnormality score threshold, an alarm is issued to remind the operator that the temperature of the stereoscopic warehouse is abnormal at any sampling moment; The method for acquiring fused data includes: acquiring the result of state evaluation of the temperature at each sampling moment by the particle filter algorithm after the initialization step is completed, obtaining the state estimation of the temperature at each sampling moment, and using the state estimation of the temperature at each sampling moment as the fused data at the corresponding sampling moment; The process of obtaining the weight of each initial particle is as follows: taking the normalized value of the stability of the data segment where any initial particle is located as the initial weight of the initial particle, taking the temperature of the initial particle except the initial particle collected at the sampling time corresponding to the initial particle as the associated data of the initial particle, determining the correlation between the initial particle and the associated data, normalizing the sum of the correlation and the initial weight to obtain the weight of the initial particle; the correlation for: , For the The data segment where the initial particle is located is the same as the The variance of the difference between the corresponding data points in the data segment where the associated data is located, For the The average value of the variance of the difference between the data segment where the initial particles are located and the corresponding data points in the data segment where the associated data are located, is the data volume of the temperature series, is the absolute value symbol.

2. The temperature abnormality warning method for a lithium battery stereoscopic warehouse according to claim 1 is characterized in that: The weights of the initial particles satisfy the following relationship: ； In the formula, For the The weight of the initial particles; For the The initial weight of the initial particles; For the The correlation between the initial particles and the associated data; is the normalization function.

3. The temperature abnormality warning method for a lithium battery stereoscopic warehouse according to claim 2 is characterized in that: The initial weight of the initial particle satisfies the following relationship: ； In the formula, For the The initial weight of the initial particles; For the The stability of the data segment where the initial particles are located; A natural constant An exponential function with base .

4. The temperature abnormality warning method for a lithium battery stereoscopic warehouse according to claim 3 is characterized in that: The method for obtaining the stability of the data segment includes: Obtain the average value of the data in a data segment consisting of the first data point in any data segment and any subsequent data point, and obtain the mean sequence of any data segment; The average value of the rate of change of each data in the mean sequence is calculated to obtain the stability of any data segment.

5. The temperature abnormality warning method for a lithium battery stereoscopic warehouse according to claim 1 is characterized in that: Said , satisfying the following relationship: ； In the formula, For the The data segment where the initial particle is located is the same as the The variance of the difference between the corresponding data points in the data segment where the associated data are located; For the The sampling time of the initial particle in the data segment is The data points of The sampling time of the associated data in the data segment is The difference between the data points For the The data segment where the initial particle is located is the same as the The average value of the differences between the corresponding data points in the data segment where the associated data are located; For the The amount of data in the data segment where the initial particle is located.

6. A temperature abnormality warning system for a lithium battery stereoscopic warehouse, characterized in that: The temperature anomaly warning system for a lithium battery stereoscopic warehouse includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as described in any one of claims 1 to 5.

Images