Error filtering processing method and device for power battery data

Abstract

The invention provides an error filtering processing method and device for power battery data, and solves the technical problem of state recognition errors caused by the fact that interference factors cannot be eliminated in the state signal acquisition process of an existing battery pack. The method comprises the following steps: acquiring time sequence state data of each battery monomer and forming time sequence state mean value data taking a battery pack as a unit; forming time sequence standard deviation data corresponding to the time sequence state mean value of each battery pack; forming a Pearson's correlation coefficient according to the time sequence state extreme value of the single batteries in the battery pack; and forming a judgment strategy according to the time sequence standard deviation data and the Pearson's correlation coefficient to judge an error state and correct a detection error. The technical problem caused by data deviation is solved from the perspective of software. And battery sampling data abnormity caused by interference factors is eliminated through real-time filtering processing. Abnormal data can be eliminated, and the probability of system misinformation is reduced. The principle of the invention can be used in any similar data processing scene.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and specifically to a method and apparatus for error filtering processing of power battery data. Background Technology

[0002] In existing technologies, battery packs / packs, formed by integrating individual battery cells, serve as energy storage units for electric vehicles, providing power for the vehicle's operation. The Battery Management System (BMS) manages these energy storage units, possessing functions such as battery data detection and monitoring, battery pack status assessment and protection, and command transmission and forwarding, coordinating and managing the onboard battery pack's operation. The basic architecture of a battery management system is as follows: Figure 1 As shown. In Figure 1 In this system, several battery cells are connected in series and parallel to form a battery pack. A slave control module uses voltage, current, and temperature acquisition modules to collect signals from each battery cell in the battery pack. Each slave control module reports the detection data in parallel to the secondary master control module. The secondary master control module stores and processes the relevant information to form real-time status data such as total voltage, total current, and SOC. After encoding and encapsulation, it is uploaded to the energy storage unit master controller and the vehicle controller. The secondary master control module accurately reports the battery status and fault information while receiving instructions from the vehicle controller to complete tasks such as battery pack switching.

[0003] The status signal acquisition of individual battery cells within the battery pack is accomplished by a slave control module consisting of a data acquisition chip and its peripheral circuitry. For example, a battery pack may include 12 series-connected batteries, and voltage signal measurement typically corresponds to 12 voltage sensors and a 12-channel 12-bit voltage acquisition ADC chip. However, due to the susceptibility of the data acquisition chip to electromagnetic interference and the instability of its internal reference voltage, during use, the voltage data of each individual cell acquired by a certain chip may simultaneously rise or fall, leading to false alarms of overcharging or over-discharging by the system and affecting the normal use of the battery pack. Summary of the Invention

[0004] In view of the above problems, embodiments of the present invention provide an error filtering processing method and apparatus for power battery data, which solves the technical problem that the existing battery pack status signal acquisition process cannot eliminate interference factors that lead to status identification errors.

[0005] The error filtering processing method for power battery data according to embodiments of the present invention includes:

[0006] Acquire the time-series state data of each battery cell and generate the average time-series state data of the battery pack as a unit;

[0007] The time-series standard deviation data of the time-series state mean for each battery pack is generated;

[0008] Pearson correlation coefficients are generated based on the time-series state extrema of individual cells within the battery pack;

[0009] Based on time series standard deviation data and Pearson correlation coefficient, a judgment strategy is formed to determine the error status and correct the detection error.

[0010] In one embodiment of the present invention, the timing state includes voltage, current or temperature.

[0011] In one embodiment of the present invention, the formation of the time-series state mean data includes:

[0012] Collect the timing status signal of each battery cell for the determined state type;

[0013] Using the battery pack as the detection unit, the time-series status data of the individual battery cells in each battery pack are generated;

[0014] The average time-series status data of each battery pack is generated based on the time-series status data.

[0015] In one embodiment of the present invention, the formation of the time-series standard deviation data includes:

[0016] The standard deviation S of the mean state of each battery pack at each time step is generated based on the mean state of each battery pack in the time series. T ,

[0017]

[0018] Where M represents the number of data acquisition chips or battery packs. Let x be the mean of the state values ​​of the battery pack measured by each acquisition chip at time T. T,p This represents the average state of the battery pack measured by the p-th acquisition chip.

[0019] In one embodiment of the present invention, the formation of the Pearson correlation coefficient includes:

[0020] At each time T, record the maximum state value of each individual battery cell within the battery pack. T and the minimum value of the state Min T The Pearson correlation coefficient r is formed based on the maximum value (Max) and minimum value (Min) of the state.

[0021]

[0022] Where n is the number of individual battery cells.

[0023] In one embodiment of the present invention, the step of determining the error state based on the time-series standard deviation data and the Pearson correlation coefficient to form a judgment strategy includes:

[0024] When data feedback indicates a battery cell status fault, retrieve the standard deviation data and Pearson correlation coefficient at the time of the fault.

[0025] A fault is identified when the standard deviation is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is greater than the correlation threshold.

[0026] In one embodiment of the present invention, it further includes:

[0027] When the standard deviation is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is less than the correlation threshold, it is determined that the collected data has drifted.

[0028] In one embodiment of the present invention, it further includes:

[0029] When the standard deviation value is greater than the standard deviation threshold, it is determined that there is electromagnetic interference affecting the data collected by each cell in the battery pack where the fault is located.

[0030] The error filtering processing device for power battery data according to an embodiment of the present invention includes:

[0031] The memory is used to store the program code of the error filtering processing method for power battery data as described above during the processing.

[0032] A processor for executing the program code.

[0033] The error filtering processing device for power battery data according to an embodiment of the present invention includes:

[0034] The data acquisition module is used to acquire the time-series status data of each battery cell and generate the time-series status average data of the battery pack as a unit.

[0035] The discrete feature extraction module is used to generate the time-series standard deviation data of the time-series state mean of each battery pack;

[0036] The deviation feature extraction module is used to generate Pearson correlation coefficients based on the time-series state extremes of individual battery cells in the battery pack.

[0037] The error filtering module is used to determine the error status and correct the detection error based on the time series standard deviation data and Pearson correlation coefficient to form a judgment strategy.

[0038] The error filtering method and apparatus for power battery data in this invention solves the technical problems caused by data deviation from a software perspective. Real-time filtering eliminates abnormal battery sampling data caused by interference factors. This removes abnormal data and reduces the probability of false alarms from the system. The principles of this invention can be applied in any similar data processing scenario. Attached Figure Description

[0039] Figure 1The diagram shows a schematic of the management architecture of a BMS in the prior art.

[0040] Figure 2 The diagram shown is a flowchart illustrating an error filtering method for power battery data according to an embodiment of the present invention.

[0041] Figure 3 The diagram shown is a schematic representation of the error filtering process for power battery data according to an embodiment of the present invention, specifically the filtering process for the voltage state of a single battery cell.

[0042] Figure 4 The diagram shown is a schematic diagram of the architecture of an error filtering processing device for power battery data according to an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer and more understandable, the invention will be further described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0044] An embodiment of the present invention provides an error filtering processing method for power battery data, as follows: Figure 2 As shown. In Figure 2 In this embodiment, the following are included:

[0045] Step 100: Obtain the time-series state data of each battery cell and form the average time-series state data of the battery pack as a unit.

[0046] Those skilled in the art will understand that specific types of physical signals can be acquired using dedicated sensors, and analog-to-digital conversion of signals with corresponding accuracy can be achieved using general-purpose analog-to-digital conversion circuits. This conversion includes necessary digital signal encoding, timing encapsulation, and other digital signal processing procedures. The timing states of individual battery cells include, but are not limited to, voltage, current, and temperature. Based on the timing states of individual battery cells, the average timing state of the entire battery pack can be determined.

[0047] Step 200: Generate the time-series standard deviation data of the time-series state mean for each battery pack.

[0048] The standard deviation of the statistically significant state mean at the corresponding time point in the time series is determined using the time-series state mean of each battery pack. The standard deviation of the mean, also known as the standard error (SE), refers to the degree of dispersion between the sample means and the population mean in the sampling distribution of the sample means. It measures the accuracy of the estimation of the population mean by the sample means; the smaller the standard error, the closer the sample means are to the population mean, and the higher the reliability of the estimation.

[0049] Step 300: Develop the Pearson correlation coefficient based on the time-series extreme values ​​of the individual battery cells in the battery pack.

[0050] The Pearson correlation coefficient is a statistic that measures the degree of linear correlation between two continuous variables, ranging from -1 to 1. It allows for quick identification of the co-current trends between variables. Combining scatter plot visualization, significance testing, and professional background knowledge avoids biased conclusions caused by relying solely on the correlation coefficient. It is a preferred method for non-normally distributed data or data with extreme values. Time-series extreme values ​​include, but are not limited to, the maximum and minimum values ​​of individual battery cells at a given time within a specific state of the battery pack.

[0051] Step 400: Based on the time series standard deviation data and Pearson correlation coefficient, a judgment strategy is formed to determine the error status and correct the detection error.

[0052] Based on the time-series status obtained from the detection of individual battery cells and battery packs, statistical calculations at different levels are performed on the status statistics to form parameters and thresholds for error type identification. This identification provides fundamental methods for eliminating or reducing errors, offering more accurate detection data to the main control module and controller.

[0053] The error filtering method for power battery data in this invention extracts features from the data discreteness and linear deviation of related variables in the time-series state data of individual battery cells through a statistical process. It then uses a feature quantization strategy to calibrate and identify specific interference error factors during signal detection and filters them, effectively eliminating abnormal data and reducing the probability of false alarms. The technical solution based on this invention can be used in any similar data processing scenario.

[0054] like Figure 2 As shown, in one embodiment of the present invention, step 100 includes:

[0055] Step 110: Collect the timing status signal of each battery cell for the determined state type.

[0056] The status type includes voltage, current, or temperature, and the corresponding physical signal is collected using a sensor of the corresponding status type.

[0057] Step 120: Using the battery pack as the detection unit, generate the timing status data of the individual battery cells in each battery pack.

[0058] Sensors are configured for each individual battery cell in the battery pack. Analog-to-digital conversion circuits and processing circuits are configured according to the number of sensors to generate a data acquisition chip for the battery pack's timing status. This data acquisition chip is a component of the slave control module. The data acquisition chip generates timing status data u for each battery cell in the battery pack. T,n(That is, forming a two-dimensional array with time T as the time and n as the battery cell number), which reflects the state value of the battery cells in the battery pack at each time. In one embodiment of the present invention, the battery pack includes 12 battery cells.

[0059] Step 130: Generate the average timing status data for each battery pack based on the timing status data.

[0060] The time-series state data of each battery cell in the battery pack is used to form the average state data of the battery pack at each moment, and then the time-series state average data x of the battery pack is formed. T,i (That is, forming a two-dimensional array with time T as the time and i as the battery pack or battery pack acquisition chip number), which reflects the average state of each battery pack at each time. In one embodiment of the present invention, the BMS sets the number of acquisition chips M according to the number of battery packs. In one embodiment of the present invention, the time-series state average data is generated through a secondary main control module.

[0061] The error filtering method for power battery data in this embodiment of the invention utilizes the acquisition chip to generate ordered data on the specific states of individual battery cells in the battery pack, and forms a time-series quantization of the average state of the battery pack. This ensures the data foundation for the BMS to perform further interference state identification.

[0062] like Figure 2 As shown, in one embodiment of the present invention, step 200 includes:

[0063] Step 210: Calculate the standard deviation S of the mean state of each battery pack at each time step based on the mean state of each battery pack. T (That is, a one-dimensional array formed by the standard deviation of the mean values ​​of the states of each battery pack at time T):

[0064]

[0065] Where M represents the number of data acquisition chips or battery packs. Let x be the mean of the state values ​​of the battery pack measured by each acquisition chip at time T. T,p This represents the average state of the battery pack measured by the p-th acquisition chip.

[0066] The error filtering method for power battery data in this embodiment of the invention utilizes the state mean of the battery pack to form a benchmark parameter for the statistical data dispersion of all battery packs under BMS control. This satisfies the data mapping basis for BMS to perform further disturbance state identification.

[0067] like Figure 2 As shown, in one embodiment of the present invention, step 300 includes:

[0068] Step 310: At each time T, record the maximum state value (Max) of each battery cell within the battery pack.T and the minimum value of the state Min T The Pearson correlation coefficient r is formed based on the maximum value (Max) and minimum value (Min) of the state:

[0069]

[0070] The error filtering method for power battery data in this invention utilizes the relevant state extreme values ​​of individual battery cells to form a benchmark parameter for the linear deviation of statistical data within each battery pack under BMS control. This satisfies the data mapping basis for further disturbance state identification by the BMS.

[0071] like Figure 2 As shown, in one embodiment of the present invention, step 400 includes:

[0072] Step 410: When the data feedback indicates a battery cell status fault, retrieve the standard deviation data and Pearson correlation coefficient at the time of the fault.

[0073] In the (secondary) main control module, when a battery cell undervoltage / overvoltage fault is detected based on the detection data, the filtering process of this embodiment of the invention is applied to determine the reliability of the data source to determine whether it is affected by interference. Based on the fault time, judgment parameter values ​​at the same time are retrieved, such as standard deviation data and Pearson correlation coefficient.

[0074] Step 420: When the standard deviation is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is greater than the correlation threshold, a fault is determined to have occurred.

[0075] In one embodiment of the present invention, the standard deviation S of the fault reporting time is checked. T If S T If |r| is ≤100, then the range of |r| is determined. If |r|>0.75, the fault is considered to have actually occurred and is reported to the vehicle controller.

[0076] Step 430: When the standard deviation value is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is less than the correlation threshold, it is determined that the collected data has drifted, and the state data of the previous time of the fault is used to replace the state data of the fault time.

[0077] In one embodiment of the present invention, the standard deviation S of the fault reporting time is checked. T If S T If |r| is ≤100, then the range of |r| is determined. If |r|≤0.75, it indicates that the correlation between the extreme values ​​of the individual voltages is low, and the collected data is too high or too low. The voltage data of the previous moment is used to replace the measured data of the current moment, and no fault information is reported to the vehicle controller.

[0078] Step 440: When the standard deviation value is greater than the standard deviation threshold, it is determined that there is electromagnetic interference causing the data collected by each cell in the battery pack where the fault is located to be affected. The state data at the time of the fault is replaced by the state data at the time of the fault.

[0079] In one embodiment of the present invention, if S T If the value is >100, it is assumed that at that moment, the voltage data of multiple battery cells measured by one or more chips are all too low or too high, caused by electromagnetic interference. The voltage data from the previous moment is used to replace the measurement data at this moment, and no fault information is reported to the vehicle controller. This is used to determine the cause of the fault and complete the data correction.

[0080] The error filtering method for power battery data in this invention utilizes time-series standard deviation data and Pearson correlation coefficient to form judgment factors for data acquisition errors and error types. By using these judgment factors and strategies, false alarm events are filtered, abnormal data is filtered out, and data errors are quantified, thereby reducing the probability of system false alarms.

[0081] The error filtering processing method for power battery data according to an embodiment of the present invention, in practical applications, includes the following filtering process for the voltage state of individual battery cells: Figure 3 As shown. In Figure 3 In this process, the filtering method detects the voltage state of individual battery cells. Time-series periodic detection generates time-series voltage data u for each battery cell within the battery pack. T,n And generate the time-series average voltage data x of each battery pack at the corresponding time. T,i Based on the time-series voltage mean data x T,i Real-time generation of standard deviation data S at the corresponding time point T Based on the maximum voltage value of the individual battery cells in the battery pack (Max). T and minimum voltage value Min T The Pearson correlation coefficient r is generated in real time for the corresponding time period. Then, the standard deviation data S is used... T和 The logical judgment strategy of Pearson correlation coefficient r determines the error data and error type, forming the necessary error correction process.

[0082] An embodiment of the present invention provides an error filtering processing device for power battery data, comprising:

[0083] A memory is used to store the program code of the error filtering processing method for power battery data in the above embodiments during the processing process;

[0084] The processor is used to execute the program code in the process of the error filtering processing method for power battery data in the above embodiments.

[0085] The processor 16 can be a DSP (Digital Signal Processor), an FPGA (Field-Programmable Gate Array), an MCU (Microcontroller Unit) system board, a SoC (System on a Chip) system board, or a PLC (Programmable Logic Controller) minimum system including I / O.

[0086] An embodiment of the present invention provides an error filtering processing device for power battery data, such as... Figure 4 As shown. In Figure 4 In this embodiment, the following are included:

[0087] The data acquisition module 10 is used to acquire the time-series state data of each battery cell and form the time-series state average data of the battery pack as a unit.

[0088] The discrete feature extraction module 20 is used to generate time series standard deviation data of the time series state mean of each battery pack;

[0089] The deviation feature extraction module 30 is used to generate Pearson correlation coefficients based on the time-series state extremes of individual battery cells in the battery pack.

[0090] The error filtering module 40 is used to determine the error status and correct the detection error based on the time series standard deviation data and Pearson correlation coefficient to form a judgment strategy.

[0091] like Figure 4 As shown, in one embodiment of the present invention, the data acquisition module 10 includes:

[0092] The individual cell signal acquisition unit 11 is used to acquire the timing status signal of each individual cell for a given state type.

[0093] The timing state forming unit 12 is used to form timing state data of individual battery cells in each battery pack using the battery pack as the detection unit;

[0094] The timing average forming unit 13 is used to form timing average data for each battery pack based on the timing state data.

[0095] like Figure 4 As shown, in one embodiment of the present invention, the discrete feature extraction module 20 includes:

[0096] Standard deviation forming unit 21 is used to generate standard deviation data S of the mean state of each battery pack at each time step based on the mean state of each battery pack in the time series. T .

[0097] like Figure 4As shown, in one embodiment of the present invention, the deviation feature extraction module 30 includes:

[0098] Correlation coefficient generating unit 31 is used to record the maximum state value Max of the individual cells in the battery pack at each time T. T and the minimum value of the state Min T The Pearson correlation coefficient r is formed based on the maximum state value Max and the minimum state value Min.

[0099] like Figure 4 As shown, in one embodiment of the present invention, the error filtering module 40 includes:

[0100] The activation unit 41 is used to retrieve the standard deviation data and Pearson correlation coefficient at the time of the fault when the data feedback indicates a battery cell state fault.

[0101] Feedback confirmation unit 42 is used to determine that a fault has occurred when the standard deviation value is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is greater than the correlation threshold.

[0102] The first identification unit 43 is used to determine that the collected data has drifted when the standard deviation value is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is less than the correlation threshold, and to use the state data of the previous time of the fault to replace the state data of the fault time.

[0103] The second identification unit 44 is used to determine that there is electromagnetic interference causing the data collected by each cell in the battery pack where the fault is located to be affected when the standard deviation value is greater than the standard deviation threshold, and to use the state data of the previous moment before the fault to replace the state data at the time of the fault.

[0104] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for error filtering processing of power battery data, characterized in that, include: Acquire the time-series state data of each battery cell and generate the average time-series state data of the battery pack as a unit; The time-series standard deviation data of the time-series state mean for each battery pack is generated; Pearson correlation coefficients are generated based on the time-series state extrema of individual cells within the battery pack; Based on time series standard deviation data and Pearson correlation coefficient, a judgment strategy is formed to determine the error status and correct the detection error.

2. The error filtering processing method for power battery data as described in claim 1, characterized in that, The timing states include voltage, current, or temperature.

3. The error filtering processing method for power battery data as described in claim 1, characterized in that, The formation of the time-series state mean data includes: Collect the timing status signal of each battery cell for the determined state type; Using the battery pack as the detection unit, the time-series status data of the individual battery cells in each battery pack are generated; The average time-series status data of each battery pack is generated based on the time-series status data.

4. The error filtering processing method for power battery data as described in claim 1, characterized in that, The formation of the time-series standard deviation data includes: The standard deviation S of the mean state of each battery pack at each time step is generated based on the mean state of each battery pack in the time series. T , Where M represents the number of data acquisition chips or battery packs. Let x be the mean of the state values ​​of the battery pack measured by each acquisition chip at time T. T,p This represents the average state of the battery pack measured by the p-th acquisition chip.

5. The error filtering processing method for power battery data as described in claim 1, characterized in that, The formation of the Pearson correlation coefficient includes: At each time T, record the maximum state value of each individual battery cell within the battery pack. T and the minimum value of the state Min T The Pearson correlation coefficient r is formed based on the maximum value (Max) and minimum value (Min) of the state. Where n is the number of individual battery cells.

6. The error filtering processing method for power battery data as described in claim 1, characterized in that, The judgment strategy based on time series standard deviation data and Pearson correlation coefficient to determine the error status includes: When data feedback indicates a battery cell status fault, retrieve the standard deviation data and Pearson correlation coefficient at the time of the fault. A fault is identified when the standard deviation is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is greater than the correlation threshold.

7. The error filtering processing method for power battery data as described in claim 6, characterized in that, Also includes: When the standard deviation is less than the standard deviation threshold and the absolute value of the Pearson correlation coefficient is less than the correlation threshold, it is determined that the collected data has drifted.

8. The error filtering processing method for power battery data as described in claim 6, characterized in that, Also includes: When the standard deviation value is greater than the standard deviation threshold, it is determined that there is electromagnetic interference affecting the data collected by each cell in the battery pack where the fault is located.

9. An error filtering processing device for power battery data, characterized in that, include: A memory for storing program code during the processing of the error filtering method for power battery data as described in any one of claims 1 to 8; A processor for executing the program code.

10. An error filtering and processing device for power battery data, characterized in that, include: The data acquisition module is used to acquire the time-series status data of each battery cell and generate the time-series status average data of the battery pack as a unit. The discrete feature extraction module is used to generate the time-series standard deviation data of the time-series state mean of each battery pack; The deviation feature extraction module is used to generate Pearson correlation coefficients based on the time-series state extremes of individual battery cells in the battery pack. The error filtering module is used to determine the error status and correct the detection error based on the time series standard deviation data and Pearson correlation coefficient to form a judgment strategy.

Images