Battery pack fault detection method, device, equipment and medium

Abstract

The embodiment of the invention provides a fault detection method and device for a battery pack, equipment and a medium, and is applied to the technical field of power electronics. According to the scheme, an interlaced voltage measurement structure is adopted, when data of the voltage sensors are abnormal, the position and the reason of a fault can be accurately positioned, only few voltage sensors need to be arranged, and the cost can be saved. Besides, during fault detection, discrete square wave signals are added to the obtained first initial voltage sequence and the second initial voltage sequence respectively, and two target voltage sequences are obtained; then determining a correlation coefficient between the two target voltage sequences; and detecting a fault of the battery pack based on the correlation coefficient. According to the embodiment of the invention, the discrete square wave signal is added into the acquired voltage sequence, so that errors existing in the acquisition process are eliminated, the calculated correlation coefficient is ensured to be closer to an actual value, and accurate battery fault detection can be carried out.

Description

Technical Field

[0001] The present application belongs to the field of power electronics technology, and in particular, relates to a method, device, equipment and medium for detecting a fault of a battery pack. Background Art

[0002] Due to manufacturing problems of the battery itself and the influence of usage conditions, various faults may occur in the battery pack, such as battery micro-short circuit. These faults seriously affect the service life and safety of the battery. Therefore, a fast and effective fault diagnosis method to ensure battery safety is crucial.

[0003] In the traditional voltage measurement structure, there is a one-to-one correspondence between the voltage sensor and the battery cell. A failure in the voltage sensor may cause a false alarm in the fault detection of the battery pack. If a redundant sensor is added, the hardware cost will increase. In addition, for a battery pack formed by multiple battery cells connected in series, multiple voltage sensors are usually provided to obtain the voltage, and then the correlation coefficient is calculated based on the different voltage sequences obtained to achieve battery fault detection. However, in practical applications, all measured voltages have measurement errors. When there is continuous input, since the voltage trend plays a dominant role in the correlation coefficient, a small measurement error will cause a large difference in the correlation coefficient. The calculated correlation coefficient is wrong, which makes it impossible to accurately detect battery faults. Summary of the invention

[0004] The embodiments of the present application provide a battery pack fault detection method, device, equipment and medium, which can accurately perform battery fault detection.

[0005] In one aspect, an embodiment of the present application provides a fault detection method for a battery pack, wherein the battery pack includes N battery cells connected in series, N-1 first voltage sensors, and 1 second voltage sensor; N is a positive integer;

[0006] Each of the first voltage sensors is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells located at both ends of the battery pack;

[0007] The method comprises:

[0008] Acquire a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor;

[0009] Adding discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence respectively to obtain a corresponding first target voltage sequence and a second target voltage sequence;

[0010] determining a correlation coefficient between the first target voltage sequence and the second target voltage sequence;

[0011] detecting a fault of the battery pack based on the correlation coefficient;

[0012] The relationship between the first initial voltage sequence and the second initial voltage sequence includes:

[0013] The voltages at the same sequence position are the voltages measured for the same battery pair at different times;

[0014] The voltages at the same sequence position are voltages measured at the same time for two adjacent battery pairs, and the two adjacent battery pairs have the same battery cell.

[0015] On the other hand, determining the correlation coefficient between the first target voltage sequence and the second target voltage sequence includes:

[0016] Acquire a first voltage average value of each voltage in the first target voltage sequence and a second voltage average value of each voltage in the second target voltage sequence;

[0017] The correlation coefficient is determined based on the first voltage average value and the second voltage average value.

[0018] On the other hand, determining the correlation coefficient based on the first voltage average value and the second voltage average value includes:

[0019] For each voltage in the first target voltage sequence, respectively obtain a difference from the first voltage mean to obtain a first difference value; and for each voltage in the second target voltage sequence, respectively obtain a difference from the second voltage mean to obtain a second difference value;

[0020] Determine a target product of the first difference and the second difference corresponding to the voltage at the same sequence position;

[0021] Summing the target products to obtain a product sum; summing the variances of the first differences to obtain a first variance sum; and summing the variances of the second differences to obtain a second variance sum;

[0022] The correlation coefficient is determined based on the product sum, the first variance sum, and the second variance sum.

[0023] On the other hand, the step of acquiring the first initial voltage sequence and the second initial voltage sequence measured by the voltage sensor includes:

[0024] Acquire the voltage collected by each of the voltage sensors at a preset target time;

[0025] Based on the voltages collected by the voltage sensor, the first initial voltage sequence is determined by taking the voltage between the first battery pair at the first end of the battery pack as the first position of the sequence, taking the voltage between the second battery pair as the last position of the sequence, and in the order from the first end to the second end of the battery pack;

[0026] Based on the voltages collected by the voltage sensor, the second initial voltage sequence is determined with the voltage between the second first battery pair at the first end of the battery pack as the first position of the sequence, the voltage between the first first battery pair at the first end of the battery pack as the last position of the sequence, and in order from the first end to the second end of the battery pack.

[0027] On the other hand, the preset target time is multiple;

[0028] After determining the correlation coefficient between the first target voltage sequence and the second target voltage sequence, the method further includes:

[0029] Combining the correlation coefficients corresponding to the preset target moments into a state space;

[0030] The detecting a fault of the battery pack based on the correlation coefficient comprises:

[0031] A failure of the battery pack is detected based on the state space.

[0032] On the other hand, the step of acquiring the first initial voltage sequence and the second initial voltage sequence measured by the voltage sensor includes:

[0033] Acquire a first voltage collected by each of the voltage sensors at a first target time;

[0034] Based on each of the first voltages, determine the first initial voltage sequence, with the first voltage between the first battery pair at the first end of the battery pack as the first position of the sequence, with the first voltage between the second battery pair as the last position of the sequence, and in order from the first end to the second end of the battery pack;

[0035] Acquire a second voltage collected by each of the voltage sensors at a second target time;

[0036] Based on each of the second voltages, the second initial voltage sequence is determined with the second voltage between the first pair of the first batteries at the first end of the battery pack as the first position of the sequence, the second voltage between the second pair of batteries as the last position of the sequence, and in order from the first end to the second end of the battery pack.

[0037] On the other hand, before adding discrete square wave signals to the voltages in the first initial voltage sequence and the second initial voltage sequence to obtain the corresponding first target voltage sequence and second target voltage sequence, the method further includes:

[0038] Get the accuracy level of battery fault detection;

[0039] The amplitude of the discrete square wave signal is determined based on the accuracy level of battery fault detection.

[0040] On the other hand, an embodiment of the present application provides a fault detection device for a battery pack, wherein the battery pack includes N battery cells connected in series, N-1 first voltage sensors and 1 second voltage sensor; N is a positive integer;

[0041] Each of the first voltage sensors is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells located at both ends of the battery pack;

[0042] The device comprises:

[0043] An acquisition module, used for acquiring a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor;

[0044] An adding module, used for adding discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence, respectively, to obtain a corresponding first target voltage sequence and a second target voltage sequence;

[0045] a determination module, configured to determine a correlation coefficient between the first target voltage sequence and the second target voltage sequence;

[0046] A detection module, configured to detect a fault of the battery pack based on the correlation coefficient;

[0047] The relationship between the first initial voltage sequence and the second initial voltage sequence includes:

[0048] The voltages at the same sequence position are the voltages measured for the same battery pair at different times;

[0049] The voltages at the same sequence position are voltages measured at the same time for two adjacent battery pairs, and the two adjacent battery pairs have the same battery cell.

[0050] In another aspect, an embodiment of the present application provides a battery pack fault detection device, the device comprising: a processor and a memory storing computer program instructions;

[0051] When the processor executes the computer program instructions, the battery pack fault detection method as described above is implemented.

[0052] On the other hand, an embodiment of the present application provides a computer-readable storage medium, on which computer program instructions are stored. When the computer program instructions are executed by a processor, the battery pack fault detection method described above is implemented.

[0053] On the other hand, an embodiment of the present application provides a computer program product. When instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the battery pack fault detection method as described above.

[0054] A method for fault detection of a battery pack provided in an embodiment of the present application adopts an interlaced voltage measurement structure, and each voltage sensor measures the voltage sum of two adjacent battery cells in the battery pack. The voltage sensors are connected in a cross manner to ensure that the voltage of each battery cell is associated with two voltage sensors, thereby improving the accuracy of fault diagnosis and increasing the detection sensitivity of internal short circuits of the battery. When the voltage sensor data is abnormal, the location and cause of the fault can be accurately located, and only fewer voltage sensors need to be set, which can save costs. In addition, when performing fault detection, discrete square wave signals are added to the first initial voltage sequence and the second initial voltage sequence respectively to obtain two target voltage sequences; then the correlation coefficient between the two target voltage sequences is determined; thereby detecting the fault of the battery pack based on the correlation coefficient. The embodiment of the present application adds a discrete square wave signal to the collected voltage sequence, thereby eliminating the error existing in the collection process, ensuring that the calculated correlation coefficient is closer to the actual value, and then accurate battery fault detection can be performed. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0056] Figure 1 A schematic diagram of a specific interleaved voltage measurement structure provided in an embodiment of the present application;

[0057] Figure 2 A schematic diagram showing a flow chart of a battery pack fault detection method provided by an embodiment of the present application;

[0058] Figure 3 A flow chart for calculating an intra-group correlation coefficient provided in an embodiment of the present application;

[0059] Figure 4A flow chart for calculating an inter-group correlation coefficient provided in an embodiment of the present application;

[0060] Figure 5 A flow chart for calculating intra-group and inter-group correlation coefficients provided in an embodiment of the present application;

[0061] Figure 6 A schematic diagram showing the structure of a battery pack fault detection device provided in an embodiment of the present application is shown;

[0062] Figure 7 A schematic diagram of the hardware structure of a battery pack fault detection device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0063] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.

[0064] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, 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 process, method, article or device. In the absence of further restrictions, an element defined by the statement "include..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0065] In order to solve the problems of the traditional solution, the embodiment of the present application provides a battery pack fault detection method, device, equipment and medium. The battery pack fault detection method provided by the embodiment of the present application is first introduced below.

[0066] There are many different numbers and connection methods of batteries in the IoT lock battery-powered system. Since the terminal voltage is the same, all battery cells are usually regarded as a whole. Therefore, the structure of the IoT lock battery-powered battery system can be divided into two types. One is to connect in parallel first and then in series, and the other is to connect in series first and then in parallel. In other words, most battery systems are based on series packs. Because voltage is very sensitive to electrical faults, except for extreme fault conditions (such as battery burnout and explosion), the remaining faults can be detected by measuring the voltage of each battery cell in series by a voltage sensor or sampling the voltage in the IoT lock integrated circuit.

[0067] In an embodiment of the present application, the battery pack includes N battery cells connected in series, N-1 first voltage sensors and 1 second voltage sensor; N is a positive integer; each first voltage sensor is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells located at both ends of the battery pack.

[0068] Figure 1 A schematic diagram of a specific interleaved voltage measurement structure provided in an embodiment of the present application; Figure 1 As shown, there are 5 battery cells 102 and 5 voltage sensors 101. The battery cells 102 are battery cell 1, battery cell 2, battery cell 3, battery cell 4 and battery cell 5 from left to right; the voltage sensors 101 are V1, V2, V3, V4 and V5 from left to right (V1 to V4 are the first voltage sensors, and V5 is the second voltage sensor).

[0069] The voltage parameters of the battery can be known from the battery design parameters, and the voltage values ​​of the two batteries at each moment can be known from the voltage sensor test. Each voltage sensor measures the sum of the voltages of the two batteries in the IoT lock. Each measuring line is directly connected to the electrode of the battery, so there is basically no contact resistance between the measuring line and the corresponding electrode. The voltage sensors are connected in a cross-connected manner to ensure that the voltage of each battery cell is associated with two voltage sensors.

[0070] Figure 1 Take the battery cell 2 from the left as an example, which is included in both voltage sensors V1 and V2. When the battery cell is in a short circuit state, the measured voltages of the two voltage sensors will drop sharply at the same time. Therefore, the fault characteristics are displayed by the two voltage sensors in cooperation with each other.

[0071] In addition, the values ​​collected by the voltage sensor can also be used to directly determine whether the voltage sensor is faulty. For example, assuming that the voltage of each battery cell is 3V; then V1 and V3 are both 6V, if V2 becomes 0V, assuming that the V2 test is normal, this means that there is an overcharge level on batteries 1 and 4 and a fault in the circuit from battery 4 to the outside, which cannot happen at the same time.

[0072] In the staggered voltage measurement structure of the embodiment of the present application, each voltage sensor measures the voltage sum of two adjacent battery cells in the battery pack. The voltage sensors are connected in a cross manner to ensure that the voltage of each battery cell is associated with two voltage sensors, thereby improving the accuracy of fault diagnosis and increasing the detection sensitivity of internal short circuits in the battery. When the voltage sensor data is abnormal, the location and cause of the fault can be accurately located, and only fewer voltage sensors need to be set, which can save costs.

[0073] Based on the interleaved voltage measurement structure proposed above, the embodiment of the present application proposes a corresponding fault detection method to detect micro-short circuits of the battery pack. Figure 2 A schematic flow chart of a battery pack fault detection method provided in one embodiment of the present application is shown.

[0074] like Figure 2 As shown, the method comprises the following steps:

[0075] S201: Acquire a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor.

[0076] The first initial voltage sequence and the second initial voltage sequence are used to detect faults in the battery pack, and the correlation coefficient of the two sequences needs to be obtained. The embodiment of the present application provides the following specific implementation method, and the relationship between the first initial voltage sequence and the second initial voltage sequence includes:

[0077] The voltages at the same sequence position are the voltages measured for the same battery pair at different times (between groups, which are two sequences in the same spatial domain).

[0078] The voltages at the same sequence position are voltages measured at the same time for two adjacent battery pairs, where the two adjacent battery pairs have a same battery cell (within a group, i.e., two sequences in the same time domain).

[0079] S202: adding discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence respectively to obtain a corresponding first target voltage sequence and a second target voltage sequence.

[0080] The Pearson correlation coefficient with a moving window is a statistical method used to evaluate the linear correlation of two sequences within a sliding window. It is a statistical method for measuring the degree of linear correlation between two variables. The correlation coefficient takes values ​​between -1 and 1, where 1 indicates a perfect positive correlation, -1 indicates a perfect negative correlation, and 0 indicates no linear correlation. The Pearson correlation coefficient is widely used in scientific research, engineering technology, data analysis and other fields, and is an important tool for measuring the linear relationship between variables.

[0081] Assuming that the battery system has no faults, their voltage correlation coefficients should be very close to 1. However, in practical applications, all measured voltages have measurement errors. When there is a continuous input, since the voltage trend plays a dominant role in the correlation coefficient, a small measurement error will cause a large difference in the correlation coefficient. In this case, the correlation coefficient will be much less than 1.

[0082] Since the battery system may be affected by measurement errors in actual use, these errors may cause the correlation coefficient to deviate from its true value, thereby affecting the accuracy of battery micro-short circuit detection. Common errors include gain error, offset error, quantization error, and noise error.

[0083] Gain error refers to the error in the proportional factor of the measurement system, which causes a fixed proportional difference between the measurement result and the actual value. Specifically, gain error will cause the measured voltage value to be too high or too low, usually due to inaccurate amplifier gain setting or improper calibration. Example: Assume that ideally, the measured voltage should be 3.7V, but due to the incorrect gain setting, the actual measured value may be 3.75V (positive gain error) or 3.65V (negative gain error).

[0084] Offset error is a fixed offset error in a measurement system, where a fixed offset is superimposed on all measurements. Offset error is usually due to a zero offset in the sensor or amplifier. Example: If a voltage measurement system has an offset error of 0.05V, a measured voltage of 3.7V may be recorded as 3.75V.

[0085] Quantization error is the error caused by the approximation of the continuous value of the signal to a discrete value during the digitization process of the analog signal. This error is inevitable in the digitization process. Example: Assume a 10-bit ADC (analog-to-digital converter) measures in the range of 0 to 5V, and its minimum resolution is 4.88mV. If the actual voltage is 3.702V, but it can only be quantized as 3.700V or 3.705V during measurement, the error is ±2.44mV.

[0086] Noise error refers to random fluctuations in measured values ​​caused by external electromagnetic interference, environmental noise, or electrical noise in the measurement system itself. Example: In an industrial environment, high-frequency noise may be superimposed on the measured voltage signal due to fluctuations in the surrounding electromagnetic field or power supply, causing random fluctuations in the measurement result in a short period of time.

[0087] The discrete square wave correction function is used to correct the error in voltage measurement. Since the measurement error may cause the correlation coefficient to deviate from its true value, thus affecting the judgment of the battery status, by introducing a discrete square wave signal that is independent of the voltage signal, this error can be eliminated to a certain extent, ensuring that the correlation coefficient is closer to the ideal value.

[0088] The discrete square wave function is a signal that is independent of the battery voltage. When it is added to the voltage measurement value, the influence of the measurement error on the correlation coefficient can be reduced. By reasonably selecting the amplitude of the square wave signal, its interference with the voltage signal can be ignored, making the correlation coefficient between the two close to 1. This provides a more accurate basis for the calculation of the correlation coefficient, thereby improving the sensitivity of battery micro-short circuit judgment.

[0089] The use of discrete square wave functions has the following advantages: 1. It can reduce measurement errors. In practice, voltage measurements are limited by sensor accuracy and have errors. Especially in battery micro-short circuit detection, small changes in battery state variables will be overwhelmed by errors. Adding discrete square wave functions can overcome this defect. 2. After adding discrete square wave signals, the impact of measurement errors on the correlation coefficient can be reduced, and the sensitivity to battery state changes can be enhanced. 3. It can enhance the accuracy of micro-short detection because the corrected correlation coefficient with discrete square wave signals is closer to the actual state of the battery. 4. It can correct the correlation. After the discrete square wave signal is added, the correlation coefficient is closer to 1, which helps to more accurately determine the correlation between batteries and thus identify micro-short circuits.

[0090] For both intra-group and inter-group situations, after adding the discrete square wave function, the intra-group correlation coefficient is closer to 1, indicating that the status of the same group of batteries is more consistent, reducing the probability of misjudging micro-short circuits. After adding the square wave, the inter-group correlation coefficient increases, indicating that the status of different groups is more correlated, enhancing the judgment of the overall health status of the battery system.

[0091] Specifically, in order to solve the error problem, the discrete square wave signal can be set to C (similar to a function), which is independent of the battery voltage measurement value. The discrete square wave signal C is added to the voltages in the first target voltage sequence and the second target voltage sequence. Then, when the battery current is zero, the signal becomes a quasi-square wave curve instead of a stable number. As a result, the correlation coefficient of the two voltages will be close to 1. In addition, by reasonably selecting the size of C, the influence of the added signal can be ignored.

[0092] Here is a specific example, C(i) is a discrete square wave function with an amplitude of c and a period of T0. Its main function is to correct the error in the measurement data and reduce the impact of the error on the correlation coefficient.

[0093]

[0094] c represents the amplitude of the discrete square wave. In specific applications, the value of c must be large enough to significantly correct the error, but not too large to avoid affecting the variance of the X and Y data. The amplitude of C should be selected to be several times the standard deviation of the error to ensure the effectiveness of the correction. At the same time, the value of c must be smaller than the variance of X and Y to ensure that the disturbance to the measured data is as small as possible.

[0095] T0 is the period of the discrete square wave, and the time window a needs to be an integer multiple of T0 to ensure that the complete square wave period can be accommodated in the window for correlation coefficient calculation, thereby avoiding the asymmetric effect of the irregular period on the result. The selection of T0 usually needs to be associated with the sampling period or measurement period of the system.

[0096] Parameter a is the length of the time window used for correlation coefficient calculation. a must be an integer multiple of T0 to ensure that the entire time window can accommodate one or more complete square wave cycles. This is to avoid incomplete cycles when calculating the correlation coefficient, which affects the accuracy of the result.

[0097] In practical applications, it is necessary to ensure the rationality of the assignment. Specifically, c is several times the error standard deviation, which can effectively correct the error; T0 is selected as a reasonable period, usually consistent with the data sampling period or an integer multiple thereof, to avoid the interference of asymmetric periods on the correlation coefficient; a is selected as an integer multiple of T0 to ensure that the complete period is included in the window.

[0098] Through the above method, it is possible to accurately determine whether the voltages between the two batteries are really consistent. If the correlation coefficient r' is close to 1, then the status of the two batteries can be considered normal; if r' deviates significantly from 1, there may be a problem (such as a micro short circuit or sensor failure).

[0099] The innovative application of the correction function can improve the accuracy of battery system diagnosis by assigning actual values ​​to the amplitude and period of the discrete square wave signal. The discrete square wave correction function is introduced to correct the voltage measurement data to reduce the impact of measurement errors on the correlation coefficient calculation results. This method effectively improves the accuracy of the correlation coefficient calculation by adding a square wave signal of a certain amplitude to the voltage data, thereby more accurately evaluating the battery status.

[0100] S203: Determine a correlation coefficient between the first target voltage sequence and the second target voltage sequence.

[0101] After adding the discrete square wave signal, the corresponding first target voltage sequence and second target voltage sequence are obtained, and then the correlation coefficient is solved based on the two sequences. Two voltage sequence situations are mentioned above, and different types of correlation coefficients are measured respectively, so that faults in different situations can be detected.

[0102] The first situation is that in the first target voltage sequence and the second target voltage sequence, the voltages at the same sequence position are voltages measured for the same battery pair at different times. In this case, the inter-group correlation coefficient can be calculated.

[0103] The second situation is that the voltages at the same sequence position are voltages measured for two adjacent battery pairs at the same time, and the two adjacent battery pairs have a same battery cell. In this case, the intra-group correlation coefficient can be calculated.

[0104] In practical applications, an analysis method combining intra-group and inter-group correlation coefficients can be used. This analysis method can flexibly adapt to different data sets and time windows and accurately reflect the mutual influence and state changes between battery cells.

[0105] S204: Detecting a fault in the battery pack based on the correlation coefficient.

[0106] Let the correlation coefficient be r:

[0107] In the case of r=1, it indicates that there is a completely positive linear relationship between the variables of the two sequences, that is, when one variable increases, the other variable also increases in a linear manner.

[0108] In the case of r=-1, it indicates that there is a perfect negative linear relationship between the two variables, that is, when one variable increases, the other variable decreases in a linear manner.

[0109] When r=0, it means there is no linear relationship between the two variables, but this does not mean there is no relationship between them. There may be a nonlinear relationship.

[0110] Assuming that the battery system has no fault, the voltage correlation coefficient of the two sets of sequences should be very close to 1. Therefore, after calculating the correlation coefficient of the two sets of sequences, the fault of the battery pack can be detected based on this principle. The specific determination method is not limited. For example, a preset interval near 1 can be set in advance. If the correlation coefficient is within this range, it is determined that there is no fault. Otherwise, it is determined that the battery pack has a fault.

[0111] The value range of the preset interval can be selected according to specific needs, and the following information can be used as a reference: 0.0<|r|<0.3 is weak correlation; 0.3≤|r|<0.5 is medium correlation; 0.5≤|r|<0.7 is strong correlation; 0.7≤|r|≤1.0 is strong correlation.

[0112] A method for fault detection of a battery pack provided in an embodiment of the present application adopts an interlaced voltage measurement structure, and each voltage sensor measures the voltage sum of two adjacent battery cells in the battery pack. The voltage sensors are connected in a cross manner to ensure that the voltage of each battery cell is associated with two voltage sensors, thereby improving the accuracy of fault diagnosis and increasing the detection sensitivity of internal short circuits of the battery. When the voltage sensor data is abnormal, the location and cause of the fault can be accurately located, and only fewer voltage sensors need to be set, which can save costs. In addition, when performing fault detection, discrete square wave signals are added to the first initial voltage sequence and the second initial voltage sequence respectively to obtain two target voltage sequences; then the correlation coefficient between the two target voltage sequences is determined; thereby detecting the fault of the battery pack based on the correlation coefficient. The embodiment of the present application adds a discrete square wave signal to the collected voltage sequence, thereby eliminating the error existing in the collection process, ensuring that the calculated correlation coefficient is closer to the actual value, and then accurate battery fault detection can be performed.

[0113] The embodiment of the present application provides a specific method for determining the correlation coefficient, by calculating the mean of each voltage in the voltage sequence, and then calculating the difference and product to quickly determine the value of the correlation coefficient. The specific steps are described below, and the correlation coefficient between the first target voltage sequence and the second target voltage sequence is determined, including:

[0114] First, obtain the first voltage mean of each voltage in the first target voltage sequence and the second voltage mean of each voltage in the second target voltage sequence. Then, determine the correlation coefficient based on the first voltage mean and the second voltage mean. In practical applications, the following method can be used to determine the correlation coefficient based on two means: for each voltage in the first target voltage sequence, obtain the difference with the first voltage mean to obtain the first difference; and for each voltage in the second target voltage sequence, obtain the difference with the second voltage mean to obtain the second difference. Then determine the target product of the first difference and the second difference corresponding to the voltage in the same sequence position; then sum each target product to obtain the product sum; and calculate the variance sum of each first difference to obtain the first variance sum; and calculate the variance sum of each second difference to obtain the second variance sum. Finally, determine the correlation coefficient based on the product sum, the first variance sum and the second variance sum.

[0115] Pearson correlation coefficient r X,Y Defined as:

[0116]

[0117] Where COV(X,Y) is the covariance of X and Y, σ X and σ Y are the standard deviations of X and Y respectively.

[0118] The covariance COV(X,Y) is defined as:

[0119]

[0120] Standard deviation X and σ Y They are defined as:

[0121]

[0122] x i and i are the voltage values ​​at the corresponding sequence position i, n is the sequence length, μ x and μ Y are the means of the voltages in the corresponding sequences X and Y, respectively.

[0123] Therefore, the original Pearson correlation coefficient formula can be written as:

[0124]

[0125] The numerator of this formula measures the sum of the products of how X and Y deviate from their means, which reflects how much they vary together. The denominator is the square root of the product of the sum of the squares of the deviations of X and Y, which is used to normalize this co-variation and ensure that the resulting correlation coefficient is dimensionless and ranges between -1 and 1.

[0126] The process of calculating the correlation coefficient is as follows: first calculate the mean, that is, calculate the mean of the voltage sensor voltage series X and Y. Then calculate the sum of the deviation products. For each battery in the group, calculate the product of the deviation of its voltage sensor variable (that is, the difference between each value and its mean), and then sum these products. Next, calculate the sum of the variances of the deviations of the voltage sensor variables X and Y respectively, and finally calculate the correlation coefficient.

[0127] The embodiment of the present application provides a specific method for calculating the correlation coefficient. By averaging, differencing, multiplying, and summing the variances of the voltages of the first target voltage sequence and the second target voltage sequence, the correlation coefficient of the two sequences can be efficiently calculated.

[0128] In practical applications, the amount of voltage sensor measurement data used as calculation samples is quite large. If the data set used in the calculation is too large or too small, the fault characteristics cannot be highlighted. Therefore, the embodiment of the present application provides a specific solution, which highlights the fault characteristics by setting the specific form of two initial voltage sequences. Specifically, obtaining the first initial voltage sequence and the second initial voltage sequence measured by the voltage sensor includes:

[0129] First, the voltage collected by each voltage sensor is obtained at a preset target time; then, based on each voltage collected by the voltage sensor, the voltage between the first first battery pair at the first end of the battery pack is used as the first position of the sequence, the voltage between the second battery pair is used as the last position of the sequence, and the first initial voltage sequence is determined in the order from the first end to the second end of the battery pack. Correspondingly, based on each voltage collected by the voltage sensor, the voltage between the second first battery pair at the first end of the battery pack is used as the first position of the sequence, the voltage between the first first battery pair at the first end of the battery pack is used as the last position of the sequence, and the second initial voltage sequence is determined in the order from the first end to the second end of the battery pack.

[0130] Here is a specific calculation process for the intraclass correlation coefficient. Figure 3 A flow chart for calculating an intra-group correlation coefficient provided in an embodiment of the present application; Figure 3 As shown, the following steps are included:

[0131] S301: Obtain a voltage value sequence X, Y at time T1.

[0132] S302: Find the mean of X and Y.

[0133] S303: Calculate the sum of the deviation products of each member in the X, Y group.

[0134] S304: Calculate the sum of the variances of the deviations of each member in the X, Y group.

[0135] S305: Solve the intra-group correlation coefficient; if the stop condition is met, enter S307; if the stop condition is not met, enter S306.

[0136] S306: Move window 1 by one position.

[0137] That is, window 1 is moved to the next sampling interval.

[0138] S307: Merge the intra-group correlation coefficients to form a state space expression of the intra-group correlation coefficients.

[0139] By calculating the intra-group correlation coefficient of the voltage measurement values ​​on the time window data in each time interval, the micro-short circuit state of the battery can be monitored. This method allows the abnormal state of the battery to be discovered in time by analyzing the correlation of the voltage data, increasing the accuracy of fault prediction.

[0140] The intra-group correlation coefficient is used to measure the correlation between multiple variables measured in the same group. In the battery detection scenario of the embodiment of the present application, the intra-group correlation coefficient is used to evaluate the linear relationship between the voltages of different battery terminals at the same time to determine the consistency of the battery status. Figure 1 Taking the structure of as an example, the scheme of the embodiment of the present application is described.

[0141] First, data collection is performed. At time T1, each voltage sensor data V1, V2, V3, V4 and V5 in the staggered voltage structure diagram at the same time is collected. By calculating the intra-group correlation coefficient of this set of data, it can be evaluated whether the voltages measured by these voltage sensors are consistent to some extent, so as to determine whether the battery is in a normal state. After data sorting, the voltage sensor variable X and the voltage sensor variable Y are set to obtain the following table:

[0142] Table 1 Voltage sequence table at time T1

[0143] X V1 V2 V3 V4 V5 Y V2 V3 V4 V5 V1

[0144] The following describes the situation where the data are normal and the situation where there are errors in the data.

[0145] Under normal circumstances, X = [V1, V2, V3, V4, V5]; Y = [V2, V3, V4, V5, V1]. Calculate the mean of the sequences X and Y:

[0146]

[0147] Since X and Y are essentially the same set of permutations and combinations of values, their means are the same.

[0148] The covariance formula is:

[0149]

[0150] Specifically for the current data, we can calculate:

[0151]

[0152] Variance calculation formula:

[0153]

[0154] For X and Y, calculate the corresponding variance:

[0155]

[0156] Since X and Y are the same permutation of values, their variances are the same.

[0157] Finally, the covariance and variance are brought into the Pearson correlation coefficient formula:

[0158]

[0159] By substituting the actual numerical calculation, we can get the correlation coefficient r between X and Y X,YAssuming the error is small, since the data of X and Y are just cyclic permutations, the correlation between them will be very close to 1, indicating that they are highly linearly correlated.

[0160] In the case of error, because the measurement error is large, the impact of the error on the Pearson correlation coefficient needs to be reconsidered, and the derivation process will have different results. In the case of large errors, the linear relationship of the original data may be significantly damaged, causing the Pearson correlation coefficient to be far away from 1 (that is, it is no longer a strong linear correlation). The error needs to be added to the model and the formula needs to be re-derived.

[0161] Assume that the measurement error is large, assuming:

[0162] X 测量 =X 真实 +∈ (15)

[0163] Y 测量 =Y 真实 +δ (16)

[0164] Among them, ∈ = [∈1, ∈2, …, ∈5] is the measurement error generated by X, and δ = [δ1, δ2, …, δ5] is the measurement error generated by Y.

[0165] Assuming that the errors are large and independently distributed, they may follow a certain distribution (such as normal distribution), that is,

[0166] Therefore, we have:

[0167] X 测量 =[V1+∈1,V2+∈2,V3+∈3,V4+∈4,V5+∈5] (17)

[0168] Y 测量 =[V2+δ1,V3+δ2,V4+δ3,V5+δ4,V1+δ5] (18)

[0169] Here X 测量 and Y 测量 All contain errors, and then recalculate the Pearson correlation coefficient.

[0170] The new mean needs to take into account the impact of the error and calculate the mean:

[0171]

[0172] Since the mean of the error is 0 (assuming the error has zero mean), we can get:

[0173]

[0174] However, when the error is large, the individual error is no longer 0, so the mean calculation is also affected.

[0175] The covariance after adding the error is calculated as follows:

[0176]

[0177] Expanding this formula yields:

[0178]

[0179] The formula is divided into four terms, represents the covariance of the real data, Representation error∈ i The product term between and the real Y, Represents the error δ i The product term between the real X, ∈ i δ i Represents the product term between errors.

[0180] If we assume that the error ∈ i and δ i If they are independent of each other and have a mean of 0, the covariance formula can be simplified to:

[0181] cov(X 测量 ,Y 测量 )≈cov(X 真实 ,Y 真实 )(25)

[0182] However, if the error is large, these assumptions often do not hold true and the impact of the error term will be large.

[0183] The calculation of the variance is also affected by the error. The new variance is:

[0184]

[0185] Expand to get:

[0186]

[0187] Similarly, for Y 测量 :

[0188]

[0189] in, and Represents the variance of the error. When the error is large, the variance will also increase significantly.

[0190] Recalculate the Pearson correlation coefficient, and finally get the formula of the Pearson correlation coefficient:

[0191]

[0192] Since the covariance term remains basically unchanged, while the variance term increases significantly (due to the sum of squares of the error term), the overall correlation coefficient r becomes smaller. In other words, a larger measurement error will weaken the correlation between X and Y, causing the Pearson correlation coefficient to be far away from 1, and may even become a weak correlation close to 0.

[0193] It can be seen that when the error is large, the derivation of the Pearson correlation coefficient needs to consider the impact of the error on the covariance and variance. The covariance may remain relatively stable, but the variance increases significantly, resulting in a decrease in the correlation coefficient. The larger the error, the weaker the correlation. This also shows that measurement accuracy has an important impact on correlation analysis, and data with large errors will obscure the original linear relationship.

[0194] The following example illustrates the effect of measurement accuracy on correlation.

[0195] Example 1: Assume that the voltage sensor measurements are as follows: V1 is 3.71, V2 is 3.82, V3 is 3.55, V4 is 4.01, and V5 is 3.69. The calculated correlation coefficient r = -0.7. This means that there is a weak correlation between these batteries. From the data, it can be seen that the voltages of V3 and V4 deviate significantly compared to the other batteries. Characterizing the intra-group correlation coefficient, a fault has been inferred.

[0196] Example 2: Assume that V1 is 3.7, V2 is 3.8, V3 is 3.7, V4 is 3.6, and V5 is 3.7. Calculate the intra-group correlation coefficient between these measured values ​​to determine whether they are consistent in voltage. The calculated correlation coefficient r = 0. Although the two sets of data appear to have a linear relationship, due to accuracy reasons, the data shifted right by one position will produce a partial positive and negative offset problem when calculating the accumulated deviation, so the correlation coefficient of this type of shifted data does not necessarily reflect a strong correlation.

[0197] The data calculation at this time can be considered that the system itself has errors. At this time, the judgment of the intra-group correlation coefficient is no longer convincing, so the explanation of the inter-group correlation coefficient is introduced below.

[0198] The embodiment of the present application can determine whether the micro-short circuit state of the battery is reasonable by calculating the intra-group correlation coefficient of the voltages collected by two voltage sensors related to a battery cell at the same time, and can quickly and accurately determine whether the battery is faulty.

[0199] In the above embodiment, one of the relationships between the first initial voltage sequence and the second initial voltage sequence is that the voltages at the same sequence position are the voltages measured at the same time for two adjacent battery pairs, and the two adjacent battery pairs have the same battery cell. At this time, the calculated correlation coefficient corresponds to a preset target time. In order to better perform fault detection, in a specific implementation, multiple preset target times can be set, and the specific number is selected according to needs.

[0200] Specifically, when sampling, a moving window can be added to perform measurements, and voltage measurements and intra-group correlation calculations can be performed at multiple times. For a moving window, the calculation of the above correlation coefficient is repeated each time the window moves, and only the data points within the window are considered each time. This means that for each window position, a correlation coefficient is calculated, thereby obtaining a sequence of correlation coefficients that changes over time. The following are the values ​​of each preset target moment:

[0201]

[0202] Then the intra-group correlation coefficients are merged. From the above analysis, the voltage correlation comparison structure of the same voltage sensor at different times has been obtained. Now the correlation test results are used as data for the correlation again. Assuming that the total measurement volume of X and Y is n, the amount of data in the sample window is designed to be a, and the state space of each correlation coefficient can be obtained:

[0203]

[0204] x i,a+i It is the data from i to a+i of the X vector. i,a+i is the data from i to a+i of the Y vector, r is x i,a+i and i,a+i The correlation coefficient between .

[0205] After determining the correlation coefficient between the first target voltage sequence and the second target voltage sequence, the correlation coefficients corresponding to the preset target moments may be combined into a state space, and finally the fault of the battery pack may be detected based on the state space.

[0206] The voltage measurement values ​​of each voltage sensor in the battery pack at different times may fluctuate due to the sensor's own errors or external interference. Relying solely on the correlation coefficient at a certain moment may not accurately reflect the true state of the battery. By combining the correlation coefficient, the overall consistency of the sensor at different time points can be more comprehensively evaluated, thereby reducing the impact of the measurement error at a single moment on the judgment result, thereby reducing the possibility of misjudgment.

[0207] After the correlation coefficients are merged within the battery pack, the relatively stable electrical characteristics of the battery at multiple time points can be better captured. If the correlation coefficients at different time points still show a high correlation after merging, it means that the battery state is relatively stable; conversely, if the correlation after merging is significantly reduced, it may indicate that there is a micro-short circuit or other abnormal conditions inside the battery. Therefore, the accuracy of short circuit detection can be improved after merging. It can be seen that after obtaining the correlation coefficients within each group at multiple times, the embodiment of the present application merges these correlation coefficients to improve the accuracy and reliability of battery micro-short circuit detection.

[0208] The above embodiment provides a specific solution for the first initial voltage sequence and the second initial voltage sequence, by calculating the intra-group correlation coefficient of the voltages collected by two voltage sensors related to a battery cell at the same time, but the intra-group correlation coefficient is insufficient in battery status detection, and the introduction of the inter-group correlation coefficient makes up for these problems. It is mainly reflected in the following aspects:

[0209] The intra-group correlation coefficient has the limitation of time dimension. It is mainly used to compare the voltage changes between multiple batteries in the same time window, and can determine their mutual relationship at the same time. However, it cannot capture the trend of battery voltage change over time, nor can it reflect the overall change trend of the battery in different time periods. The inter-group correlation coefficient can capture the law of battery change over time by calculating the correlation of voltage between different time windows, thus making up for the deficiency of the intra-group correlation coefficient in time dimension.

[0210] The intra-group correlation coefficient only focuses on the correlation at a single moment and only evaluates the correlation between batteries at a specific moment, which cannot fully reflect the health status of the battery. The inter-group correlation coefficient evaluates the long-term changes in battery voltage by comparing data from different time periods, thus providing a more comprehensive diagnostic basis.

[0211] The intra-group correlation coefficient is not sensitive enough to fault trends. Although it can detect battery micro-short circuits at the same time, if the battery fault develops gradually (such as voltage gradually decreasing or battery aging), the intra-group correlation coefficient alone may not be able to detect the accumulation of these errors in time. The inter-group correlation coefficient can detect the correlation changes of voltage data at different times and detect the slow degradation of battery performance or potential trend problems in time.

[0212] Therefore, the embodiment of the present application introduces the inter-group correlation coefficient, which specifically includes:

[0213] At the first target moment, the first voltage collected by each voltage sensor is obtained, and then based on each first voltage, the first voltage between the first first battery pair at the first end of the battery pack is used as the first position of the sequence, the first voltage between the second battery pair is used as the last position of the sequence, and the first initial voltage sequence is determined in order from the first end to the second end of the battery pack.

[0214] At the second target moment, the second voltage collected by each voltage sensor is obtained; then based on each second voltage, the second initial voltage sequence is determined with the second voltage between the first battery pair at the first end of the battery pack as the first position of the sequence, the second voltage between the second battery pair as the last position of the sequence, and in order from the first end to the second end of the battery pack.

[0215] Here is a specific calculation process for the inter-group correlation coefficient. Figure 4 A flow chart for calculating the inter-group correlation coefficient provided in the embodiment of the present application; Figure 4 As shown, the following steps are included:

[0216] S401: Obtain a voltage sequence X at time T1.

[0217] S402: Solve for the mean of the X sequence variable in group T1.

[0218] S403: Calculate the difference between the T1 group and the X sequence variables.

[0219] S404: Calculate the product of the difference at time T1.

[0220] S405: Obtain the voltage value Y at time T2.

[0221] S406: Solve for the mean of the Y sequence variable for group T2.

[0222] S407: Solve the difference between the T2 group and the Y sequence variables.

[0223] S408: Calculate the product of the difference at time T2.

[0224] S409: Calculate the inter-group correlation coefficient.

[0225] S410: Move window 2 by one position.

[0226] S411: Move window 3 by one position.

[0227] First calculate the average value: for the voltage data at each time point, calculate its average value. That is, the voltage average value of the measured values ​​of the same battery cell at each time point, T1, T2, etc. Then calculate the difference: for the voltage data at each time point, calculate the difference between it and the corresponding average value. Then calculate the product: calculate the product of the difference between each time point. Finally, use the correlation coefficient formula to calculate the correlation coefficient between each pair of time points.

[0228] The inter-group calculation can also add a moving window to measure the same battery voltage at multiple times and calculate the inter-group correlation. T1 and T2 (window 2 and window 3) differ by a moving window position of the intra-group correlation coefficient, so in the actual calculation, the inter-group correlation coefficient can be calculated at the same time for each intra-group correlation coefficient calculation.

[0229] For each battery, the correlation coefficient at different times is calculated, thus obtaining a correlation coefficient sequence that changes over time:

[0230]

[0231] The inter-group correlation coefficient is generally not merged during analysis because it is mainly used to evaluate the overall linear relationship between the variables of the same group of batteries in different time windows. The purpose of the inter-group correlation coefficient is to observe the change in the correlation coefficient between time windows, so as to determine whether the correlation of the battery group at different times has changed significantly. Therefore, the inter-group correlation coefficient within each time window should remain independent in order to accurately capture the characteristics of each time window. This independence is meaningful for analyzing the changes in battery status over time. If merged, the sensitivity to changes in the time dimension may be lost, which in turn affects the accuracy of detecting battery micro-short circuits.

[0232] The inter-group correlation coefficient is used to measure the correlation between the same battery variables at different time points, which can solve the problem of misjudgment of the intra-group coefficient. Assume that the measured voltage of battery V1 at time points T1 and T2 is V 1T1 and V 1T2 , we can calculate V 1T1 and V 1T2 If this correlation coefficient is low, it may indicate that the battery voltage is unstable over time, indicating a possible electrical fault.

[0233] The inter-group correlation coefficient is mainly used to determine whether the voltage changes of the same battery at different time points are consistent. The voltage data of V1 of the same battery, the value measured by the voltage sensor in the time window T1, and the value measured by the voltage sensor in the time window T2, these two sets of variables can be used to calculate the inter-group correlation coefficient.

[0234] According to the time window, determine the group, first count the voltage data in different time windows (the first target time is T1, the second target time is T2):

[0235] Table 2 Voltage data table under different time windows

[0236] T1-X T2-Y V1 V1′ V2 V2′ V3 V3′ V4 V4′ V5 V5′

[0237] Normally, there are two variable sequences: T1 = X and T2 = Y; X = V1, V2, V3, V4, V5; Y = V1′, V2′, V3′, V4′, V5′. i and V i ′ is the observation value of the corresponding group, which comes from two different groups. The derived Pearson correlation coefficient actually measures the linear relationship between the two variable sequences. The Pearson correlation coefficient formula is defined as follows:

[0238]

[0239] Among them, X i and Y i is the ith observation of variables X and Y, and are the means of X and Y respectively.

[0240] Compute the mean of each variable:

[0241]

[0242] Here, X i and Y i represents the observations of the two groups respectively, and n is the number of observations (in this example, the number of observations from V1 to V5 and V1' to V5' is 5). For the observations V1, V2, V3, V4, V5 of group T1 = X, calculate the mean

[0243]

[0244] Similarly, for the observations V′1, V′2, V′3, V′4, V′5 of group T2=Y, we calculate the mean

[0245]

[0246] Next, we compute the sum of the deviation products, which is the product of the deviations of each observation from its mean and sum them:

[0247]

[0248] Need to calculate:

[0249]

[0250] The following calculates the sum of squared deviations for X and Y respectively:

[0251]

[0252] Finally, substitute these calculation results into the formula of Pearson correlation coefficient to calculate Pearson correlation coefficient:

[0253]

[0254] Through the above formula, the Pearson correlation coefficient r between the two variable sequences X and Y is obtained. The following examples are given for the two cases where there is a systematic accuracy error and a measurement error.

[0255] Voltage measurement values ​​at two target times when there is a system accuracy error

[0256] Voltage code The measured value at time T1 The measured value at time T2 V1 3.7 3.6 V2 3.8 3.7 V3 3.7 3.6 V4 3.6 3.5 V5 3.7 3.7

[0257] In the case of systematic precision error, r = 0.846. The calculation of the inter-group correlation coefficient supplements the solution to the misjudgment of the intra-group correlation coefficient in the case of measurement precision error.

[0258] Voltage measurement values ​​at two target times when there is a measurement error

[0259] Voltage code The measured value at time T1 The measured value at time T2 V1 3.71 3.70 V2 3.82 3.80 V3 3.55 3.52 V4 4.01 3.85 V5 3.69 3.66

[0260] In the presence of measurement errors, r≈0.9546, indicating that the battery voltage measurements have a strong positive correlation between time T1 and T2. This means that at these two time points, the battery voltage change trend is basically the same, and the overall voltage change of the system is relatively stable. The inter-group correlation coefficient shows that the change trend of the sequence is consistent, that is, the voltage decay rate is consistent; and the intra-group correlation coefficient shows whether the sequence has a more serious system error.

[0261] The embodiment of the present application calculates the correlation coefficient by collecting the voltage at two different times through a voltage sensor, which is used to evaluate the overall linear relationship between the variables of the same group of batteries in different time windows. The purpose of the scheme of this embodiment is to observe the change of the correlation coefficient between the time windows, so as to determine whether the correlation of the battery pack at different times has changed significantly, and then accurately detect faults.

[0262] It can be seen that the intra-group correlation coefficient can determine whether there is an accuracy error in the system itself, or whether there is inconsistency in the battery cells at the beginning of the system operation. For example, in door locks, we often use 8 batteries, 4 in series and 2 in parallel. If the user mixes new and old batteries when installing the batteries, the system can make a timely judgment.

[0263] The inter-group correlation coefficient is used to determine the overall change of a group of batteries. In the actual use of door lock batteries, the capacity consumption rate of a certain battery may be relatively fast, resulting in problems with the overall power broadcast. We can promptly determine the abnormal location and improve the accuracy of subsequent fault detection. Therefore, in order to apply to various scenarios in actual applications, it is very necessary to combine within and between groups.

[0264] Here is a specific calculation process for the intra-group and inter-group correlation coefficients. Figure 5 A flow chart for calculating intra-group and inter-group correlation coefficients provided in an embodiment of the present application; Figure 5 As shown, the following steps are included:

[0265] S501: Obtain voltage sensor test value.

[0266] S502: Obtain the voltage value at time T1.

[0267] S503: Summary of intra-group correlation coefficients.

[0268] S504: The discrete square wave function in group T1 corrects the voltage measurement error; if the stop condition is met, the process goes to S508; if the stop condition is not met, the process goes to S511.

[0269] S505: Obtain the voltage value at time T2.

[0270] S506: Summary of intra-group correlation coefficients.

[0271] S507: The discrete square wave function in group T2 corrects the voltage measurement error; if the stop condition is met, enter S508; if the stop condition is not met, enter S512.

[0272] S508: Summary of intra-group correlation coefficients.

[0273] S509: Summary of inter-group correlation coefficients.

[0274] S510: Correcting the voltage measurement error by using the discrete square wave function between groups; if the stop condition is met, proceed to S513; if the stop condition is not met, proceed to S511 and S512.

[0275] S511: Move window 4 by one position.

[0276] S512: Move window 5 by one position.

[0277] S513: Summary of intra-group and inter-group correlation coefficients.

[0278] Here, the window 4 and the window 5 may differ by one sampling unit. The above stop conditions may be set according to actual needs. For example, the number of collected voltage values ​​can reflect the actual state of the battery pack.

[0279] The following is an explanation of how to add a discrete square wave signal for correction. The expression of the square wave function is:

[0280]

[0281] The correlation coefficients before and after the square wave correction are calculated respectively. Original correlation coefficient calculation:

[0282]

[0283] in, and are the means of the sequences X(i) and Y(i) respectively. When the discrete square wave C(i) is introduced, the new correlation coefficient can be written as:

[0284]

[0285] here and is the adjusted series mean.

[0286] Next, we simplify the expression. Since the mean of the discrete square wave C(i) is zero (because it is a symmetrical square wave), the corrected mean Same reason Therefore, the formula can be simplified to:

[0287]

[0288] The numerator can be expanded:

[0289]

[0290] Due to the symmetry of C(i), it cancels out the mean terms of X(i) and Y(i) (assuming X(i) and Y(i) are symmetric and unbiased), so the main remaining terms are:

[0291]

[0292] This expression shows that the correlation coefficient increases after adding the square wave, but due to the small amplitude of C(i), the corrected correlation coefficient is still close to the original correlation coefficient.

[0293] Through the correction of square wave function, the correlation coefficient r X′Y′ It will be closer to 1, indicating that the influence of measurement error on correlation is effectively suppressed. Moreover, by properly selecting c and T0, it can be ensured that the influence of discrete square wave on data is minimal, so that the corrected data can still correctly reflect the real state of the battery.

[0294] The expression of square wave function is:

[0295]

[0296] Substitute the square wave function into the calculation of the correlation coefficient:

[0297] r C (X,Y)=r(X+C,Y+C)=[R C1 R C2 … R C(n-a) ] T (50)

[0298] The modified correlation coefficient expression is:

[0299]

[0300] The actual derivation of the parameter selection is as follows:

[0301] When choosing the amplitude c, it is usually set to be several times the measurement error to ensure that the correction signal is strong enough but does not affect the variance of X and Y. For example, if the standard deviation of the measurement error is 0.01, c = 0.03 can be selected to ensure that the correction is effective.

[0302] When selecting the period T0, in order to ensure the symmetry of the signal and not interfere with the original data, T0 is selected as an integer multiple of the data sampling period. For example, if the data sampling frequency is once per second, T0 = 6 seconds can be selected.

[0303] When selecting the time window a, a should be an integer multiple of T0 to ensure that the entire window contains a complete square wave cycle and avoid the influence of the cycle incompleteness on the correlation coefficient calculation. For example, selecting a = 12 seconds ensures that there are two complete square wave cycles in one window.

[0304] C(i) is a discrete square wave function with amplitude c and period T0. It should be noted that a needs to be an integer multiple of T0. The amplitude of C should be several times the standard deviation of the error and smaller than the variance of X and Y. In this way, the discrete square wave signal has little effect on the calculation result of the correlation coefficient.

[0305] In practical applications, the accuracy level of battery fault detection can also be obtained; then, based on the accuracy level of battery fault detection, the amplitude of the discrete square wave signal is determined. Common accuracy levels include scientific research level, engineering level, and consumer level. In scientific research, the accuracy may need to reach an extremely high level. In engineering applications, the accuracy is moderate, depending on the specific application and equipment. In consumer products, the accuracy is low. Select the value under the minimum accuracy requirement of the engineering level: a can be selected as 30. T0 is set to 6 seconds. Considering the voltage measurement error and variation characteristics, c is selected as 0.01.

[0306] The embodiment of the present application sets different amplitudes based on different accuracy levels to meet actual detection needs. By correcting the errors of the intra-group correlation coefficient and the inter-group correlation coefficient, the occurrence of battery micro-short circuit can be judged more sensitively.

[0307] To solve the above technical problems, the embodiment of the present application further provides a fault detection device for a battery pack, the battery pack comprising N battery cells connected in series, N-1 first voltage sensors and 1 second voltage sensor; N is a positive integer. Each first voltage sensor is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells at both ends of the battery pack.

[0308] Figure 6 FIG. 1 is a schematic diagram showing the structure of a battery pack fault detection device provided in an embodiment of the present application. Figure 6 As shown, the device includes the following modules:

[0309] An acquisition module 601 is used to acquire a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor;

[0310] An adding module 602 is used to add discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence to obtain a corresponding first target voltage sequence and a second target voltage sequence;

[0311] A determination module 603, configured to determine a correlation coefficient between the first target voltage sequence and the second target voltage sequence;

[0312] A detection module 604, configured to detect a fault of the battery pack based on the correlation coefficient;

[0313] The device provided in the embodiment of the present application is the same as the method in the above embodiment, so both have the same embodiments and beneficial effects, which will not be repeated here.

[0314] Figure 7 FIG. 1 is a schematic diagram showing the hardware structure of a battery pack fault detection device provided in an embodiment of the present application. Figure 7As shown, the fault detection device of a battery pack may include a processor 701 and a memory 702 storing computer program instructions.

[0315] Specifically, the processor 701 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0316] The memory 702 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 702 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 702 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 702 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 702 is a non-volatile solid-state memory.

[0317] The memory 702 may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk storage medium device, an optical storage medium device, a flash memory device, an electrical, optical or other physical / tangible memory storage device. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0318] The processor 701 implements any one of the battery pack fault detection methods in the above embodiments by reading and executing computer program instructions stored in the memory 702 .

[0319] In one example, the battery pack fault detection device may further include a communication interface 703 and a bus 704. The processor 701, the memory 702, and the communication interface 703 are connected via the bus 704 and communicate with each other.

[0320] The communication interface 703 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0321] Bus 704 includes hardware, software or both, and the parts of the fault detection equipment of the battery pack are coupled to each other. For example, but not limitation, the bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 704 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.

[0322] In addition, in combination with the battery pack fault detection method in the above embodiment, the present application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when the computer program instructions are executed by a processor, any of the battery pack fault detection methods in the above embodiment is implemented.

[0323] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0324] The functional blocks shown in the block diagram above can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. The code segment can be downloaded via a computer network such as the Internet, an intranet, etc.

[0325] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0326] The above refers to the flowchart and / or block diagram of a method, device, equipment and medium for fault detection of a battery pack according to an embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of boxes in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the functions / actions specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0327] The above contents are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present application, and these modifications or replacements should be included in the protection scope of the present application.

Claims

1. A battery pack fault detection method, characterized in that: The battery pack includes N battery cells connected in series, N-1 first voltage sensors and 1 second voltage sensor; N is a positive integer; Each of the first voltage sensors is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells located at both ends of the battery pack; The method comprises: Acquire a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor; Adding discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence respectively to obtain a corresponding first target voltage sequence and a second target voltage sequence; determining a correlation coefficient between the first target voltage sequence and the second target voltage sequence; detecting a fault of the battery pack based on the correlation coefficient; The relationship between the first initial voltage sequence and the second initial voltage sequence includes: The voltages at the same sequence position are the voltages measured for the same battery pair at different times; The voltages at the same sequence position are voltages measured at the same time for two adjacent battery pairs, and the two adjacent battery pairs have the same battery cell.

2. The battery pack fault detection method according to claim 1, characterized in that: The determining a correlation coefficient between the first target voltage sequence and the second target voltage sequence includes: Acquire a first voltage average value of each voltage in the first target voltage sequence and a second voltage average value of each voltage in the second target voltage sequence; The correlation coefficient is determined based on the first voltage average value and the second voltage average value.

3. The battery pack fault detection method according to claim 2, characterized in that: The determining the correlation coefficient based on the first voltage average value and the second voltage average value includes: For each voltage in the first target voltage sequence, respectively obtain a difference from the first voltage mean to obtain a first difference value; and for each voltage in the second target voltage sequence, respectively obtain a difference from the second voltage mean to obtain a second difference value; Determine a target product of the first difference and the second difference corresponding to the voltage at the same sequence position; Summing the target products to obtain a product sum; summing the variances of the first differences to obtain a first variance sum; and summing the variances of the second differences to obtain a second variance sum; The correlation coefficient is determined based on the product sum, the first variance sum, and the second variance sum.

4. The battery pack fault detection method according to claim 2 or 3, characterized in that: The step of acquiring the first initial voltage sequence and the second initial voltage sequence measured by the voltage sensor comprises: Acquire the voltage collected by each of the voltage sensors at a preset target time; Based on the voltages collected by the voltage sensor, the first initial voltage sequence is determined by taking the voltage between the first battery pair at the first end of the battery pack as the first position of the sequence, taking the voltage between the second battery pair as the last position of the sequence, and in the order from the first end to the second end of the battery pack; Based on the voltages collected by the voltage sensor, the second initial voltage sequence is determined with the voltage between the second first battery pair at the first end of the battery pack as the first position of the sequence, the voltage between the first first battery pair at the first end of the battery pack as the last position of the sequence, and in order from the first end to the second end of the battery pack.

5. The battery pack fault detection method according to claim 4, characterized in that: After determining the correlation coefficient between the first target voltage sequence and the second target voltage sequence, the method further includes: Obtaining the correlation coefficients corresponding to different preset target moments; Combining the correlation coefficients into a state space; The detecting a fault of the battery pack based on the correlation coefficient comprises: A failure of the battery pack is detected based on the state space.

6. The battery pack fault detection method according to claim 2 or 3, characterized in that: The step of acquiring the first initial voltage sequence and the second initial voltage sequence measured by the voltage sensor comprises: Acquire a first voltage collected by each of the voltage sensors at a first target time; Based on each of the first voltages, determine the first initial voltage sequence, with the first voltage between the first battery pair at the first end of the battery pack as the first position of the sequence, with the first voltage between the second battery pair as the last position of the sequence, and in order from the first end to the second end of the battery pack; Acquire a second voltage collected by each of the voltage sensors at a second target time; Based on each of the second voltages, the second initial voltage sequence is determined with the second voltage between the first pair of the first batteries at the first end of the battery pack as the first position of the sequence, the second voltage between the second pair of batteries as the last position of the sequence, and in order from the first end to the second end of the battery pack.

7. The battery pack fault detection method according to claim 1, characterized in that: Before adding discrete square wave signals to the voltages in the first initial voltage sequence and the second initial voltage sequence to obtain the corresponding first target voltage sequence and second target voltage sequence, the method further includes: Get the accuracy level of battery fault detection; The amplitude of the discrete square wave signal is determined based on the accuracy level of battery fault detection.

8. A battery pack fault detection device, characterized in that: The battery pack includes N battery cells connected in series, N-1 first voltage sensors and 1 second voltage sensor; N is a positive integer; Each of the first voltage sensors is used to measure the voltage between a first battery pair consisting of two adjacent battery cells, and the second voltage sensor is used to measure the voltage between a second battery pair consisting of two battery cells located at both ends of the battery pack; The device comprises: An acquisition module, used for acquiring a first initial voltage sequence and a second initial voltage sequence measured by a voltage sensor; An adding module, used for adding discrete square wave signals to each voltage in the first initial voltage sequence and the second initial voltage sequence, respectively, to obtain a corresponding first target voltage sequence and a second target voltage sequence; a determination module, configured to determine a correlation coefficient between the first target voltage sequence and the second target voltage sequence; A detection module, configured to detect a fault of the battery pack based on the correlation coefficient; The relationship between the first initial voltage sequence and the second initial voltage sequence includes: The voltages at the same sequence position are the voltages measured for the same battery pair at different times; The voltages at the same sequence position are voltages measured at the same time for two adjacent battery pairs, and the two adjacent battery pairs have the same battery cell.

9. A battery pack fault detection device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the fault detection method for the battery pack according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the fault detection method for a battery pack according to any one of claims 1 to 7 is implemented.

Images