Power battery heating film fault detection method and system

Abstract

The invention belongs to the technical field of power battery detection, and particularly relates to a power battery heating film fault detection method and system. The method comprises the following steps: inputting the temperature rise rate of a power battery in a historical heating stage into a trained time sequence prediction model to obtain an upper bound and a lower bound of a temperature rise rate prediction value at a prediction moment; then acquiring an actual temperature rise rate value of the power battery at the prediction moment, and if the actual temperature rise rate value is not between the upper bound and the lower bound of the corresponding temperature rise rate prediction value and the difference between the actual temperature rise rate value and the corresponding temperature rise rate prediction value is greater than a set difference threshold value, judging that the heating film of the power battery has a fault; the problem that an existing power battery fixed threshold value early warning method is low in accuracy is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power battery detection, and in particular relates to a power battery heating film fault detection method and system. Background Art

[0002] As the core component of new energy vehicles, the performance and safety of power batteries are crucial to vehicle operation. In the cold winter, the heating performance of power batteries is particularly critical, because low temperatures can cause battery performance to decline, charging to slow down, and may even cause safety hazards. The power battery heating film is a key component used to improve the performance of batteries in low temperature environments. Under long-term use or the influence of external factors, the power battery heating film may have faults such as uneven heating, local overheating, and short circuits, resulting in battery performance degradation and even safety accidents. Therefore, in order to ensure the reliability and safety of power batteries in low temperature environments, it is necessary to monitor the working status of the heating film in real time, detect potential faults in a timely manner, and issue early warnings.

[0003] Traditional power battery heating film fault detection generally adopts a method based on fixed thresholds, that is, engineers manually set warning conditions based on data indicators. For example, when the data indicator is greater than or less than a certain set threshold, the power battery heating film is judged to be faulty and a warning is issued; although the fixed threshold method can achieve power battery heating film fault warning, the warning threshold of this method needs to be manually set by professionals, which has extremely high requirements on the experience of professionals and the accuracy of setting the threshold, and the characteristics of the monitoring indicators are different, which also makes the setting of the threshold difficult. Moreover, when the fixed warning threshold is set to a large value, it is only meaningful for the peak of the flow, while other periods are out of control; when the fixed warning threshold is set to a small value, it cannot meet the peak state warning, and the peak flow is in the warning state for a long time, which loses the meaning of the alarm, and the accuracy of fault judgment is low, and timely warning cannot be issued. Summary of the invention

[0004] The object of the present invention is to provide a method and system for detecting a power battery heating film fault, so as to solve the problem of low accuracy of the existing power battery fixed threshold warning method.

[0005] In order to solve the above technical problems, the present invention provides a method for detecting a power battery heating film fault, the method comprising:

[0006] Obtain the temperature rise rate of the power battery in the historical heating stage, input the temperature rise rate of the power battery in the historical heating stage into the trained time series prediction model, and output the upper and lower bounds of the predicted value of the temperature rise rate of the power battery at the prediction time;

[0007] The actual value of the temperature rise rate of the power battery at the prediction time is obtained. If the actual value of the temperature rise rate is not between the upper and lower limits of the corresponding temperature rise rate prediction value and the difference between the actual value of the temperature rise rate and the upper and lower limits of the corresponding temperature rise rate prediction value is greater than the set difference threshold, it is determined that the power battery heating film is faulty.

[0008] Furthermore, the difference is the degree of deviation between the actual value of the temperature rise rate and the upper and lower limits of the corresponding predicted value of the temperature rise rate; the degree of deviation is calculated as follows:

[0009]

[0010] Among them, when the actual value of the temperature rise rate at the prediction time is higher than the upper limit of the corresponding temperature rise rate prediction value, the baseline value is the upper limit of the temperature rise rate prediction value; when the actual value of the temperature rise rate at the prediction time is lower than the lower limit of the corresponding temperature rise rate prediction value, the baseline value is the lower limit of the temperature rise rate prediction value;

[0011] The set difference threshold is a set deviation degree threshold.

[0012] Furthermore, the temperature rise rate in the historical heating stage includes the highest temperature temperature rise rate and the lowest temperature temperature rise rate; the upper and lower bounds of the predicted value of the temperature rise rate at the prediction moment output by the time series prediction model include the upper and lower bounds of the predicted value of the highest temperature temperature rise rate at the prediction moment and the upper and lower bounds of the predicted value of the lowest temperature temperature rise rate;

[0013] Correspondingly, the conditions for determining that the power battery heating film has a fault are as follows: the actual value of the highest temperature temperature rise rate at the prediction moment is not between the upper and lower limits of the corresponding highest temperature temperature rise rate prediction value, and the difference between the actual value of the highest temperature temperature rise rate and the upper and lower limits of the corresponding highest temperature temperature rise rate prediction value is greater than the set difference threshold, or the actual value of the lowest temperature temperature rise rate at the prediction moment is not between the upper and lower limits of the corresponding lowest temperature temperature rise rate prediction value, and the difference between the actual value of the lowest temperature temperature rise rate and the upper and lower limits of the corresponding lowest temperature temperature rise rate prediction value is greater than the set difference threshold.

[0014] Furthermore, the set difference threshold includes a first set difference threshold and a second set difference threshold; if the difference is greater than the first set difference threshold and less than or equal to the second set difference threshold, it is determined that a serious fault occurs in the power battery heating film; if the difference is greater than the second set difference threshold, it is determined that a catastrophic level fault occurs in the power battery heating film; the second set difference threshold is greater than the first set difference threshold.

[0015] Furthermore, the time series prediction model is a prophet model.

[0016] Furthermore, the calculation formula of the maximum temperature rise rate is:

[0017]

[0018] Among them, Ratio1 is the maximum temperature rise rate, Temp start1 Temp is the maximum temperature of the monomer when heating is turned on. start1 It is the maximum temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

[0019] Furthermore, the calculation formula of the minimum temperature rise rate is:

[0020]

[0021] Among them, Ratio2 is the minimum temperature rise rate, Temp start2 Temp is the lowest temperature of the monomer when heating is turned on. end2 It is the lowest temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

[0022] In order to solve the above technical problems, the present invention also provides a power battery heating film fault detection system, comprising a processor, wherein the processor executes a computer program to implement the steps of the power battery heating film fault detection method as described above.

[0023] The beneficial effects are as follows: the present invention is an improved invention creation. The present invention is no longer like the prior art, which only determines whether the power battery heating film has a fault according to the fixed temperature rise rate threshold set by professional and technical personnel, but takes into account the working state and flow of the power battery heating film and adopts a dynamic interval for determination; by obtaining the temperature rise rate of the power battery in the historical heating stage, the temperature rise rate of the power battery in the historical heating stage is input into the trained time series prediction model, and the upper and lower bounds of the temperature rise rate prediction value of the power battery at the prediction time are output; then the actual temperature rise rate value of the power battery at the prediction time is obtained, if the actual temperature rise rate value is not between the upper and lower bounds of the corresponding temperature rise rate prediction value and the difference between the actual temperature rise rate value and the corresponding temperature rise rate prediction value is greater than the set difference threshold, it is determined that the power battery heating film has a fault; by using the upper and lower bounds of the temperature rise rate prediction value such as the dynamic interval determination method, it is determined whether the power battery heating film has a fault, and the interval range of the prediction value corresponding to different times is different, which is more in line with the performance of the power battery, improves the accuracy of the power battery heating film fault determination, and reduces the risk of power battery fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flowchart of a method for detecting a power battery heating film fault according to an embodiment of the method of the present invention;

[0025] Figure 2 is the maximum temperature temperature rise rate prediction data baseline of the method embodiment of the present invention;

[0026] Figure 3 It is the minimum temperature temperature rise rate prediction data baseline of the method embodiment of the present invention. DETAILED DESCRIPTION

[0027] The present invention obtains the temperature rise rate of a power battery in a historical heating stage, inputs the temperature rise rate of the power battery in the historical heating stage into a trained time series prediction model, and outputs the upper and lower bounds of the predicted value of the temperature rise rate of the power battery at the prediction moment; obtains the actual value of the temperature rise rate of the power battery at the prediction moment, and if the actual value of the temperature rise rate is not between the upper and lower bounds of the corresponding predicted value of the temperature rise rate and the difference between the actual value of the temperature rise rate and the corresponding predicted value of the temperature rise rate is greater than a set difference threshold, it is determined that a heating film of the power battery is faulty; and the problem of low accuracy of the existing fixed threshold warning method for power batteries is solved.

[0028] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0029] Method Example:

[0030] A power battery heating film fault detection method of this embodiment is as follows: Figure 1 As shown, the specific steps include:

[0031] Step 1: Extract the parameter values ​​of non-liquid heating vehicles when the heating relay and parking heating are turned on, specifically:

[0032] During the heating phase and the vehicle is currently in charging state, if the two pieces of data are within 300 seconds and the heating-on status of the two pieces of data is consistent, the heating parameters are extracted to exclude data with a long time interval and errors.

[0033] The extracted parameters include: maximum single cell maximum temperature, maximum temperature single cell battery number, maximum single cell minimum temperature, minimum temperature single cell battery number, single cell starting maximum temperature, single cell starting minimum temperature, single cell ending maximum temperature and single cell ending minimum temperature; among them, the temperature values ​​corresponding to the minimum current of the battery cell ≤0 and the total battery current <-5A in the collected parameters are deleted, and the collected parameters such as the maximum single cell maximum temperature, the maximum temperature single cell battery number, the maximum single cell minimum temperature and the minimum temperature single cell battery number can be used to locate the faulty battery cell.

[0034] Step 2: According to the heating parameters obtained in step 1, calculate the maximum temperature rise rate and the minimum temperature rise rate during the heating of the power battery heating film, so as to construct a historical data baseline. The calculation formula is as follows:

[0035]

[0036] Among them, Ratio1 is the maximum temperature rise rate, Temp start1 Temp is the maximum temperature of the monomer when heating is turned on. start1 It is the maximum temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

[0037]

[0038] Among them, Ratio2 is the minimum temperature rise rate, Temp start2 Temp is the lowest temperature of the monomer when heating is turned on. end2 It is the lowest temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

[0039] Step 3: Based on the historical data baseline obtained in step 2, train and test the constructed time series prediction model.

[0040] The time series prediction model of this embodiment is a prophet model. As other implementation methods, they can be selected according to actual needs, such as a seasonal autoregressive moving average model, a differential autoregressive moving average model, etc. Among them, the prediction output result of this embodiment using the prophet model includes two prediction intervals, namely, the prediction interval of the maximum temperature rise rate and the prediction interval of the minimum temperature rise rate.

[0041] Compared with traditional time series forecasting models, the prophet model has better flexibility, can easily adapt to the seasonality of multiple seasons, and can make different assumptions about trends through analysis. The measured values ​​do not need to be equally spaced, and there is no need to interpolate missing values, so the fitting speed is fast.

[0042] The prophet model has easily interpretable parameters that can be analyzed to impose on the forecast. The prophet model is essentially a model for predicting time series data based on a self-additive model. The model as a whole can be decomposed into three parts: trend term (growth), seasonal term (seasonality) and holiday term (holidays), which can be specifically expressed as:

[0043] p(t)=g(t)+s(t)+h(t)+ε t (3)

[0044] Among them, ε t It is an error term. Generally, it is assumed to obey a normal distribution with a mean of 0. It is mainly used to reflect abnormal changes that are not reflected in the model.

[0045] g(t) is a trend function, including parameters for adjusting the smoothness of the model and assumptions of varying degrees. It is mainly used to model non-periodic changes such as piecewise linear growth or logistic saturation growth in time series, and to detect trend trends based on change points in the data. The logistic saturation growth model is used, and a logistic regression model is used for the basic trend term:

[0046]

[0047] Among them, K is the growth rate, b(t) is the offset, and C(t) is the model capacity; g(t) approaches C(t) as t increases, and g(t) is positively correlated with C(t).

[0048] s(t) is a periodic change. In order to fit and predict the effect of seasons, Prophet proposed a flexible model based on Fourier series and estimated it according to the following formula:

[0049]

[0050] Where N is the total number of cycles, T is a fixed cycle (T for annual data is 365.25, T for weekly data is 7), 2n is the number of cycles expected to be used in the model, a1, a2...a n , b 1, b2...b n are the parameters that need to be estimated in the model.

[0051] h(t) is the holiday effect. Holidays and major events will cause predictable fluctuations in the time series. Independent models can be established for the impact of different holidays and major events at different times. A time window is set for each independent model, which takes into account the impact of holidays in the window period (such as the days before and after the Mid-Autumn Festival). The impact in the same window period is the same value; it is estimated according to the following formula:

[0052]

[0053] Where L is the holiday set, i is the i-th holiday, is the impact factor of the corresponding holiday on the predicted value, and is the time t included in the window period. Define Z(t)=[l(t∈D1),...l(t∈D L )], we can get:

[0054] h(t)=Z(t)k,k∈Normal(0,v2 ) (7)

[0055] Step 4: Using the prophet model trained in step 3, obtain the prediction data baseline of the maximum temperature rise rate prediction value and the minimum temperature rise rate prediction value, which is composed of the prediction value intervals at each moment.

[0056] Step 5: Obtain the actual value of the highest temperature rise rate and the actual value of the lowest temperature rise rate at the predicted moment, and compare them with the predicted value obtained in step 4 respectively. If the difference between the actual value and the predicted value is greater than the set difference threshold, it is determined that the power battery heating film is faulty.

[0057] According to the predicted data baseline of the highest temperature temperature rise rate and the predicted data baseline of the lowest temperature temperature rise rate predicted by the prophet model in step 4, the degree of deviation of the actual values ​​of the temperature rise rates at the predicted moments of the highest temperature temperature rise rate and the lowest temperature temperature rise rate from the upper and lower bounds of the corresponding predicted data baseline (the values ​​are the abnormal points) is calculated respectively. If the calculated degree of deviation meets any condition of the warning rule, a fault warning is triggered.

[0058] The degree of deviation is calculated as:

[0059] Among them, the baseline value is the value in the predicted data baseline, and the actual value is the value obtained by measurement and calculation corresponding to the baseline value; when the abnormal point is higher than the upper limit of the predicted data baseline, the baseline value is correspondingly taken as the upper limit of the predicted data baseline, and when the abnormal point is lower than the lower limit of the predicted data baseline, the baseline value is correspondingly taken as the lower limit of the predicted data baseline. As other implementation methods, the value of the baseline value can be set according to actual needs, for example, the baseline value is set to take the predicted data baseline value as the upper limit of the predicted value, the degree of deviation between each abnormal point and the upper limit is calculated, and the difference threshold is set accordingly.

[0060] The early warning rules are: when the deviation degree of the highest temperature temperature rise rate abnormal point is greater than the first set difference threshold and is less than or equal to the second set difference threshold, it is determined that a serious fault has occurred in the power battery heating film; when the deviation degree of the highest temperature temperature rise rate abnormal point is greater than the first set difference threshold, it is determined that a catastrophic level fault has occurred in the power battery heating film; when the deviation degree of the lowest temperature temperature rise rate abnormal point is greater than the first set difference threshold and is less than or equal to the second set difference threshold, it is determined that a serious fault has occurred in the power battery heating film; when the deviation degree of the lowest temperature temperature rise rate abnormal point is greater than the second set difference threshold, it is determined that a catastrophic level fault has occurred in the power battery heating film; the first set difference threshold is less than the second set difference threshold.

[0061] In this embodiment, the first set difference threshold is 70%, and the second set threshold is 80%. As other implementation methods, they can be set according to actual needs.

[0062] The model is trained based on data from September 2022 to September 2023 to obtain a historical baseline, and then verified based on the data from September 2023 to the present. Figure 2 , Figure 3 As shown, the prediction is made using the power battery heating film fault detection method of this embodiment. The prediction points between the upper and lower limits of the prediction baseline in each figure are normal, and the points outside the upper and lower limits are abnormal points, and the degree of deviation needs to be calculated.

[0063] Step 5: Based on the fault warning results, the problematic vehicle will be accurately sent to designated personnel or professional technicians for further inspection and repair.

[0064] The present invention provides a method for detecting a power battery heating film fault, which can monitor various parameters of the power battery heating film during operation in real time and improve the accuracy of battery fault warning. At the same time, the reliability of power battery heating film fault judgment is effectively improved by combining the heating film system parameters and battery status monitoring data. When a fault is identified, a warning signal can be issued in time, and corresponding measures can be taken to reduce the risk of power battery failure caused by the failure of the heating film. The present invention has high real-time performance and accuracy, helps to improve the safety and reliability of the battery, and reduce the risk of failure. It has broad application prospects in the fields of new energy vehicles, energy storage systems, etc.

[0065] System Example:

[0066] A power battery heating film fault detection system of the present invention includes a memory, a processor and an internal bus. The processor and the memory communicate and exchange data with each other through the internal bus. The memory includes at least one software function module stored in the memory. The processor executes various functional applications and data processing by running the computer program and module stored in the memory to implement a power battery heating film fault detection method introduced in the method embodiment of the present invention. The principle, implementation process and achievable effect of the method have been fully introduced in the method embodiment and will not be repeated here.

Claims

1. A method for detecting a power battery heating film fault, characterized in that: The method includes: Obtain the temperature rise rate of the power battery in the historical heating stage, input the temperature rise rate of the power battery in the historical heating stage into the trained time series prediction model, and output the upper and lower bounds of the predicted value of the temperature rise rate of the power battery at the prediction time; The actual value of the temperature rise rate of the power battery at the prediction time is obtained. If the actual value of the temperature rise rate is not between the upper and lower limits of the corresponding temperature rise rate prediction value and the difference between the actual value of the temperature rise rate and the upper and lower limits of the corresponding temperature rise rate prediction value is greater than the set difference threshold, it is determined that the power battery heating film is faulty.

2. The power battery heating film fault detection method according to claim 1, characterized in that: The difference is the degree of deviation between the actual value of the temperature rise rate and the upper and lower limits of the corresponding predicted value of the temperature rise rate; the degree of deviation is calculated as follows: Among them, when the actual value of the temperature rise rate at the prediction time is higher than the upper limit of the corresponding temperature rise rate prediction value, the baseline value is the upper limit of the temperature rise rate prediction value; when the actual value of the temperature rise rate at the prediction time is lower than the lower limit of the corresponding temperature rise rate prediction value, the baseline value is the lower limit of the temperature rise rate prediction value; The set difference threshold is a set deviation degree threshold.

3. The power battery heating film fault detection method according to claim 1, characterized in that: The temperature rise rate in the historical heating stage includes the highest temperature temperature rise rate and the lowest temperature temperature rise rate; the upper and lower bounds of the predicted value of the temperature rise rate at the prediction moment output by the time series prediction model include the upper and lower bounds of the predicted value of the highest temperature temperature rise rate at the prediction moment and the upper and lower bounds of the predicted value of the lowest temperature temperature rise rate; Correspondingly, the conditions for determining that the power battery heating film has a fault are as follows: the actual value of the highest temperature temperature rise rate at the prediction moment is not between the upper and lower limits of the corresponding highest temperature temperature rise rate prediction value, and the difference between the actual value of the highest temperature temperature rise rate and the upper and lower limits of the corresponding highest temperature temperature rise rate prediction value is greater than the set difference threshold, or the actual value of the lowest temperature temperature rise rate at the prediction moment is not between the upper and lower limits of the corresponding lowest temperature temperature rise rate prediction value, and the difference between the actual value of the lowest temperature temperature rise rate and the upper and lower limits of the corresponding lowest temperature temperature rise rate prediction value is greater than the set difference threshold.

4. The power battery heating film fault detection method according to claim 1, characterized in that: The set difference threshold includes a first set difference threshold and a second set difference threshold; if the difference is greater than the first set difference threshold and less than or equal to the second set difference threshold, it is determined that a serious fault occurs in the power battery heating film; if the difference is greater than the second set difference threshold, it is determined that a catastrophic level fault occurs in the power battery heating film; the second set difference threshold is greater than the first set difference threshold.

5. The power battery heating film fault detection method according to claim 1, characterized in that: The time series prediction model is a prophet model.

6. The power battery heating film fault detection method according to claim 3, characterized in that: The calculation formula of the maximum temperature rise rate is: Among them, Ratio1 is the maximum temperature rise rate, Temp start1 Temp is the maximum temperature of the monomer when heating is turned on. start1 It is the maximum temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

7. The power battery heating film fault detection method according to claim 3, characterized in that: The calculation formula of the minimum temperature rise rate is: Among them, Ratio2 is the minimum temperature rise rate, Temp start2 Temp is the lowest temperature of the monomer when heating is turned on. end2 It is the lowest temperature of the monomer at the end of heating, and DuringTime is the duration of heating.

8. A power battery heating film fault detection system, comprising a processor, characterized in that: The processor executes a computer program to implement the steps of the power battery heating film fault detection method according to any one of claims 1 to 7.

Images