A power battery insulation early warning detection system based on big data

Abstract

The application discloses a kind of power battery insulation early warning detection systems based on big data, belong to vehicle battery early warning technical field, according to different state of vehicle, set resistance, test insulation resistance information entropy, cumulative information entropy, weighted information entropy, and monthly iteration calculation is carried out to vehicle, through iteration calculation result is entered next time calculation, again according to the absolute value of Z score and the comparison of host factory or supplier according to abnormal function self-defined value, whether vehicle is finally judged to alarm, the possibility that the application assesses whole vehicle insulation problem by the calculation of cloud data, can be distinguished by multidimension under high pressure parking, go high pressure driving, parked charging and other multi-working conditions, maximum degree excludes external introduction circuit or data detection / upload abnormality caused insulation false alarm, utilizes cloud big data to early warn whole vehicle insulation risk, passenger safety is guaranteed, reduce the harassment and security anxiety of customer due to false alarm, improve product quality.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle battery early warning, and specifically relates to a big data-based power battery insulation early warning detection system. Background Art

[0002] With the continuous popularization of new energy vehicles, an increasing amount of vehicle data and alarm information is uploaded to national or enterprise platforms. Among them, the vehicle high-voltage insulation alarm accounts for a relatively large proportion in all alarm statistics, and most of them are false alarms caused by the introduction of external circuits. At present, most vehicles implement different strategies for insulation alarms under charging and discharging conditions, but limited by the real-time performance of vehicle data, it is still impossible to identify whether an insulation problem actually exists. Therefore, the present invention combines the big data of the cloud platform, and performs insulation fault early warning judgment in combination with the relevant data of the whole life cycle of the vehicle through the judgment method described in this patent. Summary of the Invention

[0003] In view of the above problems, the present invention provides a big data-based power battery insulation early warning detection system. According to different vehicle states, resistances are set, and the information entropy, cumulative information entropy and weighted information entropy of the insulation resistance are tested. Iterative calculation is performed on the vehicle every month, and the iterative calculation result is included in the next calculation. Then the absolute value of the Z-score is compared with a value defined by the automaker or supplier according to the abnormality function to finally determine whether the vehicle gives an alarm. Through the calculation of cloud data, the present invention evaluates the possibility of insulation problems in the whole vehicle, can distinguish multiple working conditions such as high-voltage off parking, high-voltage on driving, and parking charging through multi-dimensional classification, maximally eliminate insulation false alarms caused by external introduced circuits or abnormal data detection / upload, use cloud big data to give early warning of whole vehicle insulation risks in advance, ensure passenger safety, reduce customer harassment and safety anxiety caused by false alarms, and improve product quality.

[0004] The technical solution of the present invention is as follows: a big data-based power battery insulation early warning detection system, comprising the following steps:

[0005] The insulation early warning module of the platform judges the vehicle state through signals uploaded by the vehicle; the states include a high-voltage off state, a high-voltage on state, and a charging state;

[0006] Eliminate non-alarm data when the high-voltage insulation fault flag is set or the high-voltage insulation detection circuit fault flag is set;

[0007] When the insulation resistance R≥R1, calculate as R=R1; when R1>R≥R2, calculate as the actual value of R; when R<R2, calculate as R=R2;

[0008] S4. Calculate the probability p of each insulation resistance value R i occurring in the sample iwhere i is the number of samples;

[0009] S5. Randomly select 800 vehicles on the network each month. The selected vehicles must upload more than 48 hours of data in the current month. For each vehicle, extract 1-10 hours of valid insulation resistance data.

[0010] S6. Using the formula:

[0011] E ANew =-∑p i *log2(p i )

[0012] Calculate E ANew E ANew This represents the entropy of the insulation resistance information for the current month.

[0013] S7. Using the formula:

[0014]

[0015] Calculate E A E A Let E be the cumulative information entropy, n be the iteration number, and E be the cumulative information entropy. AOld The cumulative information entropy from the previous month is iterated monthly.

[0016] S8. Record the iteration count n:

[0017] n = n + 1

[0018] n = Min[n, 11]

[0019] Record E A Execute S9, output E A Execute S10;

[0020] S9, Monthly record of cumulative information entropy E A And in the information entropy correction next month, it will be recorded as E in S7. AOld ;

[0021] S10, Output the cumulative information entropy E for the current month. A Up to S18;

[0022] S11. Record the insulation resistance data for the day, using the formula:

[0023] E0=-∑p i *log2(p i )

[0024] Calculate the information entropy E0 of the insulation resistance under various states between 0:00:00 and 23:59:59;

[0025] S12, Using the formula:

[0026] E=0.5*θ*E0+0.3*(2-θ)*E1+0.2*(2-θ)*E2

[0027] Calculate the weighted information entropy E of the vehicle on the current day, where E1 is the information entropy of the previous day and E2 is the information entropy of the previous two days.

[0028] S13. Output the weighted information entropy E of the vehicle for the day, and store the vehicle information entropy E. o Execute S14, then perform Z-score calculation and execute S15;

[0029] S14. Record and store the E0 value for more than 2 days. Use it as E1 in the entropy calculation of the next day and E2 in the entropy calculation of the next two days. Use it to perform the calculation of the next S15.

[0030] S15. Through the formula:

[0031] |Z|=(EE A ) / σ

[0032] Calculate the absolute value of the Z-score, where σ is the standard deviation;

[0033] S16. Determine if |Z|≤X, where X is defined by the OEM or supplier based on the exception function. If the result is yes, proceed to S17; if the result is no, proceed to S18.

[0034] S17. Output insulation warning judgment result: Vehicle is normal, process ends;

[0035] S18. Record the vehicle's VIN code in the insulation warning key vehicle storage area;

[0036] S19. Determine if |Z|≥Y, where Y is defined by the OEM or supplier based on the exception function. If the result is negative, proceed to step 20; if the result is positive, proceed to step 24.

[0037] S20. Determine if the vehicle is in a parked charging state. If yes, proceed to S21; otherwise, proceed to S22.

[0038] S21. If the insulation detection value is determined to be low due to the introduction of an external charger and the insulation value does not reach the enterprise's predetermined warning level, only the insulation warning will be introduced to focus on the vehicle warehouse, and the process will be terminated.

[0039] S22. Determine if the vehicle is recorded as "verification complete". If yes, proceed to S21; otherwise, proceed to S23.

[0040] S23. Re-sample and record the vehicle VIN, mark the vehicle as "verification completed", clear this status every day, and execute S11.

[0041] S24. Based on the vehicle status, if the vehicle is in a low-voltage state, execute S25; if the vehicle is in a high-voltage state, execute S29; if the vehicle is in a parked charging state, execute S31.

[0042] S25. Determine whether the vehicle has recorded an insulation resistance detection circuit fault within the past 30 days. If yes, proceed to S26; if no, proceed to S27.

[0043] S26. The insulation resistance detection circuit of the vehicle is deemed unreliable. Only the power battery pack insulation detection fault is recorded, and no warning is given. The process is terminated.

[0044] S27. Record early warning of insulation fault in power battery pack;

[0045] S28, Insulation warning triggered, process terminated;

[0046] S29. Determine whether the vehicle has triggered the high-voltage insulation warning within 3 days. If yes, proceed to S25; if no, proceed to S30.

[0047] S30, record the early warning of insulation fault in the high-voltage system of the whole vehicle, and execute S28;

[0048] S31. Crawl data for 10 minutes before and after the vehicle's parking and charging phase on the same day;

[0049] S32. Determine whether "R > R1 remains for more than 30 seconds within 60 seconds before the charging state changes from "not charging" to "charging while parked" or "not charging" to "abnormal" or "R > R1 remains for more than 30 seconds within 120 seconds after the charging state changes from "charging while parked" to "charging complete" or "charging while parked" to "not charging" or "charging while parked" to "abnormal" or "abnormal" to "not charging". If neither condition is met, proceed to S33; if either condition is met, proceed to S34.

[0050] S33. Record the vehicle charging insulation fault warning and execute S28;

[0051] S34. Determine whether the vehicle has been recorded as a key vehicle for monitoring within the past 30 days. If yes, proceed to S33; if no, proceed to S35.

[0052] S35. It is believed that the low insulation detection value is caused by the introduction of an external charger, and there is no risk of low insulation resistance in the vehicle's high-voltage system, resulting in a false alarm.

[0053] S36. Record the warning information, but no action is required; the process terminates.

[0054] Furthermore, in step S1, the lower high voltage state indicates that the high voltage of the vehicle is not connected; the upper high voltage state indicates that the high voltage of the vehicle is connected when the vehicle is not stopped for charging; and the charging state indicates that the vehicle is in the state of plugging in the charging gun while parked.

[0055] Furthermore, in step S3, R1 is calculated by the battery system based on AΩ / V, where A is the threshold set by the OEM, V is the voltage unit, Ω is the resistance unit, and R2 is the threshold for the battery system to trigger an insulation alarm. Non-discrete data is calculated using R = R2, and discrete data is discarded.

[0056] Furthermore, in step S3, AΩ / V is 1000Ω / V.

[0057] Furthermore, in step S7, n is an integer and 0≤n≤11, and n=0 is calculated for the first time.

[0058] Furthermore, in step S12, it is detected whether the vehicle VIN is a key vehicle of concern. If so, a correction factor θ = B is set, which can be set by the OEM or supplier as 2 ≥ B ≥ 1; if not, a correction factor θ = 1 is set.

[0059] Furthermore, in step S12, θ = B, and B is 1.2.

[0060] Furthermore, in step S16, X is 3.5.

[0061] Furthermore, in step S19, Y is 4.

[0062] The beneficial effects of this invention are as follows:

[0063] This invention uses cloud data calculations to assess the likelihood of insulation problems in the entire vehicle. It can differentiate between various operating conditions such as parking under high voltage, driving under high voltage, and charging while parked, minimizing false alarms caused by external circuits or abnormal data detection / upload. By utilizing cloud big data to provide early warnings of vehicle insulation risks, it ensures passenger safety, reduces customer annoyance and safety anxiety caused by false alarms, and improves product quality. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of a big data insulation testing system.

[0065] Figure 2 This is a schematic diagram of the vehicle-end insulation detection circuit.

[0066] Figure 3 This is a schematic diagram of the insulation fault early warning judgment logic.

[0067] In the picture:

[0068] Figure 2 The red line represents the positive electrode circuit of the battery system, and the black line represents the negative electrode circuit of the battery system. Detailed Implementation

[0069] It should be noted that in the description of this invention, the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", "clockwise", "counterclockwise", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation.

[0070] In this invention, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link," etc., should be interpreted broadly. For example, "fixed" can mean a fixed connection, a detachable connection, or an integral part; a connection can be a mechanical connection or an electrical connection; a link can be a direct connection or an indirect connection through an intermediate medium, and can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0071] This invention requires the vehicle to upload relevant data to the platform, including: VIN code, vehicle status, charging status, insulation resistance value, high-voltage insulation fault flag, and high-voltage insulation detection circuit fault flag. The platform cleans the uploaded data, including filling in missing data and correcting discrete data, and then stores the cleaned data in its database. Except for the VIN code, all other data should be uploaded at a frequency of ≤5 seconds.

[0072] Step 1: Insulation warning. The system will determine the following three operating conditions: low high voltage, high high voltage, and charging. Low high voltage indicates that the vehicle's high voltage is not connected; high high voltage indicates that the vehicle's high voltage is connected when charging is not stopped; and charging indicates that the vehicle is in a plug-in charging state while parked. The signals uploaded to the platform must differentiate between these states.

[0073] Step 2: All data involved in the calculation are non-alarm data excluding those under the high-voltage insulation fault flag position or the high-voltage insulation detection circuit fault flag position.

[0074] Step 3

[0075] When the insulation resistance R ≥ R1, it is calculated as R = R1.

[0076] When R1 > R ≥ R2, calculate based on the value of R itself.

[0077] When R < R2, calculate according to R = R2.

[0078] R1 is calculated by the battery system based on AΩ / V, where A is the threshold set by the OEM, V is the voltage unit, Ω is the resistance unit, and R2 is the threshold for the battery system to trigger an insulation alarm. Because the vehicle-side insulation resistance detection time and alarm judgment cycle are relatively long, there are cases where the uploaded resistance value is lower than R2 in some frames but the high-voltage insulation fault flag is not set. Non-discrete data is calculated according to R=R2, and discrete data is discarded.

[0079] A is the threshold value, and Ω / V is the unit. In battery systems, because each battery has a different rated voltage, the insulation resistance requirements for the battery system also differ.

[0080] For example, if the battery design rated voltage of car model A is 400V and the battery design voltage of car model B is 350V, then according to the national standard requirement that the insulation resistance alarm threshold is ≥100Ω / V, the insulation resistance of battery A shall not be less than 100Ω / V*400V=40kΩ, and the insulation resistance of battery B shall not be less than 100Ω / V*350V=35kΩ.

[0081] Therefore, the insulation warning threshold AΩ / V mentioned in this article is a standard requirement, and the specific threshold resistors R1 and R2 are obtained from the voltage of system A.

[0082] Step 4: Calculate the insulation resistance value R for each insulation resistance. i The probability p that occurs in the sample i where i is the number of samples;

[0083] Commonly used mathematical calculation methods, p i =Event frequency / Total number of samples.

[0084] Step 5: For the low-voltage state, randomly select 800 vehicles on the network each month. The selected vehicles are required to upload more than 48 hours of data in the current month. The vehicle status for each vehicle is equal to 1 hour of effective insulation resistance data under the low-voltage state.

[0085] For high voltage conditions, 800 vehicles on the network are randomly selected each month. The selected vehicles are required to upload more than 48 hours of data in the current month. The vehicle status for each vehicle selected is equal to 10 hours of effective insulation resistance data under high voltage conditions.

[0086] For vehicles in the parking and charging state, 800 vehicles in the network are randomly selected each month. The selected vehicles are required to upload more than 48 hours of data in the current month. The vehicle status sampled for each vehicle is equal to 10 hours of effective insulation resistance data in the parking and charging state.

[0087] Step 6: Calculate the information entropy E of the vehicle's insulation resistance for the current month under the vehicle's state (lower high voltage state / upper high voltage state / parking and charging state). ANew The information entropy of each vehicle model is calculated separately;

[0088] E ANew =-∑p i *log2(p i )

[0089] Step 7: Monthly calculation of cumulative information entropy E A The process involves iteration. The initial accumulated information entropy is provided by the OEM or supplier. The number of iterations is recorded as n. The initial calculation is n = 0, where n is an integer and 0 ≤ n ≤ 11.

[0090] Step 8: Calculate the cumulative information entropy E A And monthly, the cumulative information entropy E A The iteration process begins, with the initial accumulated information entropy provided by the OEM supplier, E. AOld This is the cumulative information entropy from the previous iteration;

[0091] (n represents E) A The number of iterations, n is an integer where 0 ≤ n ≤ 11, and the first calculation is n = 0.

[0092] Step 9: Record the number of iterations n:

[0093] n = n + 1

[0094] n = Min[n, 11]

[0095] Record E A Execute step 10 and output E. A Perform step 11;

[0096] Step 10: Record the monthly information entropy E A And in the information entropy correction of the following month, it will be recorded as the cumulative information entropy E in step 8. AOld =E A ;

[0097] Step 11: Output the information entropy E for the current month. A Proceed to step 19;

[0098] Step 12: Record the insulation resistance data for each vehicle on the same day, and calculate the information entropy E0 of the insulation resistance of the vehicle in the three states from 0:00:00 to 23:59:59 on the same day: low high voltage state / high high voltage state / parking and charging state. Calculate the three states separately.

[0099] E0=-∑p i *log2(p i )

[0100] Step 13: Check if the vehicle's VIN is a vehicle of high interest. If yes, proceed to step 14; otherwise, proceed to step 15.

[0101] Step 14: Set the correction factor θ = B, which can be set by the OEM or supplier: 2 ≥ B ≥ 1.

[0102] Step 15: Set the correction coefficient θ = 1;

[0103] Step 16: Combining the vehicle's information entropy over the past three days, the current information entropy is weighted to obtain the weighted information entropy E for the current day, where E1 is the information entropy of the previous day and E2 is the information entropy of the previous two days (Note: The weighted information entropy E in this step is only used to calculate the insulation resistance warning situation for the current day, and E1 and E2 in the formula are still the information entropy calculated according to step 12 the previous day).

[0104] E=0.5*θ*E0+0.3*(2-θ)*E1+0.2*(2-θ)*E2

[0105] Step 17: Output the weighted information entropy E of the vehicle for the day, and store the vehicle information entropy E. o Execute step 18, calculate the Z-score, and then execute step 19.

[0106] Step 18: Record and store the E0 value for more than 2 days. This value will be used as E1 in the entropy calculation of the next day and E2 in the entropy calculation of the next two days. This value will be used to perform the next calculation in Step 16.

[0107] Step 19: Calculate the absolute value of the Z-score as the anomaly coefficient for judgment;

[0108] |Z|=(EE A ) / σ

[0109] Where σ is the standard deviation.

[0110] Step 20: Determine if |Z|≤X (X is defined by the OEM or supplier based on the exception function). If the result is yes, proceed to step 21; if the result is no, proceed to step 22.

[0111] Step 21: Output insulation warning judgment result: Vehicle is normal, process ends;

[0112] Step 22: Record the vehicle's VIN code in the insulation warning key vehicle storage area;

[0113] Step 23: Determine if |Z|≥Y (Y is defined by the OEM or supplier based on the exception function). If the result is negative, proceed to step 24; if the result is positive, proceed to step 28.

[0114] Step 24: Determine if the vehicle is in a parked charging state. If yes, proceed to step 25; if no, proceed to step 26.

[0115] Step 25: If the insulation detection value is low due to the introduction of an external charger and the insulation value does not reach the enterprise's predetermined warning level, only the insulation warning will be introduced to focus on the vehicle warehouse, and the process will be terminated.

[0116] Step 26: Determine if the vehicle is recorded as "verification complete". If yes, proceed to step 25; if no, proceed to step 27.

[0117] Step 27: Re-sample and record the vehicle VIN, mark the vehicle as "verification completed", clear this status daily, and execute step 12;

[0118] Step 28: Determine the vehicle status. If the vehicle status is in the lower high voltage state, proceed to step 29. If the vehicle status is in the upper high voltage state, proceed to step 33. If the vehicle status is in the parking charging state, proceed to step 35.

[0119] Step 29: Determine if the vehicle has recorded any insulation resistance detection circuit faults within the past 30 days. If yes, proceed to step 30; if no, proceed to step 31.

[0120] Step 30: If the insulation resistance detection circuit of the vehicle is deemed unreliable, only record the power battery pack insulation detection fault, do not take any warning action, and terminate the process.

[0121] Step 31: Record the early warning of insulation fault in the power battery pack;

[0122] Step 32: Insulation warning is triggered, and the process terminates;

[0123] Step 33: Determine whether the vehicle has triggered a high-voltage insulation warning within the past 3 days. If yes, proceed to step 29; if no, proceed to step 34.

[0124] Step 34: Record the early warning of insulation fault in the high-voltage system of the whole vehicle, and then proceed to step 32;

[0125] Step 35: Crawl the vehicle's charging status for the day = data for 10 minutes before and after the parking and charging phase;

[0126] Step 36: Determine if "R > R(1) for more than 30 seconds within 60 seconds before the charging state changes from "not charging → charging while parked" or "not charging → abnormal" or "R > R(1) for more than 30 seconds after the charging state changes from "charging while parked → charging completed" or "charging while parked → not charging" or "charging while parked → abnormal" or "abnormal → not charging". If no, proceed to step 37; if yes, proceed to step 38.

[0127] Step 37: Record the vehicle charging insulation fault warning and proceed to step 32;

[0128] Step 38: Determine whether the vehicle has been recorded as a key monitoring vehicle within the past 30 days (excluding the day of the warning). If yes, proceed to step 37; if no, proceed to step 39.

[0129] Step 39: It is assumed that the low insulation detection value is caused by the introduction of an external charger, and there is no risk of low insulation resistance fault in the vehicle's high-voltage system, resulting in a false alarm.

[0130] Step 40: Record the warning information, but no further action is required; the process terminates.

[0131] Based on the above steps, online vehicle insulation fault early warning is completed using big data. The early warning results can be pushed to the R&D and after-sales departments through the platform to support OEMs in data analysis and after-sales maintenance.

[0132] In step 3, AΩ / V is set to 1000Ω / V. 100Ω / V is the first-level insulation alarm threshold required by national standards, and 500Ω / V is the second-level insulation alarm threshold commonly used in the industry. Since the power battery system directly triggers the insulation alarm when it is below 500Ω / V, in order to provide early warning for the present invention, and in combination with the fault performance of past models, 1000Ω / V is set as the threshold for early warning judgment.

[0133] In step 14, θ = B, and B is set to 1.2. This is mainly because after one Z-score judgment, vehicles with low risk of insulation warning are screened out. To avoid the result error caused by sample specificity, the sample is collected again and the insulation information entropy is corrected to increase the proportion of the information entropy on the day. The focus is on assessing the insulation status on the day. θ = B = 1.2 is obtained based on historical data.

[0134] In step 20, X is set to 3.5. According to the 3σ principle, it is considered that when |Z|≤3.5, the insulation information entropy is consistent with the cumulative information entropy, and there is no risk of insulation failure.

[0135] In step 23, Y is set to 4. According to the 3σ principle, it is believed that when |Z|≥4, the insulation information entropy deviates from the cumulative information entropy, indicating a significant risk of insulation failure.

[0136] This invention combines big data information entropy to calculate the cumulative information entropy of vehicle models and iterates it monthly. At the same time, it corrects the daily information entropy of a single vehicle within three days and performs repeated verification correction according to a strategy.

[0137] This invention can distinguish the vehicle state according to the strategy, break down the working conditions for different states, and further refine the judgment strategy.

[0138] This invention can eliminate insulation abnormalities caused by changes in the insulation circuit due to the introduction of an external charger, based on a strategy.

[0139] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be included within the scope of protection of the present invention. Furthermore, all content not described in detail in this specification is prior art known to those skilled in the art.

[0140] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

Claims

1. A power battery insulation early warning detection system based on big data, characterized in that, Includes the following steps: The insulation warning module of the S1 platform judges the vehicle status by the signals uploaded by the vehicle; the status includes low high voltage status, high high voltage status, and charging status. S2 removes non-alarm data at the high-voltage insulation fault mark position or the high-voltage insulation detection circuit fault mark position; S3 When the insulation resistance R≥R1, it is calculated as R=R1; when R1>R≥R2, it is calculated as R itself; when R<R2, it is calculated as R=R2. R1 is calculated by the battery system based on AΩ / V, where A is the threshold set by the OEM, V is the voltage unit, and Ω is the resistance unit. R2 is the threshold for the battery system to trigger the insulation alarm. Non-discrete data is calculated as R=R2, and discrete data is discarded. S4. Calculate the insulation resistance value R for each insulation resistance. i The probability P that occurs in the sample i where i is the number of samples; S5. Randomly select 800 vehicles on the network each month. The selected vehicles must upload more than 48 hours of data in the current month. For each vehicle, extract 1-10 hours of valid insulation resistance data. S6. Using the formula: Calculate E ANew E ANew This represents the entropy of the insulation resistance information for the current month. S7. Using the formula: Calculate E A E A Let E be the cumulative information entropy, n be the iteration number, and E be the cumulative information entropy. AOld The cumulative information entropy from the previous month is used for monthly iteration; S8. Record the iteration count n: n=n+1 n = Min[n, 11] Record E A Execute S9, output E A Execute S10; S9, Monthly record of cumulative information entropy E A And in the information entropy correction next month, it will be recorded as E in S7. AOld ; S10, Output the cumulative information entropy E for the current month. A Up to S18; S11. Record the insulation resistance data for the day, using the formula: Calculate the information entropy E0 of the insulation resistance under various states between 0:00:00 and 23:59:59; S12, Using the formula: Calculate the weighted information entropy E of the vehicle for the day, where E1 is the information entropy of the previous day, E2 is the information entropy of the two days prior to the current day, and θ is the correction coefficient. Detect whether the vehicle VIN is a key vehicle to be monitored. If so, set the correction coefficient θ=B, which is set by the OEM or supplier: 2≥B≥1. If not, set the correction coefficient θ=1. S13. Output the weighted information entropy E of the vehicle for the day, and store the vehicle information entropy E. o Execute S14, then perform Z-score calculation and execute S15; S14. Record and store the E0 value for more than 2 days. Use it as E1 in the entropy calculation of the next day and E2 in the entropy calculation of the next two days. Use it to perform the calculation of the next S15. S15. Through the formula: Calculate the absolute value of the Z-score, where σ is the standard deviation; S16. Determine if |Z| ≤ X, where X is defined by the OEM or supplier based on the exception function. If the result is yes, proceed to S17; if the result is no, proceed to S18. S17. Output insulation warning judgment result: Vehicle is normal, process ends; S18. Record the vehicle's VIN code in the insulation warning key vehicle storage area; S19. Determine if |Z|≥Y, where Y is defined by the OEM or supplier based on the exception function. If the result is negative, proceed to step 20. If the result is yes, proceed to step 24; S20. Determine if the vehicle is in a parking charging state. If yes, execute S21. If not, proceed to S22; S21. If the insulation detection value is determined to be low due to the introduction of an external charger and the insulation value does not reach the enterprise's predetermined warning level, only the insulation warning will be introduced to focus on the vehicle warehouse, and the process will be terminated. S22. Determine whether the vehicle is recorded as "verification completed". If yes, proceed to S21; otherwise, proceed to S23. S23. Re-sample and record the vehicle VIN, mark the vehicle as "verification completed", clear this status daily, and execute S11; S24. Based on the vehicle status, if the vehicle is in a low-voltage state, execute S25; if the vehicle is in a high-voltage state, execute S29; if the vehicle is in a parked charging state, execute S31. S25. Determine whether the vehicle has recorded an insulation resistance detection circuit fault within the past 30 days. If yes, proceed to S26; if no, proceed to S27. S26. The insulation resistance detection circuit of the vehicle is deemed unreliable. Only the insulation detection fault of the power battery pack is recorded, and no warning is given. The process is terminated. S27. Record early warning of insulation fault in power battery pack; S28, Insulation warning triggered, process terminated; S29. Determine whether the vehicle has triggered the high-voltage insulation warning within 3 days. If yes, proceed to S25; if no, proceed to S30. S30, record the early warning of insulation fault in the high-voltage system of the whole vehicle, and execute S28; S31. Crawl data for 10 minutes before and after the vehicle's parking and charging phase on the same day; S32. Determine whether "R > R1 remains greater than 30 seconds for more than 30 seconds before the charging status changes from "not charging" to "charging while parked" or "not charging" to "abnormal"; or whether "R > R1 remains greater than 30 seconds for more than 30 seconds after the charging status changes from "charging while parked" to "charging complete" or "charging while parked" to "not charging" or "charging while parked" to "abnormal" or "abnormal" to "not charging". If none of these conditions are met, proceed to S33; if one of these conditions is met, proceed to S34. S33. Record the vehicle charging insulation fault warning and execute S28; S34. Determine whether the vehicle has been recorded as a key vehicle for monitoring within 30 days. If yes, proceed with S33. If not, proceed to S35; S35. It is believed that the low insulation detection value is caused by the introduction of an external charger, and there is no risk of low insulation resistance in the vehicle's high-voltage system, resulting in a false alarm. S36. Record the warning information, but no action is required; the process terminates.

2. The power battery insulation early warning detection system based on big data as described in claim 1, characterized in that, In step S1, the lower high voltage state indicates that the high voltage of the vehicle is not connected; the upper high voltage state indicates that the high voltage of the vehicle is connected when the vehicle is not parked for charging; and the charging state indicates that the vehicle is in the state of plugging in the charging gun while parked for charging.

3. The power battery insulation early warning detection system based on big data as described in claim 1, characterized in that, In step S3, AΩ / V is 1000Ω / V.

4. The power battery insulation early warning detection system based on big data as described in claim 1, characterized in that, In step S7, n is an integer and 0≤n≤11, and n=0 is calculated for the first time.

5. The power battery insulation early warning detection system based on big data as described in claim 4, characterized in that, In step S12, θ = B, and B is 1.

2.

6. The power battery insulation early warning detection system based on big data as described in claim 1, characterized in that, In step S16, X is 3.

5.

7. The power battery insulation early warning detection system based on big data as described in claim 1, characterized in that, In step S19, Y is 4.

Images