Quantitative diagnosis method for fault of battery voltage sensor

Abstract

The invention discloses a quantitative diagnosis method for a fault of a battery voltage sensor. The method comprises the following steps: collecting time t, a highest monomer voltage Vmax and a lowest monomer voltage Vmin in a certain time period; four parameters, namely dVmax, dVmin, Vjump and detaV, are calculated; inspecting a fault starting time point t1 of the voltage sensor; inspecting a fault end time point t2 of the voltage sensor; the voltage abnormity degree Vabavg is calculated; calculating a voltage abnormal time ratio tab; and quantitatively estimating the fault degree of the voltage sensor. The method can be used for accurately judging the safety risk level of cloud diagnosis and accurately guiding the after-sales maintenance scheme, the algorithm universality is extremely high, a battery digital model is not needed, calculation is simple, the quantitative result can be used for accurately judging the looseness degree of the voltage sensor, the blank problem in the field of existing technical schemes can be filled, and the method is suitable for popularization and application. And online application in a vehicle end battery management system can also be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of battery application, and more specifically to a quantitative diagnosis method for battery voltage sensor failure. Background Art

[0002] Battery fault diagnosis can be achieved by using advanced technical means to monitor and analyze the vehicle or equipment without stopping or disassembling it, to determine whether the vehicle or equipment has abnormalities or faults, and the diagnosis mainly relies on battery voltage, current, temperature and other parameters. However, when there are problems with the voltage, current and temperature sensors themselves, abnormal data will be generated, which will cause abnormal judgment results and misjudgments such as overvoltage, undervoltage, and fire alarms. In particular, when the voltage sensor is loose, it is easy to cause an external short circuit, which can cause a fire in the battery system in severe cases.

[0003] At present, the battery management system can avoid false alarms by jointly judging multiple monitoring parameters or extending the judgment time. However, when there is an abnormality in the voltage sensor and the abnormal voltage value remains within the normal voltage range, it is difficult for the BMS to judge the fault of the voltage sensor. In addition, the abnormal voltage value can easily lead to misjudgment of the cloud battery diagnosis results, resulting in erroneous warnings and increased offline after-sales investigation costs.

[0004] Existing battery fault diagnosis methods, such as a method for diagnosing the fault of an intermediate voltage sensor of a locomotive traction converter disclosed in Chinese patent publication number CN105137376A, a vehicle power battery fault detection system and detection method disclosed in Chinese patent publication number CN102033205A, and a method, system, device and medium for diagnosing the fault of an energy storage system sensor disclosed in Chinese patent publication number CN118033519A. Some of these diagnostic methods require the use of battery models, which increases the algorithm's requirements for hardware computing power; some require the preparation of data sets and label training, which increases the workload; some require an increase in the number of voltage sensors, which increases the cost. Summary of the invention

[0005] The present invention provides a quantitative diagnosis method for battery voltage sensor faults, aiming to solve the problem that voltage sensor faults cannot be identified in the prior art, accurately provide the cause of vehicle battery system fault alarms, improve after-sales troubleshooting efficiency, and avoid misjudgment of battery problems by battery cloud diagnosis, thereby reducing risk warning false alarms and reducing after-sales troubleshooting costs.

[0006] The present invention adopts the following technical solution:

[0007] A quantitative diagnosis method for battery voltage sensor failure comprises the following steps:

[0008] S1. Collect time and voltage data within a certain period of time. The voltage data includes the highest single cell voltage V max and the minimum cell voltage V min ;

[0009] S2. Calculate dV according to formulas (1)-(4) max , dV min 、V jump , detaV four parameters;

[0010] dV max (i) = V max (i)-V max (i-1) (1)

[0011] dV min (i) = V min (i)-V min (i-1) (2)

[0012] V jump (i) = dV max (i)-dV min (i) (3)

[0013] detaV=V max (i)-V min (i) (4)

[0014] Among them, dV max V max Change in value over time, dV min V min Changes over time, V jump Represents dV max With dV min The difference between V max With V min The difference between , i is the time series number;

[0015] S3, inspection voltage sensor fault start time t1;

[0016] S4, inspection voltage sensor fault end time point t2;

[0017] S5. Calculate the voltage abnormality V according to formula (5) ab_avg :

[0018]

[0019] Where base(k) is the detaV value corresponding to the voltage data before time t1(k) when there is no fault, and k is the fault number sequence number;

[0020] S6. Calculate the voltage abnormal time ratio t according to formula (6) ab :

[0021] S7: The voltage abnormality level V calculated in step S5 ab_avg The voltage abnormality time ratio t calculated in step S6 ab , quantitatively estimate the fault degree of voltage sensor.

[0022] In the above step S2, dV is calculated. max and dV min When calculating dV max and dV min When the absolute value is less than 10mV, it is directly set to 0.

[0023] The inspection method at the voltage sensor fault start time t1 in step S3 is as follows: max The value is greater than b and dV min is less than -b, and V jump It must be greater than the threshold a; the time point at which the above conditions are met during the inspection is one or more.

[0024] The inspection method at the voltage sensor failure end time t2 in step S4 is as follows: max The value is judged to be less than -b and dV min is greater than b, and V jump The value must be less than the threshold value - a. The time point at which the above conditions are met during the inspection is one or more.

[0025] The above threshold a is set in the range of 0.02-0.05, and can be adjusted accordingly according to different application scenarios; the threshold b is set in the range of 0.00-0.01, and can be adjusted accordingly according to different application scenarios.

[0026] The base(k) in the above step S5 directly takes the detaV value at the moment before t1(k); or takes the average value of the detaV value in the time period from t2(k-1)+1 to t1(k)-1 or the value with the largest probability density distribution; or takes the average value of all detaV arrays that are not in the time period corresponding to the start time and end time of the fault or the value with the largest probability density distribution.

[0027] The V in step S5 above ab_avg It can also be calculated using the following formula: V ab_avg =max(V ab ), and select different formulas for calculation according to different application scenarios.

[0028] The voltage sensor fault degree in step S7 can be estimated by fuzzy mathematics method. ab_avg ,t ab The fault level is determined by adding, taking the maximum value or taking the minimum value. The V ab_avg ,t ab The final fault level is confirmed based on the two classification results. The confirmation method is to add, take the maximum and minimum values. You can also directly use V when different evaluation effects are required in different scenarios. ab_avg , t ab One or more direct evaluations.

[0029] The present invention is applicable to the identification of battery voltage sensor faults in all battery systems, and there is no need to distinguish between lithium iron phosphate batteries, ternary batteries or lithium titanate batteries, etc. when used; it can also be applied to online diagnosis of battery management systems and cloud diagnosis.

[0030] It can be seen from the above description of the present invention that, compared with the prior art, the present invention has the following advantages:

[0031] The present invention designs a calculation method for identifying the degree of voltage sensor failure by utilizing the principle that all single cells have the same electrochemical polarization direction, and utilizes the coordinated judgment of four parameters to accurately identify the start and end time of the voltage sensor failure, thereby supporting the accurate calculation of the voltage sensor voltage abnormality amplitude and voltage abnormality time ratio, and realizing the accurate grading of the voltage sensor failure degree, especially the sensor loose failure, which is beneficial to the cloud diagnosis for the diagnosis of battery abnormality, and can be used to accurately determine the safety risk level of cloud diagnosis and to accurately guide after-sales maintenance plans. The algorithm is highly versatile, does not require the use of a battery digital model, is simple to calculate, and the quantitative results can be used to accurately determine the degree of voltage sensor looseness, which can fill the gaps in the field of existing technical solutions, and can also be realized in online applications in vehicle-side battery management systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 Schematic diagram of a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0033] Refer to the following Figure 1 The specific implementation of the present invention is described. In order to fully understand the present invention, many details are described below, but for those skilled in the art, the present invention can be implemented without these details. For well-known components, methods and processes, they are not described in detail below.

[0034] This embodiment provides a quantitative diagnosis method for battery voltage sensor failure. Figure 1, including the following steps:

[0035] S1. Collect the time t and voltage data within a certain period of time. The voltage data includes the highest single cell voltage V max and the minimum cell voltage V min .

[0036] S2. Calculate dV according to formulas (1)-(4) max , dV min 、V jump , detaV four parameters.

[0037] dV max (i) = V max (i)-V max (i-1) (1)

[0038] dV min (i) = V min (i)-V min (i-1) (2)

[0039] V jump (i) = dV max (i)-dV min (i) (3)

[0040] detaV=V max (i)-V min (i) (4)

[0041] Among them, dV max V max Change in value over time, dV min V min Changes over time, V jump Represents dV max With dV min The difference between V max With V min , i is the time series number.

[0042] In step S2, dV is calculated. max and dV min When calculating dV max and dV min When the absolute value is less than 10mV, it is directly set to 0.

[0043] S3, inspect the voltage sensor fault start time t1. The inspection method at t1 is: based on dV max The value is greater than b and dV minis less than -b, and V jump It must be greater than the threshold a; the time point at which the above conditions are met during the inspection is one or more.

[0044] The threshold a is set in the range of 0.02-0.05 and can be adjusted accordingly according to different application scenarios.

[0045] S4, inspect the voltage sensor fault end time t2; the inspection method for t2 is: according to dV max The value is judged to be less than -b and dV min is greater than b, and V jump The value must be less than the threshold value - a. The time point at which the above conditions are met during the inspection is one or more.

[0046] The threshold b is set in the range of 0.00-0.01 and can be adjusted accordingly according to different application scenarios.

[0047] S5. Calculate the voltage abnormality V according to formula (5) ab_avg :

[0048]

[0049] Where base(k) is the detaV value corresponding to the voltage data before time t1(k) when there is no fault, and k is the fault number sequence number;

[0050] The above base(k) can directly take the detaV value before t1(k); it can also take the average value of the detaV value in the time period from t2(k-1)+1 to t1(k)-1 or the value with the largest probability density distribution; it can also be the average value of all detaV arrays that are not in the time period corresponding to the start time and end time of the fault or the value with the largest probability density distribution.

[0051] S6. Calculate the voltage abnormal time ratio t according to formula (6) ab :

[0052] S7: The voltage abnormality level V calculated in step S5 ab_avg The voltage abnormality time ratio t calculated in step S6 ab , quantitatively estimate the fault degree of voltage sensor.

[0053] The voltage sensor fault degree in step S7 can be estimated by fuzzy mathematics method. ab_avg ,t ab After classification, expert knowledge is used to determine the fault level, and the confirmation method is addition, maximum value or minimum value.

[0054] It should be noted that the V ab_avg It can also be calculated using the following formula: V ab_avg =max(V ab ), and select different formulas for calculation according to different application scenarios.

[0055] It should be noted that the fault degree of the voltage sensor in step S7 can also be determined by a threshold value. ab_avg ,t ab The final fault level is confirmed based on the two classification results. The confirmation method is to add, take the maximum and minimum values. You can also directly use V when different evaluation effects are required in different scenarios. ab_avg , t ab One or more direct evaluations.

[0056] The batteries involved above can be new energy vehicle battery systems or battery energy storage systems.

[0057] The quantitative diagnosis method of the present invention is applicable to the identification of battery voltage sensor faults in all battery systems, and there is no need to distinguish between lithium iron phosphate batteries, ternary batteries or lithium titanate batteries, etc. when using it; it can be applicable to online diagnosis of battery management systems and also to cloud diagnosis.

[0058] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited thereto. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.

Claims

1. A quantitative diagnosis method for battery voltage sensor failure, characterized in that: The following steps are involved: S1. Collect time and voltage data within a certain period of time. The voltage data includes the highest single cell voltage V max and the minimum cell voltage V min ; S2. Calculate dV according to formulas (1)-(4) max , dV min 、V jump , detaV four parameters; dV max (i)=V max (i)-V max (i-1) (1) dV min (i)=V min (i)-V min (i-1) (2) V jump (i)=dV max (i)-dV min (i) (3) detaV=V max (i)-V min (i) (4) Among them, dV max V max Change in value over time, dV min V min Changes over time, V jump Represents dV max With dV min The difference between V max With V min The difference between , i is the time series number; S3, inspection voltage sensor fault start time t1; S4, inspection voltage sensor fault end time point t2; S5. Calculate the voltage abnormality V according to formula (5) ab_avg : Where base(k) is the detaV value corresponding to the voltage data before time t1(k) when there is no fault, and k is the fault number sequence number; S6. Calculate the voltage abnormal time ratio t according to formula (6) ab : S7: The voltage abnormality level V calculated in step S5 ab_avg The voltage abnormality time ratio t calculated in step S6 ab , quantitatively estimate the fault degree of voltage sensor.

2. A quantitative diagnosis method for battery voltage sensor failure according to claim 1, characterized in that: If the voltage sensor error is 10mV, the step S2 calculates dV max and dV min When the absolute value is less than 10mV, it is directly set to 0.

3. The quantitative diagnosis method of a battery voltage sensor fault according to claim 1, characterized in that: The inspection method of the voltage sensor failure start time t1 in step S3 is: based on dV max The value is greater than b and dV min is less than -b, and V jump Must be greater than threshold a; The inspection reaches one or more time points that meet the above conditions.

4. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 3, characterized in that: The inspection method of the voltage sensor failure end time point t2 in step S4 is: based on dV max The value is judged to be less than -b and dV min is greater than b, and V jump The value must be less than the threshold value - a. The time point at which the above conditions are met during the inspection is one or more.

5. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 4, characterized in that: The threshold a is set within the range of 0.02-0.05; the threshold b is set within the range of 0.00-0.

01.

6. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 1, characterized in that: The base(k) of step S5 directly takes the detaV value at the moment before t1(k); or takes the average value of the detaV value in the time period from t2(k-1)+1 to t1(k)-1 or the value with the largest probability density distribution; or takes the average value of all detaV arrays that are not in the time period corresponding to the start time and end time of the fault or the value with the largest probability density distribution.

7. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 1, characterized in that: The V of step S5 ab_avg It can also be calculated using the following formula: V ab_avg =max(V ab ).

8. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 1, characterized in that: The voltage sensor fault degree in step S7 is calculated by fuzzy mathematics method. ab_avg ,t ab After classification, expert knowledge is used to determine the fault level, and the confirmation method is addition, maximum value or minimum value.

9. A quantitative diagnosis method for battery voltage sensor failure as claimed in claim 1, characterized in that: The voltage sensor fault degree of step S7 is determined by threshold value. ab_avg ,t ab The two classification results are used to determine the final fault level by adding, taking the maximum and minimum values.

10. A quantitative diagnosis method for battery voltage sensor failure according to claim 1, characterized in that: The voltage sensor fault degree in step S7 is directly used when different evaluation effects are required in different scenarios. ab_avg , t ab One or more direct evaluations.

Images