A method for automatically obtaining the threshold value of capacitor withstand voltage insulation test

Abstract

The present invention discloses a method for automatically acquiring a capacitor withstand voltage insulation test threshold, comprising: S1. Sample collection: using a discrete TV / IR data sample set CpData obtained by equipment testing, the discrete TV / IR data samples are filtered step by step to obtain a new sample set; S2. Calculating the TV / IR threshold: obtaining the average value Eqv and standard value Std of the discrete TV / IR data samples by the average value and standard deviation square root calculation formula, and then quickly calculating the TV / IR threshold by using the UpDn threshold algorithm and combining the HalfSeak dichotomy method; S3. Calculating the number of capacitors removed from the upper and lower parts; S4. Acquiring the ideal threshold. The beneficial effects of the present invention: improving the accuracy of the step-by-step screening of capacitors, replacing the manual point counting method, and greatly improving the efficiency of TV / IR threshold calculation.

Description

Technical Field

[0001] The invention belongs to the technical field of industrial production control systems, and in particular relates to a method for automatically acquiring a capacitor withstand voltage insulation test threshold. Background Art

[0002] In capacitor testing and sorting, the withstand voltage insulation TV / IR test is an essential parameter. Different manufacturers have their own process standards, but in the process of formulating the standards, the thresholds of each TV / IR station need to be determined. In many cases, it takes a lot of time to formulate a capacitor test process standard, and it is necessary to repeatedly adjust the upper and lower limits to find the standard process parameters and carry out the work processing.

[0003] There are still some problems with the existing methods for automatically acquiring the thresholds for capacitor withstand voltage insulation tests: during the on-site machine adjustment process, it is found that each batch of products of the user process will be tested on the machine in advance to obtain the product process parameters. The most time-consuming part is the acquisition of the thresholds of each TV / IR station, which often cannot be determined for several hours, thereby reducing work efficiency and increasing machine time occupancy. At the same time, the accuracy of the obtained thresholds is low. For this reason, we propose a method for automatically acquiring the thresholds for capacitor withstand voltage insulation tests. Summary of the invention

[0004] The purpose of the present invention is to provide a method for automatically acquiring a capacitor withstand voltage insulation test threshold value to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for automatically obtaining a capacitor withstand voltage insulation test threshold, comprising the following steps:

[0006] S1. Sample collection: using discrete TV / IR data samples obtained by equipment testing, the discrete TV / IR data samples are filtered step by step to obtain a new sample set;

[0007] S2. Calculate TV / IR threshold: obtain the average value Eqv and standard value Std of the discrete TV / IR data samples by the mean value and standard deviation square root calculation formula, and then quickly calculate the TV / IR threshold by the UpDn threshold algorithm combined with the dichotomy method;

[0008] S3. Calculate the number of capacitors to be rejected: Use the UpDn threshold algorithm to evaluate the quality of the capacitors and the rationality of the rejection ratio proposed by the process, give the upper and lower limits of the rejection data and mark them intuitively in the histogram, and calculate the number of capacitors to be rejected by the given rejection ratio Koff and the total number of samples;

[0009] S4. Ideal threshold acquisition: Through human-computer interaction, semi-automatic upper and lower limit adjustment is achieved to obtain the ideal threshold.

[0010] Preferably, the calculation of the TV / IR threshold in S2 is specifically as follows: Based on the multi-station TV / IR data set of the sampling capacitor, since some devices do not distinguish between withstand voltage and insulation, they are collectively referred to as IR1, IR2..IRn. Calculate the average value Eqv(i) and standard value Std(i) of TV / IR(i) at each station, and according to the given rejection ratio Koff(i) at each station, obtain a new sample set through step-by-step filtering, and then obtain the threshold [Dn(i), Up(i)] of TV / IR at each station according to the UpDn and HalfSeak algorithms. Where: i is the serial number of the withstand voltage and insulation TV / IR test site, Up(i) and Dn(i) are the IRi thresholds [Dn(i), Up(i)] of the capacitor TV / IR leakage current at the i-th station. The rejection ratio Koff(i) is set according to the user's process requirements. The rejection ratio includes the upper limit rejection ratio and the lower limit rejection ratio. When not given separately, it is evenly distributed, that is, each accounts for 50%, corresponding to Koff / 2 in the following calculation formula.

[0011] Preferably, the specific steps for obtaining the threshold [Dn(i), Up(i)] of TV / IR at each station include:

[0012] S2011. Sampling data: IR leakage currents IR1, IR2..IRn corresponding to Cp;

[0013] S2012. Calculate Eqv(i), STD(i) of the leakage current data of IR1, IR2..IRn and the leakage current distribution histogram (i = 1 to n);

[0014] S2013. According to the center value Eqv(1), square root of standard deviation Std(1), and rejection ratio of the capacitor IR1 leakage current, combined with the external capacitor distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the qualified threshold [Dn1, Up1] of IR1;

[0015] S2014. If IR1 ∈ [Dn1, Up1], reject the unqualified samples of IR1, reorganize the sampling data, Select *from CpData where (IR1> Up1 and IR1 <Dn1), and obtain the IR2 sample set;

[0016] S2015. According to the center value Eqv(2), square root of standard deviation Std(2), and rejection ratio of the capacitor IR2 leakage current, combined with the external capacitor distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the qualified threshold [Dn2, Up2] of IR2;

[0017] S2016.IR1∈[Dn1,Up1], IR2∈[Dn2,Up2],remove the samples with unqualified IR1 and IR2, and reorganize the sampling data Select * from CpData where (IR1>Up1 andIR1<Dn1) and (IR2> Up2 and IR2< Dn2), get the IR3 sample set;

[0018] S2017. Based on the central value Eqv(3) of the leakage current of capacitor IR3, the standard deviation square root Std(3), and the elimination ratio, and combined with the external capacitance distribution [Eqv-Std, Eqv+Std], the UpDn&HalfSeak threshold algorithm is called to obtain the IR3 qualified threshold [Dn3, Up3];

[0019] S2018. IR1∈[Dn1,Up1], IR2∈[Dn2,Up2],IR3∈[Dn3,Up3],remove the samples with unqualified IR1..IR3 and reorganize the sampling data Select * from CpData where (IR1>Up1 and IR1<Dn1)and (IR2> Up2 and IR2< Dn2) and (IR3>Up3 and IR3< Dn3), get the IR4 sample set; and so on, get the IRn sample set;

[0020] S2019. Based on the central value Eqv(n) of the leakage current of capacitor IRn, the root standard deviation Std(n), and the elimination ratio, and combined with the external capacitance distribution [Eqv-Std, Eqv+Std], the UpDn&HalfSeak threshold algorithm is called to obtain the IRn qualified threshold [Dnn, Upn].

[0021] Preferably, the formula for the average value Eqv(i) of the discrete TV / IR data samples is:

[0022] ;

[0023] Where x1(i) is the sample set of TV / IR and N is the total number of stations in the insulation TV / IR test.

[0024] Preferably, the formula for the standard deviation square mean (referred to as standard value) Std(i) of the discrete TV / IR data samples is:

[0025] .

[0026] Preferably, the UpDn threshold algorithm in S2 specifically includes the following steps:

[0027] S2021. Given: leakage current data set x(i), mean value Eqv, standard deviation root Std, elimination ratio Koff, total number of valid data Nt, normal distribution data, etc.; threshold range [Eqv-k1*Std, Eqv+ k2*Std];

[0028] S2022. Determine whether (Count(X(i))>Equ+Std)> Nt*Koff / 2) and(Count(X(i)> Equ+2*Std)< Nt*Koff / 2). If not, proceed to S2023. If yes, proceed to S2024.

[0029] S2023. The data distribution is rather chaotic, modify the test parameters, re-sample, and end;

[0030] S2024. Determine whether (Count(X(i))<Equ-Std)> Nt*Koff / 2)and (Count(X(i)< Equ-2*Std)< Nt*Koff / 2), if not, go to S2025, if yes, go to S2026;

[0031] S2025. The data distribution is relatively scattered, modify the test parameters, re-sample, and end;

[0032] S2026. The upper threshold is between (Eqv+ Std, Eqv+2*Std) and the lower threshold is between (Eqv-2* Std, Eqv-Std). The threshold is in a reasonable range;

[0033] S2027. Obtain the upper threshold value between (Eqv+ Std, Eqv+ 2*Std) through the normal distribution graph, and use the binary search algorithm HalfSeak(x, a, b), where a=1, b=2;

[0034] S2028. Obtain the lower limit of the threshold between (Eqv-2*Std, Eqv-Std) using the normal distribution data, and use the binary search algorithm HalfSeak(x, a, b), a=-2, b=-1, and end.

[0035] Preferably, the specific steps of adopting the dichotomy method are:

[0036] S2031. Take capacitance c1=a, c2=b;

[0037] S2032. Let c = (c1 + c2) / 2;

[0038] S2033. Determine whether (Count(X(i)> Equ+c*Std)> Nt*Koff / 2). If so, c1=c, return to S2032. If not, enter S2034.

[0039] S2034. Determine whether (Count(X(i)> Equ+c*Std)= Nt*Koff / 2), if not, c2=c, return to S2032, if yes, end the process.

[0040] Preferably, the elimination work in S3 adopts the principle of sequential elimination of multi-station TV / IR tests, and screens the data level by level to filter out samples that have been eliminated by the previous TV / IR level.

[0041] Preferably, the TV / IR threshold is related to the capacity, test voltage and test time of the capacitor, and different test equipments have different test positions, movement times and charging times, so upper and lower limits need to be adjusted.

[0042] Preferably, the upper and lower limit adjustment is to achieve upper and lower limit setting under fewer sample conditions to achieve a fast calibration function, and the fewer sample conditions are 300-1000 capacitors.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] (1) The present invention uses the TV / IR test data set of capacitor withstand voltage insulation, the TV / IR normal distribution characteristics of batch capacitors, the histogram, the average value, the root square value of the standard deviation and the rejection ratio provided by the user, and the UpDn+HalfSeak algorithm to quickly and automatically calculate the threshold, thereby improving the accuracy of the step-by-step screening of capacitors, increasing work efficiency and reducing machine time occupancy;

[0045] (2) The method for automatically acquiring the threshold value of a capacitor withstand voltage insulation test in the present invention is applicable to the automatic acquisition of the capacitor leakage current and insulation resistance threshold values, and is also applicable to the automatic acquisition of the pure insulation test threshold value. In addition, it can be extended to TVIR tests of more stations, and the capacitor test TV / IR upper and lower limit analysis algorithm is also applicable to the automatic acquisition of threshold values ​​in the test sorting of resistors, inductors and other products, thereby having a wider range of application. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a flowchart of the average value and standard value calculation, step-by-step filtering, and UpDn+HalfSeak algorithm call of the present invention.

[0047] Figure 2 This is a flowchart of the UpDn algorithm of the present invention.

[0048] Figure 3It is a flowchart of the HalfSeak algorithm of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] See also Figure 1-Figure 3 The present invention provides a technical solution: a method for automatically obtaining a capacitor withstand voltage insulation test threshold, comprising the following steps:

[0051] S1. Sample collection: using discrete TV / IR data samples obtained by equipment testing, the discrete TV / IR data samples are filtered step by step to obtain a new sample set;

[0052] S2. Calculate TV / IR threshold: obtain the average value Eqv and standard value Std of the discrete TV / IR data samples by the mean value and standard deviation square root calculation formula, and then quickly calculate the TV / IR threshold by the UpDn threshold algorithm combined with the dichotomy method;

[0053] S3. Calculate the number of capacitors to be rejected: Use the UpDn threshold algorithm to evaluate the quality of the capacitors and the rationality of the rejection ratio proposed by the process, give the upper and lower limits of the rejection data and mark them intuitively in the histogram, and calculate the number of capacitors to be rejected by the given rejection ratio Koff and the total number of samples;

[0054] S4. Ideal threshold acquisition: Through human-computer interaction, semi-automatic upper and lower limit adjustment is achieved to obtain the ideal threshold.

[0055] In this embodiment, preferably, the S2 calculates the TV / IR threshold specifically as follows: based on the TV / IR data set of multiple stations of the sampled capacitor, the average value Eqv(i) and the standard value Std(i) of TV / IRi of each station are calculated, and according to the given rejection ratio Koff(i) of each station, a new sample set is obtained by step-by-step filtering, and then the TV / IR threshold of each station [Dn(i), Up(i)] is obtained according to the UpDn and HalfSeak algorithms, where i is the total number of stations in the insulation TV / IR test, and Up(i) and Dn(i) are the TV / IR thresholds of the capacitor at the i-th station [Dn(i), Up(i)].

[0056] In this embodiment, preferably, the specific steps of obtaining the TV / IR threshold [Dn(i), Up(i)] of each station include:

[0057] S2011. Sampling data: IR leakage currents IR1, IR2..IRn corresponding to Cp;

[0058] S2012. Calculate Eqv(i), STD(i) of the leakage current data of IR1, IR2..IRn and the leakage current distribution histogram (i = 1 to n), taking n = 4 as an example;

[0059] S2013. According to the center value Eqv(1), root mean square standard deviation Std(1), rejection ratio of the IR leakage current of capacitor IR1, and combined with the external capacitor distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the qualified threshold [Dn1, Up1] of IR1;

[0060] S2014. If IR1 ∈ [Dn1, Up1], reject the unqualified samples of IR1, reorganize the sampling data, Select *from CpData where IR1> Up1 and IR1 <Dn1, to obtain the IR2 sample set;

[0061] S2015. According to the center value Eqv(2), root mean square standard deviation Std(2), rejection ratio of the IR leakage current of capacitor IR2, and combined with the external capacitor distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the qualified threshold [Dn2, Up2] of IR2;

[0062] S2016. If IR1∈[Dn1, Up1] and IR2∈[Dn2, Up2], reject the unqualified samples of IR1 and IR2, reorganize the sampling data Select * from CpData where (IR1>Up1 and IR1 <Dn1) and (IR2>Up2 and IR2< Dn2), to obtain the IR3 sample set;

[0063] S2017. According to the center value Eqv(3), root mean square standard deviation Std(3), rejection ratio of the IR leakage current of capacitor IR3, and combined with the external capacitor distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the qualified threshold [Dn3, Up3] of IR3;

[0064] S2018. Reject the samples with unqualified IR1..IR3, reorganize the sampling data. Select * from CpData where (IR1>Up1 and IR1 <Dn1) and (IR2>Up2 and IR2< Dn2) and (IR3>Up3 and IR3<Dn3) to obtain the IR4 sample set;

[0065] S2019. According to the center value Eqv(4), root mean square standard deviation Std(4), and rejection ratio of the capacitance IR4 leakage current, combined with the external capacitance distribution of [Eqv - Std, Eqv + Std], call the UpDn&HalfSeak threshold algorithm to obtain the IR4 qualified threshold [Dn4, Up4].

[0066] In this embodiment, the formula for the average value Eqv(i) of the discrete TV / IR data samples is:

[0067] ;

[0068] where x1(i) is the sample set of TV / IR, and N is the total number of stations in the insulation TV / IR test.

[0069] In this embodiment, the formula for the mean square standard deviation (abbreviated as standard value) Std(i) of the discrete TV / IR data samples is:

[0070] .

[0071] In this embodiment, the UpDn threshold algorithm in S2 specifically includes the following steps:

[0072] S2021. Given: leakage current data set X(i), average value Eqv(i), root mean square standard deviation Std(i), rejection ratio Koff(i), total number of valid data Nt, normal distribution data, etc.; for the sake of convenience of description, (i) is omitted, the threshold range is [Eqv - k1*Std, Eqv + k2*Std], where k1 and k2 are constants to be obtained; Count(..) is the capacitance count that meets the.. condition, CpData is the data set of capacitance test parameters, i.e., the database, and Cp is the capacitance;

[0073] S2022. Determine whether (Count(X(i)) > Equ + Std) > Nt*Koff / 2) and (Count(X(i) > Equ + 2*Std) < Nt*Koff / 2). If not, go to S2023; if so, go to S2024;

[0074] S2023. The data distribution is relatively chaotic. Modify the test parameters, take samples again, and end;

[0075] S2024. Determine whether (Count(X(i))<Equ-Std)> Nt*Koff / 2)and (Count[X(i)< Equ-2*Std)< Nt*Koff / 2), if not, go to S2025, if yes, go to S2026;

[0076] S2025. The data distribution is relatively scattered, modify the test parameters, re-sample, and end;

[0077] S2026. The upper threshold is between (Eqv+ Std, Eqv+2*Std) and the lower threshold is between (Eqv-2* Std, Eqv-Std). The threshold is in a reasonable range;

[0078] S2027. Obtain the upper threshold value between (Eqv+ Std, Eqv+ 2*Std) through the normal distribution graph, and use the binary search algorithm HalfSeak(x, a, b), where a=1, b=2;

[0079] S2028. Obtain the lower limit of the threshold between (Eqv-2*Std, Eqv-Std) using the normal distribution data, and use the binary search algorithm HalfSeak(x, a, b), a=-2, b=-1, and end.

[0080] In this embodiment, preferably, the specific steps of adopting the dichotomy method are:

[0081] S2031. Take capacitance c1=a, c2=b;

[0082] S2032. Determine whether c1>=0 and c2>=0, if so, return to S2033, if not, enter S2036;

[0083] S2033. Let c = (c1 + c2) / 2; / / (upper limit)

[0084] S2034. Determine whether (Count(X(i)> Equ+c*Std)> Nt*Koff / 2). If so, c1=c, return to S2033. If not, enter S2035.

[0085] S2035. Determine whether (Count(X(i)> Equ+c*Std)= Nt*Koff / 2). If not, c2=c, return to S2033. If yes, end the process.

[0086] S2036. Let c = (c1 + c2) / 2; / / (lower limit)

[0087] S2037. Determine whether (Count(X(i)< Equ+c*Std)> Nt*Koff / 2). If so, c1=c, return to S2036. If not, enter S2038.

[0088] S2038. Determine whether (Count(X(i)> Equ+c*Std)= Nt*Koff / 2), if not, c2=c, return to S2036, if yes, end the process.

[0089] In this embodiment, preferably, the elimination work in S3 adopts the principle of multi-station TV / IR test sequential elimination, and screens the data level by level to filter out samples that have been eliminated by the previous TV / IR level.

[0090] In this embodiment, preferably, the TV / IR threshold is related to the capacity of the capacitor, the test voltage, and the test time, and different test equipment has different test positions, movement times, and charging times, and upper and lower limits need to be adjusted.

[0091] In this embodiment, preferably, the upper and lower limit adjustment is to achieve upper and lower limit setting under fewer sample conditions to achieve a fast calibration function, and the fewer sample conditions are 300-1000 capacitors.

[0092] Principles and advantages of the present invention:

[0093] The present invention uses a capacitor withstand voltage insulation TV / IR test data set, utilizes the TV / IR normal distribution characteristics of batch capacitors, and uses a histogram, an average value, a standard deviation square root value, and a rejection ratio provided by a user to quickly and automatically calculate a threshold value using an UpDn+HalfSeak algorithm, thereby improving the accuracy of step-by-step screening of capacitors, increasing work efficiency, and reducing machine time occupancy; the capacitor withstand voltage insulation test threshold automatic acquisition method of the present invention is suitable for automatic acquisition of capacitor leakage current and insulation resistance thresholds, and is also suitable for automatic acquisition of pure insulation test thresholds. In addition, it can be expanded to TVIR tests of more stations, and the capacitor test TV / IR upper and lower limit analysis algorithm is also suitable for automatic acquisition of thresholds in test sorting of products such as resistors and inductors, thereby having a higher scope of application.

[0094] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for automatically obtaining a capacitor withstand voltage insulation test threshold, characterized in that: The following steps are involved: S1. Sample collection: using equipment testing to obtain discrete withstand voltage / insulation data samples, and gradually filtering the discrete withstand voltage / insulation data samples to obtain a new sample set; S2. Calculate the withstand voltage / insulation threshold: obtain the average value Eqv and standard value Std of the discrete withstand voltage / insulation data samples through the mean value and standard deviation square root calculation formula, and then quickly calculate the withstand voltage / insulation threshold through the threshold algorithm combined with the binary algorithm; S3. Calculate the number of capacitors to be rejected: Through the threshold algorithm, evaluate the quality of capacitors and the rationality of the rejection ratio proposed by the process, give the upper and lower limits of the rejection data and mark them intuitively in the histogram, and calculate the number of capacitors to be rejected by the given rejection ratio Koff and the total number of samples; S4. Threshold acquisition: Through human-computer interaction, the upper and lower limits are adjusted to obtain the threshold.

2. The method for automatically obtaining a capacitance withstand voltage insulation test threshold according to claim 1, characterized in that: The S2 calculates the withstand voltage / insulation threshold specifically as follows: based on the multi-station withstand voltage / insulation data sets IR1, IR2...IRn of the sampled capacitor, the average value Eqv(i) and the standard value Std(i) of the withstand voltage / insulation of each station are calculated, and according to the given rejection ratio Koff(i) of each station, a new sample set is obtained by step-by-step filtering, and then the withstand voltage / insulation threshold [Dn(i), Up(i)] of each station is obtained according to the threshold algorithm and combined with the binary algorithm, wherein: i is the withstand voltage / insulation test site number, Up(i) and Dn(i) are the withstand voltage / insulation leakage current IRi thresholds of the capacitor at the i-th station [Dn(i), Up(i)], the rejection ratio Koff(i) is set by the user's process requirements, and the rejection ratio includes an upper limit rejection ratio and a lower limit rejection ratio. When not given separately, they are evenly distributed, that is, each accounts for 50%, corresponding to Koff / 2 in the following calculation formula.

3. The method for automatically obtaining a capacitance withstand voltage insulation test threshold according to claim 2, characterized in that: The specific steps of obtaining the withstand voltage / insulation threshold [Dn(i), Up(i)] of each station include: S2011. Sampling data: IR leakage current IR1, IR2…IRn corresponding to capacitor Cp; S2012. Calculate Eqv(i), STD(i) and leakage current distribution histogram of leakage current data of IR1, IR2 ... IRn; S2013. According to the central value Eqv(1), the standard deviation square root Std(1) and the elimination ratio of the capacitor leakage current IR1, and combined with the external capacitor distribution [Eqv-Std, Eqv+Std], the threshold algorithm and the binary algorithm are called to obtain the IR1 qualified threshold [Dn1, Up1]; S2014. IR1 ∈ [Dn1, Up1], remove the unqualified samples of IR1 and obtain the IR2 sample set; S2015. Based on the central value Eqv(2) of the leakage current of capacitor IR2, the standard deviation square root Std(2), and the elimination ratio, and combined with the external capacitance distribution [Eqv-Std, Eqv+Std], the threshold algorithm and the binary algorithm are called to obtain the IR2 qualified threshold [Dn2, Up2]; S2016. IR1 ∈ [Dn1, Up1], IR2 ∈ [Dn2, Up2]. Reject the samples with unqualified IR1 and IR2, and obtain the IR3 sample set; S2017. Based on the center value Eqv(3), the root mean square of the standard deviation Std(3), and the rejection ratio of the capacitance IR3 leakage current, combined with the external capacitance distribution of [Eqv - Std, Eqv + Std], call the threshold algorithm and the bisection method algorithm to obtain the qualified threshold [Dn3, Up3] of IR3; S2018. IR1 ∈ [Dn1, Up1], IR2 ∈ [Dn2, Up2], IR3 ∈ [Dn3, Up3]. Reject the samples with unqualified IR1...IR3, and obtain the IR4 sample set; and so on, to obtain the IRn sample set; S2019. Based on the center value Eqv(n), the root mean square of the standard deviation Std(n), and the rejection ratio of the capacitance IRn leakage current, combined with the external capacitance distribution of [Eqv - Std, Eqv + Std], call the threshold algorithm and the bisection method algorithm to obtain the qualified threshold [Dnn, Upn] of IRn.

4. The method for automatically acquiring a capacitance withstand voltage insulation test threshold according to claim 1, characterized in that: The formula for the average value Eqv(i) of the discrete withstand voltage / insulation data samples is: ； where X(j) is the sample set of IRi, and Nt is the total number of stations in the insulation TV / IR test.

5. The method for automatically acquiring a capacitor withstand voltage insulation test threshold according to claim 4, characterized in that: The formula for the mean square of the standard deviation Std(i) of the discrete withstand voltage / insulation samples is: 。 6. The method for automatically acquiring a capacitor withstand voltage insulation test threshold according to claim 1, characterized in that: The threshold algorithm in S2 specifically includes the following steps: S2021. Given: data set X(j), average value Eqv(i), root mean square of the standard deviation Std(i), rejection ratio Koff(i), total number of stations Nt in the withstand voltage / insulation test, normal distribution data; threshold range [Eqv(i) - k1*Std(i), Eqv(i) + k2*Std(i)], where k1 and k2 are constants to be obtained; Count(...) is the capacitance count that meets the... condition, CpData is the data set of capacitance test parameters, i.e., the database, and Cp is the capacitance; S2022. Determine whether ((Count(X(i)) > Eqv + Std) > Nt*Koff / 2) and (Count(X(i) > Eqv + 2*Std) < Nt*Koff / 2). If not, go to S2023; if so, go to S2024; S2023. Modify the test parameters, re-sample, and end; S2024. Determine whether ((Count(X(i)) < Eqv - Std) > Nt*Koff / 2) and (Count(X(i) < Eqv - 2*Std) < Nt*Koff / 2). If not, go to S2025; if so, go to S2026; S2025. Modify the test parameters, re-sample, and end; S2026. If the upper threshold is between (Eqv + Std, Eqv + 2*Std) and the lower threshold is between (Eqv - 2*Std, Eqv - Std), then the threshold is in a reasonable range; S2027. Obtain the upper threshold value between (Eqv+ Std, Eqv+ 2*Std) through the normal distribution graph, and use the binary search algorithm HalfSeak(x, a, b), where a=1, b=2; S2028. Obtain the lower limit of the threshold between (Eqv-2*Std, Eqv-Std) using the normal distribution data, and use the binary search algorithm HalfSeak(x, a, b), a=-2, b=-1, and end.

7. The method for automatically acquiring a capacitance withstand voltage insulation test threshold according to claim 6, characterized in that: The specific steps of adopting the dichotomy method are: S2031. Take capacitance c1=a, c2=b; S2032. Let c = (c1 + c2) / 2; S2033. Determine whether (Count(X(i)>Eqv+c*Std)>Nt*Koff / 2). If so, c1=c, return to S2032. If not, enter S2034. S2034. Determine whether (Count(X(i)>Eqv+c*Std)= Nt*Koff / 2), if not, c2=c, return to S2032, if yes, end the process.

8. The method for automatically acquiring a capacitor withstand voltage insulation test threshold according to claim 1, characterized in that: The elimination work in S3 adopts the principle of multi-station withstand voltage / insulation test sequential elimination, screens the data level by level, and filters out samples that have been eliminated by the previous level withstand voltage / insulation.

9. The method for automatically acquiring a capacitance withstand voltage insulation test threshold according to claim 1, characterized in that: The withstand voltage / insulation threshold is related to the capacity of the capacitor, the test voltage, and the test time. Different test equipment has different test positions, movement times, and charging times, so upper and lower limits need to be adjusted.

10. The method for automatically acquiring a capacitance withstand voltage insulation test threshold according to claim 1, characterized in that: The upper and lower limit adjustment is specifically to achieve upper and lower limit setting under the condition that the sample condition is 300-1000 capacitors.

Images