DC fast charging contactor life prediction system and method based on big data mining

Abstract

The present invention discloses a system and method for predicting the life of a DC fast charging contactor based on big data mining, and relates to the technical field of electronic component life prediction. The present invention includes a judgment parameter acquisition module for obtaining the service life condition judgment parameters of the DC fast charging contactor; a database processing module, based on the service life condition judgment parameters of the DC fast charging contactor, forms a database of the service life condition judgment parameters of the DC fast charging contactor; a data screening module, based on the service life condition judgment parameter database of the DC fast charging contactor, constructs a data screening model to form life prediction historical learning data; a life prediction module, based on the life prediction historical learning data, forms a DC fast charging contactor life prediction model; and a warning module. By adopting the method of big data mining, the continuous changes of data are used to analyze the changes between various parameters and the life of the DC fast charging contactor, so as to achieve accurate prediction of the life and prevent problems before they occur.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic component life prediction, and in particular to a system and method for predicting the life of a DC fast-charging contactor based on big data mining. Background Art

[0002] As one of the important components of electric vehicle charging piles, the service life of DC fast charging contactors has always been one of the important consideration parameters of electric vehicle charging piles. As the number of times the DC fast charging contactor is used increases, it will cause the DC fast charging contactor to malfunction, which will in turn cause the performance of the product to decline. At present, in most cases, the service life of the DC fast charging contactor is assessed by the number of uses. The same service life is generally calibrated at the factory, and an alarm is processed when a certain number of uses is reached; or relevant conditional parameters are set, and temporary measurements are used. When certain parameters are in an abnormal state, an alarm is processed. However, neither of them can realize digital prediction of life. The first one can only reflect a vague performance situation, while the second one requires real-time measurement, which can only evaluate the present and cannot realize prediction; therefore, it is currently impossible to realize self-life prediction of DC fast charging contactors based on the use of DC fast charging contactors in different environments and with different parameters. Summary of the invention

[0003] The purpose of the present invention is to provide a system and method for predicting the life of a DC fast charging contactor based on big data mining to solve the problems raised in the prior art.

[0004] To achieve the above object, the present invention provides the following technical solution: a DC fast charging contactor life prediction method based on big data mining, the method comprising:

[0005] Obtain service life condition judgment parameters of DC fast charging contactor;

[0006] Based on the service life condition judgment parameters of the DC fast charging contactor, a service life condition judgment parameter database of the DC fast charging contactor is formed, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor;

[0007] Based on the life condition judgment parameter database of DC fast charging contactors, a data screening model is constructed to form life prediction historical learning data;

[0008] Based on the historical learning data of life prediction, a DC fast charging contactor life prediction model is formed;

[0009] The service life condition judgment parameter of the DC fast charging contactor to be detected is obtained, and the predicted service life of the DC fast charging contactor is determined based on the DC fast charging contactor life prediction model.

[0010] According to the above technical solution, the internal use parameters of the DC fast charging contactor include:

[0011] The voltage difference between the two main contacts;

[0012] Temperature of main contacts;

[0013] The number of times the main contacts are on and off;

[0014] The output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state.

[0015] In the above parameter collection process, a voltage collector, a temperature sensor and a counter are set, wherein the voltage collector can measure the voltage difference between the two main contacts, and measure the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the pull-in working state; the temperature sensor can measure the temperature of the main contact; and the counter can record the number of on and off times of the main contact;

[0016] According to the above technical solution, the environmental parameters of the DC fast charging contactor include:

[0017] The number of times the charging pile to which the DC fast charging contactor belongs is used;

[0018] The continuous use time of the charging pile to which the DC fast charging contactor belongs;

[0019] The intermittent duration of the charging pile to which the DC fast charging contactor belongs.

[0020] According to the above technical solution, the construction of the data screening model includes:

[0021] Obtaining a service life condition judgment parameter when a DC fast charging contactor fails, wherein the failure refers to the DC fast charging contactor being in a scrapped state;

[0022] Data screening is performed in the service life condition judgment parameters when the DC fast charging contactor fails, specifically including:

[0023] Construct a time period T, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T, and obtain the voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, and the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state under each working condition within the time period T; form a set of data under each working condition;

[0024] Get the usage count of the charging pile to which the DC fast charging contactor belongs;

[0025] Obtain the continuous use time of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the first output value;

[0026] Obtain the intermittent duration of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the second output value;

[0027] Where N refers to the number of groups of working data within the time period T;

[0028] Write each set of data and the number of times the corresponding DC fast charging contactor's charging pile is used, a random first output value, and a random second output value into the same data column. The first output value and the second output value are not selected repeatedly during the selection process. They serve as the life prediction historical learning data of the DC fast charging contactor. Each set of life prediction historical learning data corresponds to a life time; the life time refers to the time period between the start of use of the DC fast charging contactor and the occurrence of a fault.

[0029] According to the above technical solution, the DC fast charging contactor life prediction model includes:

[0030] Taking the life time as the dependent variable and the historical learning data of the life prediction of the DC fast charging contactor as the independent variable, a set of multivariate linear regression functions is formed as the initial function:

[0031]

[0032] in, Refers to life span; They refer to the regression coefficients corresponding to each parameter respectively; The voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state, the number of times the charging pile to which the DC fast charging contactor belongs is used, the random first output value, and the random second output value respectively correspond to;

[0033] A loss function is formed for the initial function, and the weak learner corresponding to the minimum loss function is used as the initial weak learner of the initial training set; a negative gradient processing formula is constructed to form a negative gradient for each set of life prediction historical learning data i in the multivariate linear regression function. :

[0034]

[0035] in, For each set of lifespan prediction historical learning data, for The corresponding loss function; Use the model of the previous round of learners; t represents the current number of iterations; Refers to differential;

[0036] Based on the negative gradient, a regression tree is formed, and the leaf node area of ​​the tth regression tree is recorded as , use regression tree to fit, form the best fitting value, and construct a strong learner of weak learner based on the best fitting value:

[0037]

[0038] in, represents the strong learner obtained in the tth round of iteration; Represents the learner of the previous round; represents the best fitting value; j and J represent the leaf nodes and leaf regions on the regression tree respectively; I represents the best fitting value Combination, represents the decision tree fitting function of this round;

[0039] The formed strong learner replaces the original multiple linear regression function to form a new multiple linear regression function output.

[0040] According to the above technical solution, the service life condition judgment parameters of the DC fast charging contactor to be tested are obtained and substituted into the new multivariate linear regression function to form the predicted life span of the DC fast charging contactor, and the remaining life span is calculated. If the remaining life span is lower than the threshold set by the system, it is fed back to the administrator port.

[0041] A DC fast charging contactor life prediction system based on big data mining, the system includes:

[0042] A judgment parameter acquisition module is used to obtain the service life condition judgment parameters of the DC fast charging contactor;

[0043] A database processing module, based on the service life condition judgment parameters of the DC fast charging contactor, forms a service life condition judgment parameter database of the DC fast charging contactor, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor;

[0044] The data screening module builds a data screening model based on the life condition judgment parameter database of the DC fast charging contactor to form the life prediction historical learning data;

[0045] The life prediction module forms a DC fast charging contactor life prediction model based on the life prediction history learning data;

[0046] The warning module obtains the service life condition judgment parameter of the DC fast charging contactor to be detected, and determines the predicted service life of the DC fast charging contactor based on the DC fast charging contactor life prediction model;

[0047] The output end of the judgment parameter acquisition module is connected to the input end of the database processing module; the output end of the database processing module is connected to the input end of the data screening module; the output end of the data screening module is connected to the input end of the life prediction module; the output end of the life prediction module is connected to the input end of the warning module.

[0048] According to the above technical solution, the data screening module also includes:

[0049] A screening time period T is determined, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T.

[0050] According to the above technical solution, the life prediction module also includes:

[0051] Taking the life time as the dependent variable and the historical learning data of the life prediction of DC fast charging contactor as the independent variable, a set of multiple linear regression functions is formed as the initial function, and the initial function is learned and analyzed to output a new multiple linear regression function.

[0052] According to the above technical solution, the warning module also includes:

[0053] Obtain the service life condition judgment parameters of the DC fast charging contactor to be tested, substitute them into the new multivariate linear regression function, form the predicted life span of the DC fast charging contactor, calculate the remaining life span, and if the remaining life span is lower than the threshold set by the system, feedback is sent to the administrator port.

[0054] Compared with the prior art, the beneficial effects of the present invention are: the present invention can realize the life prediction of the DC fast charging contactor, adopt the big data mining method, utilize the continuous changes of data to analyze the changes between various parameters and the life of the DC fast charging contactor, realize accurate prediction of the life, and prevent problems before they occur. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a flow chart of a method for predicting the life of a DC fast charging contactor based on big data mining. DETAILED DESCRIPTION

[0056] 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.

[0057] Example: Figure 1As shown, the present invention provides a method for predicting the life of a DC fast charging contactor based on big data mining, the method comprising:

[0058] Obtain service life condition judgment parameters of DC fast charging contactor;

[0059] Based on the service life condition judgment parameters of the DC fast charging contactor, a service life condition judgment parameter database of the DC fast charging contactor is formed, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor;

[0060] The internal use parameters of the DC fast charging contactor include:

[0061] The voltage difference between the two main contacts;

[0062] Temperature of main contacts;

[0063] The number of times the main contacts are on and off;

[0064] The output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state.

[0065] The environmental parameters of the DC fast charging contactor include:

[0066] The number of times the charging pile to which the DC fast charging contactor belongs is used;

[0067] The continuous use time of the charging pile to which the DC fast charging contactor belongs;

[0068] The intermittent duration of the charging pile to which the DC fast charging contactor belongs.

[0069] The constructing of the data screening model comprises:

[0070] Obtaining a service life condition judgment parameter when a DC fast charging contactor fails, wherein the failure refers to the DC fast charging contactor being in a scrapped state;

[0071] Data screening is performed in the service life condition judgment parameters when the DC fast charging contactor fails, specifically including:

[0072] Construct a time period T, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T, and obtain the voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, and the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state under each working condition within the time period T; form a set of data under each working condition;

[0073] Get the usage count of the charging pile to which the DC fast charging contactor belongs;

[0074] Obtain the continuous use time of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the first output value;

[0075] Obtain the intermittent duration of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the second output value;

[0076] Where N refers to the number of groups of working data within the time period T;

[0077] Write each set of data and the number of times the corresponding DC fast charging contactor's charging pile is used, a random first output value, and a random second output value into the same data column. The first output value and the second output value are not selected repeatedly during the selection process. They serve as the life prediction historical learning data of the DC fast charging contactor. Each set of life prediction historical learning data corresponds to a life time; the life time refers to the time period between the start of use of the DC fast charging contactor and the occurrence of a fault.

[0078] Based on the historical learning data of life prediction, a DC fast charging contactor life prediction model is formed;

[0079] The service life condition judgment parameter of the DC fast charging contactor to be detected is obtained, and the predicted service life of the DC fast charging contactor is determined based on the DC fast charging contactor life prediction model.

[0080] The DC fast charging contactor life prediction model includes:

[0081] Taking the life time as the dependent variable and the historical learning data of the life prediction of the DC fast charging contactor as the independent variable, a set of multivariate linear regression functions is formed as the initial function:

[0082]

[0083] in, Refers to life span; They refer to the regression coefficients corresponding to each parameter respectively; The voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state, the number of times the charging pile to which the DC fast charging contactor belongs is used, the random first output value, and the random second output value respectively correspond to;

[0084] A loss function is formed for the initial function, and the weak learner corresponding to the minimum loss function is used as the initial weak learner of the initial training set; a negative gradient processing formula is constructed to form a negative gradient for each set of life prediction historical learning data i in the multivariate linear regression function. :

[0085]

[0086] in, For each set of lifespan prediction historical learning data, for The corresponding loss function; Use the model of the previous round of learners; t represents the current number of iterations; Refers to differential;

[0087] Based on the negative gradient, a regression tree is formed, and the leaf node area of ​​the tth regression tree is recorded as , use regression tree to fit, form the best fitting value, and construct a strong learner of weak learner based on the best fitting value:

[0088]

[0089] in, represents the strong learner obtained in the tth round of iteration; Represents the learner of the previous round; represents the best fitting value; j and J represent the leaf nodes and leaf regions on the regression tree respectively; I represents the best fitting value Combination, represents the decision tree fitting function of this round;

[0090] The formed strong learner replaces the original multiple linear regression function to form a new multiple linear regression function output.

[0091] Obtain the service life condition judgment parameters of the DC fast charging contactor to be tested, substitute them into the new multivariate linear regression function, form the predicted life span of the DC fast charging contactor, calculate the remaining life span, and if the remaining life span is lower than the threshold set by the system, feedback is sent to the administrator port.

[0092] In this embodiment, a DC fast charging contactor life prediction system based on big data mining is also included, and the system includes:

[0093] A judgment parameter acquisition module is used to obtain the service life condition judgment parameters of the DC fast charging contactor;

[0094] A database processing module, based on the service life condition judgment parameters of the DC fast charging contactor, forms a service life condition judgment parameter database of the DC fast charging contactor, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor;

[0095] The data screening module builds a data screening model based on the life condition judgment parameter database of the DC fast charging contactor to form the life prediction historical learning data;

[0096] The life prediction module forms a DC fast charging contactor life prediction model based on the life prediction history learning data;

[0097] The warning module obtains the service life condition judgment parameter of the DC fast charging contactor to be detected, and determines the predicted service life of the DC fast charging contactor based on the DC fast charging contactor life prediction model;

[0098] The output end of the judgment parameter acquisition module is connected to the input end of the database processing module; the output end of the database processing module is connected to the input end of the data screening module; the output end of the data screening module is connected to the input end of the life prediction module; the output end of the life prediction module is connected to the input end of the warning module.

[0099] The data screening module also includes:

[0100] A screening time period T is determined, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T.

[0101] The life prediction module also includes:

[0102] Taking the life time as the dependent variable and the historical learning data of the life prediction of DC fast charging contactor as the independent variable, a set of multiple linear regression functions is formed as the initial function, and the initial function is learned and analyzed to output a new multiple linear regression function.

[0103] The warning module also includes:

[0104] Obtain the service life condition judgment parameters of the DC fast charging contactor to be tested, substitute them into the new multivariate linear regression function, form the predicted life span of the DC fast charging contactor, calculate the remaining life span, and if the remaining life span is lower than the threshold set by the system, feedback is sent to the administrator port.

[0105] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for predicting the life of a DC fast charging contactor based on big data mining, characterized in that: The method includes: Obtain service life condition judgment parameters of DC fast charging contactor; Based on the service life condition judgment parameters of the DC fast charging contactor, a service life condition judgment parameter database of the DC fast charging contactor is formed, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor; Based on the life condition judgment parameter database of DC fast charging contactors, a data screening model is constructed to form life prediction historical learning data; Based on the historical learning data of life prediction, a DC fast charging contactor life prediction model is formed; Obtain the service life condition judgment parameters of the DC fast charging contactor to be detected, and determine the predicted service life of the DC fast charging contactor based on the DC fast charging contactor life prediction model; the environmental parameters of the DC fast charging contactor include: The number of times the charging pile to which the DC fast charging contactor belongs is used; The continuous use time of the charging pile to which the DC fast charging contactor belongs; The intermittent duration of the charging pile to which the DC fast charging contactor belongs; The constructing of the data screening model comprises: Obtaining a service life condition judgment parameter when a DC fast charging contactor fails, wherein the failure refers to the DC fast charging contactor being in a scrapped state; Data screening is performed in the service life condition judgment parameters when the DC fast charging contactor fails, specifically including: Construct a time period T, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T, and obtain the voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, and the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state under each working condition within the time period T; form a set of data under each working condition; Get the usage count of the charging pile to which the DC fast charging contactor belongs; Obtain the continuous use time of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the first output value; Obtain the intermittent duration of the charging pile to which the DC fast charging contactor belongs, sort them from large to small, and take the first N groups of data as the second output value; Where N refers to the number of groups of working data within the time period T; Write each set of data and the number of times the charging pile to which the corresponding DC fast charging contactor belongs is used, the random first output value, and the random second output value into the same data column, and the first output value and the second output value are not repeatedly selected during the selection process, as the life prediction historical learning data of the DC fast charging contactor, and each set of life prediction historical learning data corresponds to a life time; the life time refers to the time period between the start of use of the DC fast charging contactor and the occurrence of a fault; The DC fast charging contactor life prediction model includes: Taking the life time as the dependent variable and the historical learning data of the life prediction of the DC fast charging contactor as the independent variable, a set of multivariate linear regression functions is formed as the initial function: ； in, Refers to life span; They refer to the regression coefficients corresponding to each parameter respectively; The voltage difference between the two main contacts, the temperature of the main contacts, the number of on-off times of the main contacts, the output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state, the number of times the charging pile to which the DC fast charging contactor belongs is used, the random first output value, and the random second output value respectively correspond to; A loss function is formed for the initial function, and the weak learner corresponding to the minimum loss function is used as the initial weak learner of the initial training set; a negative gradient processing formula is constructed to form a negative gradient for each set of life prediction historical learning data i in the multivariate linear regression function. : ； in, For each set of lifespan prediction historical learning data, for The corresponding loss function; Use the model of the previous round of learners; t represents the current number of iterations; Refers to differential; Based on the negative gradient, a regression tree is formed, and the leaf node area of ​​the tth regression tree is recorded as , use regression tree to fit, form the best fitting value, and construct a strong learner of weak learner based on the best fitting value: ； in, represents the strong learner obtained in the tth round of iteration; Represents the learner of the previous round; represents the best fitting value; j and J represent the leaf nodes and leaf regions on the regression tree respectively; I represents the best fitting value Combination, represents the decision tree fitting function of this round; The formed strong learner replaces the original multiple linear regression function to form a new multiple linear regression function output.

2. The method for predicting the life of a DC fast charging contactor based on big data mining according to claim 1 is characterized in that: The internal use parameters of the DC fast charging contactor include: The voltage difference between the two main contacts; Temperature of main contacts; The number of times the main contacts are on and off; The output voltage value of the auxiliary contact to which the detection voltage is applied when the main contact is in the energized working state.

3. The method for predicting the life of a DC fast charging contactor based on big data mining according to claim 2 is characterized in that: Obtain the service life condition judgment parameters of the DC fast charging contactor to be tested, substitute them into the new multivariate linear regression function, form the predicted life span of the DC fast charging contactor, calculate the remaining life span, and if the remaining life span is lower than the threshold set by the system, feedback is sent to the administrator port.

4. A DC fast charging contactor life prediction system based on big data mining, using a DC fast charging contactor life prediction method based on big data mining as claimed in any one of claims 1 to 3, characterized in that: The system includes: A judgment parameter acquisition module is used to obtain the service life condition judgment parameters of the DC fast charging contactor; A database processing module, based on the service life condition judgment parameters of the DC fast charging contactor, forms a service life condition judgment parameter database of the DC fast charging contactor, wherein the service life condition judgment parameters include internal use parameters of the DC fast charging contactor and environmental parameters of the DC fast charging contactor; The data screening module builds a data screening model based on the life condition judgment parameter database of the DC fast charging contactor to form the life prediction historical learning data; The life prediction module forms a DC fast charging contactor life prediction model based on the life prediction history learning data; The warning module obtains the service life condition judgment parameter of the DC fast charging contactor to be detected, and determines the predicted service life of the DC fast charging contactor based on the DC fast charging contactor life prediction model; The output end of the judgment parameter acquisition module is connected to the input end of the database processing module; the output end of the database processing module is connected to the input end of the data screening module; the output end of the data screening module is connected to the input end of the life prediction module; the output end of the life prediction module is connected to the input end of the warning module.

5. The DC fast charging contactor life prediction system based on big data mining according to claim 4 is characterized in that: The data screening module also includes: A screening time period T is determined, where the time period T refers to a period of time selected from the time when the fault occurs as the end time of the time period T.

6. The DC fast charging contactor life prediction system based on big data mining according to claim 4 is characterized in that: The life prediction module also includes: Taking the life time as the dependent variable and the historical learning data of the life prediction of DC fast charging contactor as the independent variable, a set of multiple linear regression functions is formed as the initial function, and the initial function is learned and analyzed to output a new multiple linear regression function.

7. The DC fast charging contactor life prediction system based on big data mining according to claim 6 is characterized in that: The warning module also includes: Obtain the service life condition judgment parameters of the DC fast charging contactor to be tested, substitute them into the new multivariate linear regression function, form the predicted life span of the DC fast charging contactor, calculate the remaining life span, and if the remaining life span is lower than the threshold set by the system, feedback is sent to the administrator port.

Images