Fault analysis method for direct current charging pile

Abstract

The invention discloses a DC charging pile fault analysis method, which comprises the following steps of S1, constructing sample features for a plurality of fault-free charging piles in a working state; s2, constructing sample features for a plurality of charging piles with known fault types in a working state; s3, constructing a sample set based on the obtained sample features; s4, constructing a classifier model based on a machine learning method, and training the classifier model by using the sample set to obtain a mature classifier model; and S5, for a to-be-analyzed charging pile, constructing sample features, and sending the sample features into the mature classifier model for fault analysis. According to the invention, the sample features can be constructed through the output current and voltage of the charging pile, the classifier is trained in combination with machine learning to carry out fault analysis on the charging pile, and even non-professionals can directly judge the fault condition according to the output result of the classifier.

Description

Technical Field

[0001] The present invention relates to the field of charging piles, and in particular to a fault analysis method for a DC charging pile. Background Art

[0002] The large-scale installation and popularization of charging piles will help electric vehicles to charge more conveniently; however, during the use of charging piles, malfunctions often occur. At present, the maintenance of charging piles often requires professional maintenance personnel to go to the site and perform manual maintenance. In many cases, the fault analysis of charging piles is based on manual analysis, which has strict requirements on the professional ability of the staff. For non-professional staff, it is difficult to judge whether the charging pile is faulty and the type of fault. Summary of the Invention

[0003] The purpose of the present invention is to overcome the shortcomings of the existing technology and provide a DC charging pile fault analysis method. The method can construct sample features through the output current and voltage of the charging pile, and train a classifier to perform fault analysis on the charging pile in combination with machine learning. Even non-professionals can directly determine the fault situation based on the output results of the classifier.

[0004] The object of the present invention is achieved through the following technical solutions:

[0005] A DC charging pile fault analysis method includes the following steps:

[0006] S1. For multiple fault-free charging piles in working state, collect information on output voltage and output current, extract voltage characteristics and current characteristics of each fault-free charging pile, and construct sample characteristics of each fault-free charging pile;

[0007] S2. For charging piles with various known fault types in working state, collect information on output voltage and output current, extract voltage and current characteristics of charging piles with each known fault type, and construct sample characteristics of charging piles with various known fault types;

[0008] S3. Construct a sample set based on the sample features obtained in steps S1 to S2;

[0009] S4. Constructing a classifier model based on a machine learning method, training the classifier model using the sample set constructed in step S3 to obtain a mature classifier model;

[0010] S5. For the charging pile to be analyzed, when it is in working state, collect the output voltage and output current, extract the voltage features and current features, construct the sample features and send them to the mature classifier model for fault analysis.

[0011] The beneficial effects of the present invention are as follows: the present invention can construct sample features through the output current and voltage of the charging pile, and train a classifier in combination with machine learning to perform fault analysis on the charging pile. Even non-professionals can directly determine the fault situation based on the output results of the classifier. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0013] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0014] like Figure 1 As shown, a DC charging pile fault analysis method is characterized by comprising the following steps:

[0015] S1. For multiple fault-free charging piles in working state, collect information on output voltage and output current, extract voltage characteristics and current characteristics of each fault-free charging pile, and construct sample characteristics of each fault-free charging pile;

[0016] The step S1 comprises:

[0017] S101. For any fault-free charging pile, after it is in working state, collect the current information I1, I2, ..., I of the charging pile within K moments. K And voltage information U1, U2, ..., U K ;

[0018] S102. Calculate I1, I2, ..., I K Average value and U1,U2,...,U K Average value

[0019] S103. Calculate I1, I2, ..., I K and the average Deviation As a current characteristic;

[0020] S104. Calculate U1, U2, ..., U K and the average Deviation As voltage characteristics;

[0021] S105. Construct a sample feature matrix of the current fault-free charging pile:

[0022]

[0023] S106. For each fault-free charging pile, repeat steps S101 to S105 to obtain sample features of each fault-free charging pile.

[0024] S2. For charging piles with various known fault types in working state, collect information on output voltage and output current, extract voltage and current characteristics of charging piles with each known fault type, and construct sample characteristics of charging piles with various known fault types;

[0025] The step S2 comprises:

[0026] S201. For any known fault type of charging pile;

[0027] S202. Put the charging pile of the currently known fault type into working state and collect the current information I1′, I2′, ..., I′ of the charging pile within K moments K and voltage information U1′, U2′, ..., U′ K ;

[0028] S203. Calculate I1′, I2′, ..., I′ K Average value and U1′,U2′,...,U′ K Average value

[0029] S204. Calculate I1′, I2′, ..., I′ K and the average Deviation As a current characteristic;

[0030] S205. Calculate U1′, U2′, ..., U′ K and the average Deviation As voltage characteristics;

[0031] S206. Construct a sample feature matrix of charging piles with currently known fault types:

[0032]

[0033] S207. For a charging pile with a known fault type, repeat steps S202 to S206 multiple times to obtain multiple sample feature matrices;

[0034] S208. For each charging pile with a known fault type, repeat steps S201 to S207 to obtain sample features of the charging pile with each known fault type.

[0035] S3. Construct a sample set based on the sample features obtained in steps S1 to S2;

[0036] The step S3 comprises:

[0037] For each sample feature obtained in step S1, add a label of 0;

[0038] For each sample feature obtained in step S2, the charging pile fault type corresponding to the sample feature is used as a label:

[0039] Assume there are M types of known fault types for charging piles. Types 1 to M are used to represent the fault types, and 1 to M are used to mark the features of each sample.

[0040] After labeling is completed, all sample features are added into a set to form a sample set.

[0041] S4. Constructing a classifier model based on a machine learning method, training the classifier model using the sample set constructed in step S3 to obtain a mature classifier model; the classification algorithm in the machine learning algorithm includes a KNN algorithm and a BP neural network algorithm;

[0042] S5. For the charging pile to be analyzed, when it is in working state, collect the output voltage and output current, extract the voltage features and current features, construct the sample features and send them to the mature classifier model for fault analysis.

[0043] The step S5 comprises:

[0044] S501. For the charging pile to be analyzed, when it is in working state, collect the current information I1″, I2″, ..., I′ of the charging pile within K moments K ′ and voltage information U1″, U2″, ..., U′ K ';

[0045] S502. Calculate I1″, I2″, ..., I′ K The average value of ′ and U1″,U2″,...,U′ K The average value of ′

[0046] S503. Calculate I1″, I2″, ..., I′ K ′ and the average value Deviation As a current characteristic;

[0047] S504. Calculate U1″, U2″, ..., U′ K ′ and the average value Deviation As voltage characteristics;

[0048] S505. Construct the sample feature matrix of the charging pile to be analyzed:

[0049]

[0050] S506. Input the sample feature matrix of the charging pile to be analyzed into a mature classifier model, and perform fault analysis on the DC charging pile based on the results output by the model:

[0051] If the classifier model outputs 0, it is considered that the charging pile to be analyzed is not faulty;

[0052] If the classifier model outputs m, where m is an integer from 1 to M, then the charging pile to be analyzed currently has the mth type of fault.

[0053] The foregoing description is a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein and should not be construed as excluding other embodiments. Instead, the present invention can be used in other combinations, modifications, and environments and can be modified within the scope of the concept described herein through the above teachings or techniques or knowledge in the relevant field. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention are intended to be protected by the appended claims.

Claims

1. A DC charging pile fault analysis method, characterized by: The following steps are involved: S1. For multiple fault-free charging piles in working state, collect information on output voltage and output current, extract voltage characteristics and current characteristics of each fault-free charging pile, and construct sample characteristics of each fault-free charging pile; S2. For charging piles with various known fault types in working state, collect information on output voltage and output current, extract voltage and current characteristics of charging piles with each known fault type, and construct sample characteristics of charging piles with various known fault types; S3. Construct a sample set based on the sample features obtained in steps S1 to S2; S4. Constructing a classifier model based on a machine learning method, and training the classifier model using the sample set constructed in step S3 to obtain a mature classifier model; S5. For the charging pile to be analyzed, when it is in working state, collect the output voltage and output current, extract the voltage features and current features, construct the sample features and send them to the mature classifier model for fault analysis.

2. A DC charging pile fault analysis method according to claim 1, characterized in that: The classification algorithm in the machine learning algorithm includes one of a KNN algorithm and a BP neural network algorithm.

3. A DC charging pile fault analysis method according to claim 1, characterized in that: The step S1 comprises: S101. For any fault-free charging pile, after it is in working state, collect the current information I1, I2, ..., I of the charging pile within K moments. K And voltage information U1, U2, ..., U K ; S102. Calculate I1, I2, ..., I K Average value and U1,U2,...,U K Average value S103. Calculate I1, I2, ..., I K and the average Deviation As a current characteristic; S104. Calculate U1, U2, ..., U K and the average Deviation As voltage characteristics; S105. Construct a sample feature matrix of the current fault-free charging pile: S106. For each fault-free charging pile, repeat steps S101 to S105 to obtain sample features of each fault-free charging pile.

4. A DC charging pile fault analysis method according to claim 1, characterized in that: The step S2 comprises: S201. For any known fault type of charging pile; S202. Put the charging pile of the currently known fault type into working state and collect the current information I1′, I2′, ..., I′ of the charging pile within K moments K and voltage information U1′, U2′, ..., U′ K ; S203. Calculate I1′, I2′, ..., I′ K Average value and U1′,U2′,...,U′ K Average value S204. Calculate I1′, I2′, ..., I′ K and the average Deviation As a current characteristic; S205. Calculate U1′, U2′, ..., U′ K and the average Deviation As voltage characteristics; S206. Construct a sample feature matrix of charging piles with currently known fault types: S207. For a charging pile with a known fault type, repeat steps S202 to S206 multiple times to obtain multiple sample feature matrices. S208. For each charging pile with a known fault type, repeat steps S201 to S207 to obtain sample features of the charging pile with each known fault type.

5. A DC charging pile fault analysis method according to claim 1, characterized in that: The step S3 comprises: For each sample feature obtained in step S1, add a label of 0; For each sample feature obtained in step S2, the charging pile fault type corresponding to the sample feature is used as a label: Assume that there are M types of known fault types for charging piles. Types 1 to M are used to represent the fault types, and 1 to M are used to mark the features of each sample. After labeling is completed, all sample features are added into a set to form a sample set.

6. A DC charging pile fault analysis method according to claim 1, characterized in that: The step S5 comprises: S501. For the charging pile to be analyzed, when it is in working state, collect the current information I1″, I2″, ..., I′ of the charging pile within K moments K ′ and voltage information U1″, U2″, ..., U′ K '; S502. Calculate I1″, I2″, ..., I′ K The average value of ′ and U1″,U2″,...,U′ K The average value of ′ S503. Calculate I1″, I2″, ..., I′ K ′ and the average value Deviation As a current characteristic; S504. Calculate U1″, U2″, ..., U′ K ′ and the average value Deviation As voltage characteristics; S505. Construct the sample feature matrix of the charging pile to be analyzed: S506. Input the sample feature matrix of the charging pile to be analyzed into a mature classifier model, and perform fault analysis on the DC charging pile based on the results output by the model: If the classifier model outputs 0, it is considered that the charging pile to be analyzed is not faulty; If the classifier model outputs m, where m is an integer from 1 to M, then the charging pile to be analyzed currently has the mth type of fault.

Images