CN116502058B

CN116502058B - AI fault detection analysis method and system applied to charging pile system and cloud platform

Info

Publication number: CN116502058B
Application number: CN202310772965.1A
Authority: CN
Inventors: 许青松; 陈锐; 韩振雨; 罗光盛; 葛淼
Original assignee: Changyuan Shenrui Energy Technology Co ltd
Current assignee: Changyuan Shenrui Energy Technology Co ltd
Priority date: 2023-06-28
Filing date: 2023-06-28
Publication date: 2023-09-26
Anticipated expiration: 2043-06-28
Also published as: CN116502058A

Abstract

The application provides an AI fault detection analysis method, an AI fault detection analysis system and a cloud platform applied to a charging pile system, wherein the AI fault detection analysis cloud platform respectively acquires a plurality of task deterministic parameters of a plurality of operation fault hidden danger item analysis tasks and a plurality of deterministic inhibition coefficients between every two operation fault hidden danger item analysis tasks, clears tasks with false detection and error induction from the plurality of operation fault hidden danger item analysis tasks according to the plurality of task deterministic parameters and the plurality of deterministic inhibition coefficients, and then determines a target analysis task needing to dynamically monitor fault hidden danger, so that predictability, accuracy and reliability of dynamic monitoring can be ensured when determining fault hidden danger detection guidance, and the unmanned charging pile system is subjected to accurate and efficient fault detection analysis by utilizing an artificial intelligence technology.

Description

AI fault detection analysis method and system applied to charging pile system and cloud platform

Technical Field

The application relates to the technical field of artificial intelligence, in particular to an AI fault detection and analysis method and system applied to a charging pile system and a cloud platform.

Background

Charging pile (Charging pile) refers to a Charging device for providing energy for electric vehicles, has a function similar to that of an oiling machine in a gas station, can be fixed on the ground or on a wall, is installed in public buildings (public buildings, malls, public parking lots and the like) and residential district parking lots or Charging stations, and can charge electric vehicles of various types according to different voltage levels.

The input end of the charging pile is directly connected with an alternating current power grid, and the output end of the charging pile is provided with a charging plug for charging the electric automobile. The charging pile generally provides two charging modes of conventional charging and quick charging, people can use a specific charging card to swipe the card on a man-machine interaction operation interface provided by the charging pile for corresponding charging operation and cost data printing, and a charging pile display screen can display data such as charging amount, cost and charging time.

Most of charging stations of the charging pile type in the current industry are unattended, and the failure rate of the charging pile system in the operation process is high. Therefore, how to accurately and reliably perform fault detection analysis of the charging pile system is a difficulty in operation of the charging equipment industry.

Disclosure of Invention

The application provides an AI fault detection analysis method and system applied to a charging pile system and a cloud platform.

The first aspect is an AI fault detection analysis method applied to a charging pile system, applied to an AI fault detection analysis cloud platform, the method comprising:

determining a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks through a selected operation report text of the charging stake system and a preamble charging stake system operation report text which is related to the upstream and downstream of the selected operation report text; wherein, each operation fault hidden trouble item analysis task in the plurality of operation fault hidden trouble item analysis tasks corresponds to a task certainty parameter;

determining a deterministic inhibition coefficient between every two operation fault hidden trouble item analysis tasks in the plurality of operation fault hidden trouble item analysis tasks according to the operation report text of the selected charging pile system;

and determining a target operation fault hidden danger item analysis task from the operation fault hidden danger item analysis tasks according to the task deterministic parameters and the deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item.

In some optional examples, the determining, by the selected charging stake system operation report text and the preamble charging stake system operation report text associated with the selected charging stake system operation report text upstream and downstream, a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks includes:

determining a plurality of operation fault hidden danger item analysis tasks and a plurality of hidden danger analysis decision scores corresponding to the operation fault hidden danger item analysis tasks according to the operation report text of the selected charging pile system, wherein each operation fault hidden danger item analysis task in the operation fault hidden danger item analysis tasks corresponds to one hidden danger analysis decision score;

determining a plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks according to the operation report text of the selected charging pile system and the operation report text of the prefronous charging pile system, wherein each operation fault hidden danger item analysis task corresponds to one upstream and downstream viscosity index;

and determining a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks according to the plurality of hidden danger analysis decision scores and the plurality of upstream and downstream viscosity indexes.

In some optional examples, the determining, according to the selected charging pile system operation report text and the preamble charging pile system operation report text, a plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks includes:

determining a plurality of past operation fault hidden danger item analysis tasks in an operation report text of the preamble charging pile system;

determining a plurality of task chain related variables between a first operation fault hidden danger item analysis task and the plurality of past operation fault hidden danger item analysis tasks, wherein the first operation fault hidden danger item analysis task is one operation fault hidden danger item analysis task in the plurality of operation fault hidden danger item analysis tasks;

determining the maximum associated variable in the plurality of task chain associated variables as a first upstream and downstream viscosity index corresponding to the first operation fault hidden danger item analysis task;

and determining a plurality of first upstream and downstream viscosity indexes corresponding to the plurality of first operation fault hidden danger item analysis tasks so as to determine the plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks.

In some optional examples, the determining, according to the operation report text of the selected charging pile system, a deterministic inhibition coefficient between each two operation fault hidden trouble item analysis tasks in the plurality of operation fault hidden trouble item analysis tasks includes:

Determining task chain matching degree and hidden danger detail matching degree between every two operation fault hidden danger item analysis tasks in the operation report text of the selected charging pile system;

and determining the certainty factor between every two operation fault hidden danger item analysis tasks according to the task chain matching degree and the hidden danger detail matching degree.

In some optional examples, determining the task chain matching degree between every two operation fault hidden danger item analysis tasks in the operation report text of the selected charging pile system includes:

respectively acquiring an analysis report text set of a first operation fault hidden danger item analysis task and an analysis report text set of a second operation fault hidden danger item analysis task, wherein the first operation fault hidden danger item analysis task and the second operation fault hidden danger item analysis task are every two operation fault hidden danger item analysis tasks;

and determining the task chain matching degree between the first operation fault hidden danger item analysis task and the second operation fault hidden danger item analysis task according to the analysis report text set of the first operation fault hidden danger item analysis task and the analysis report text set of the second operation fault hidden danger item analysis task so as to determine the task chain matching degree between every two operation fault hidden danger item analysis tasks.

In some optional examples, the determining, according to the plurality of task deterministic parameters and the deterministic suppression coefficients, a target operation fault hidden danger item analysis task from the plurality of operation fault hidden danger item analysis tasks includes:

taking the task deterministic parameters as the contribution degree of task analysis suppression elements of a task analysis suppression matrix;

the deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks is used as the contribution degree of query characteristics between two task analysis inhibition elements corresponding to every two operation fault hidden danger item analysis tasks, and a task analysis inhibition matrix is generated;

determining at least one inhibition matrix block in the task analysis inhibition matrix, and determining a first inhibition matrix block from the at least one inhibition matrix block according to task certainty parameters and certainty suppression coefficients included in the at least one inhibition matrix block;

and determining the operation fault hidden danger item analysis task included in the first inhibition matrix block as the target operation fault hidden danger item analysis task.

In some optional examples, the determining a first suppression matrix segment from the at least one suppression matrix segment according to the task deterministic parameter and the deterministic suppression coefficient included in the at least one suppression matrix segment includes:

Determining at least one group of report text corresponding to at least one inhibition matrix block respectively, wherein each inhibition matrix block in the at least one inhibition matrix block corresponds to one group of report text, and the group of report text comprises at least one report text;

determining at least one deduction weight corresponding to each suppression matrix block in the at least one suppression matrix block according to task certainty parameters and certainty suppression coefficients included in the at least one report text, wherein each report text in the at least one report text corresponds to one deduction weight;

determining a target deduction weight with the maximum deduction weight from the least one deduction weight corresponding to each inhibition matrix block until determining the least one target deduction weight corresponding to the least one inhibition matrix block;

determining at least one local semantic vector set corresponding to the least one target deduction weight from the at least one inhibition matrix partition;

and splicing the at least one local semantic vector set into the first suppression matrix partition.

In some optional examples, after the determining the first target operational fault potential term included in the target operational fault potential term analysis task, the method further includes:

Determining a fault hidden danger trend analysis result corresponding to the first target operation fault hidden danger item and a detection interference trend analysis result corresponding to a fault detection interference item according to the operation report text of the selected charging pile system, wherein the fault detection interference item is an operation item with the highest contact factor with the target operation fault hidden danger item in the first target operation fault hidden danger items;

determining a past hidden danger state expression vector set corresponding to the first target operation fault hidden danger item and a past interference state expression vector set corresponding to the fault detection interference item according to a previous charging pile system operation report text set before the selected charging pile system operation report text;

determining a current fault hidden danger trend positioning identifier and a current hidden danger state expression vector corresponding to a second target fault hidden danger item through a later charging pile system operation report text which is related with the selected charging pile system operation report text in the upstream-downstream direction, wherein the second target fault hidden danger item is a target fault hidden danger item included in a target fault hidden danger item analysis task of the later charging pile system operation report text;

determining the relevance of the charging fault items between the first target operation fault hidden danger item and the second target operation fault hidden danger item according to the fault hidden danger trend analysis result, the past hidden danger state expression vector set, the current fault hidden danger trend positioning identification and the current hidden danger state expression vector;

Determining the relevance of a charging interference item according to the analysis result of the detected interference trend, the past interference state expression vector set, the current fault hidden danger trend positioning identifier and the current hidden danger state expression vector;

and determining fault hidden danger detection guidance of the first target operation fault hidden danger item according to the charging fault item association and the charging interference item association.

In some optional examples, the determining the fault hidden danger detection guidance of the first target operation fault hidden danger item according to the charging fault item association and the charging interference item association includes:

determining fault hidden danger deduction tracing information between the first target operation fault hidden danger item and the second target operation fault hidden danger item according to the charging fault item association and the charging interference item association;

and extracting auxiliary decision operation matters related to the first target operation fault hidden danger item from the second target operation fault hidden danger item by combining the fault hidden danger deduction tracing information so as to determine fault hidden danger detection guidance of the first target operation fault hidden danger item.

The second aspect is an AI fault detection analysis cloud platform comprising a memory and a processor; the memory is coupled to the processor; the memory is used for storing computer program codes, and the computer program codes comprise computer instructions; wherein the computer instructions, when executed by the processor, cause the AI fault detection analysis cloud platform to perform the method of the first aspect.

A third aspect is a computer readable storage medium having stored thereon a computer program which, when run, performs the method of the first aspect.

The fourth aspect is an AI fault detection analysis system applied to a charging pile system, including an AI fault detection analysis cloud platform and a charging pile system in communication with each other; the AI fault detection analysis cloud platform is used for: determining a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks through a selected operation report text of the charging stake system and a preamble charging stake system operation report text which is related to the upstream and downstream of the selected operation report text; wherein, each operation fault hidden trouble item analysis task in the plurality of operation fault hidden trouble item analysis tasks corresponds to a task certainty parameter; determining a deterministic inhibition coefficient between every two operation fault hidden trouble item analysis tasks in the plurality of operation fault hidden trouble item analysis tasks according to the operation report text of the selected charging pile system; and determining a target operation fault hidden danger item analysis task from the operation fault hidden danger item analysis tasks according to the task deterministic parameters and the deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item.

According to one embodiment of the application, a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks are determined through selecting a charging pile system operation report text and a preamble charging pile system operation report text with upstream-downstream association, wherein each operation fault hidden danger item analysis task in the plurality of operation fault hidden danger item analysis tasks corresponds to one task certainty parameter; determining a deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks in a plurality of operation fault hidden danger item analysis tasks by selecting an operation report text of a charging pile system; determining a target operation fault hidden danger item analysis task from a plurality of operation fault hidden danger item analysis tasks through a plurality of task deterministic parameters and deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item. The AI fault detection analysis cloud platform is used for respectively acquiring a plurality of task certainty parameters of a plurality of operation fault hidden danger item analysis tasks and a plurality of certainty suppression coefficients between every two operation fault hidden danger item analysis tasks, clearing tasks with false detection and error induction from the plurality of operation fault hidden danger item analysis tasks according to the plurality of task certainty parameters and the plurality of certainty suppression coefficients, and then determining a target analysis task needing to dynamically monitor the fault hidden danger, so that predictability, accuracy and reliability of dynamic monitoring can be ensured when determining the detection guide of the fault hidden danger, and the unmanned charging pile system is subjected to accurate and efficient fault detection analysis by utilizing an artificial intelligence technology.

Drawings

Fig. 1 is a flow chart of an AI fault detection analysis method applied to a charging pile system according to an embodiment of the present application.

Fig. 2 is a block diagram of an AI fault detection and analysis device applied to a charging pile system according to an embodiment of the present application.

Description of the embodiments

Hereinafter, the terms "first," "second," and "third," etc. are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first", "a second", or "a third", etc., may explicitly or implicitly include one or more such feature.

Fig. 1 shows a flow chart of an AI fault detection and analysis method applied to a charging pile system, which is provided by an embodiment of the present application, and the AI fault detection and analysis method applied to the charging pile system may be implemented by an AI fault detection and analysis cloud platform, which may include a memory and a processor; the memory is coupled to the processor; the memory is used for storing computer program codes, and the computer program codes comprise computer instructions; and when the processor executes the computer instructions, the AI fault detection analysis cloud platform is caused to execute the technical scheme described in the following steps.

STEP101, determining a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks through a selected operation report text of the charging pile system and a preamble charging pile system operation report text with upstream-downstream association, wherein each operation fault hidden danger item analysis task in the operation fault hidden danger item analysis tasks corresponds to one task certainty parameter.

In the embodiment of the application, the operation fault hidden danger item in the operation fault hidden danger item analysis task can be an electronic lock locking fault item and an electronic lock physical loop fault item.

In the embodiment of the application, the AI fault detection analysis cloud platform determines an operation fault hidden danger item analysis task including an operation fault hidden danger item in the operation report text of the selected charging pile system, wherein the operation fault hidden danger item analysis task can be a text window including the operation fault hidden danger item.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines a plurality of operation fault hidden danger item analysis tasks and a plurality of hidden danger analysis decision scores corresponding to the operation fault hidden danger item analysis tasks by selecting the operation report text of the charging pile system, and each operation fault hidden danger item analysis task in the operation fault hidden danger item analysis tasks corresponds to one hidden danger analysis decision score.

For example, the AI decision component performs decision trust operation on the operation fault hidden danger item analysis task in the operation report text of the selected charging pile system to obtain a hidden danger analysis decision score corresponding to the operation fault hidden danger item analysis task, where the AI decision component may be a neural network component such as a classifier that may provide the hidden danger analysis decision score of the operation fault hidden danger item analysis task.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines a plurality of upstream and downstream viscosity indexes corresponding to a plurality of operation fault hidden danger item analysis tasks by selecting the operation report text of the charging pile system and the operation report text of the preface charging pile system, and each operation fault hidden danger item analysis task corresponds to one upstream and downstream viscosity index which can be understood as the decision trust degree of time sequence continuity.

For example, the process of determining a plurality of upstream and downstream viscosity indexes corresponding to a plurality of operation fault hidden danger item analysis tasks by the AI fault detection analysis cloud platform through selecting an operation report text of the charging pile system and an operation report text of the pre-charging pile system includes: determining a plurality of past operation fault hidden danger item analysis tasks in an operation report text of the preamble charging pile system by the AI fault detection analysis cloud platform; then, the AI fault detection analysis cloud platform determines a plurality of task chain related variables between a first operation fault hidden danger item analysis task and a plurality of past operation fault hidden danger item analysis tasks, wherein the first operation fault hidden danger item analysis task is one operation fault hidden danger item analysis task in the plurality of operation fault hidden danger item analysis tasks; the AI fault detection analysis cloud platform determines the maximum associated variable in the plurality of task chain associated variables as a first upstream and downstream viscosity index corresponding to a first operation fault hidden danger item analysis task; and determining a plurality of first upstream and downstream viscosity indexes corresponding to the plurality of first operation fault hidden danger item analysis tasks, and further obtaining a plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks by the AI fault detection analysis cloud platform.

For example, the AI fault detection analysis cloud platform determines a plurality of task cross indexes (a plurality of task chain related variables) between a first operation fault hidden danger item analysis task in the operation report text of the selected charging pile system and a plurality of past operation fault hidden danger item analysis tasks (the front operation fault hidden danger item analysis tasks) in the operation report text of the front charging pile system one by one, and then the AI fault detection analysis cloud platform determines a maximum task cross index from the plurality of task cross indexes, and then the value of the maximum task cross index can be the first upstream and downstream viscosity indexes of the first operation fault hidden danger item analysis tasks. The AI fault detection analysis cloud platform is realized by adopting the scheme above for a plurality of operation fault hidden danger item analysis tasks, so that a plurality of upstream and downstream viscosity indexes corresponding to the operation fault hidden danger item analysis tasks are obtained.

In some examples, the upstream and downstream viscosity indexes (which may also be understood as time sequence correlations) corresponding to the operation fault hidden danger item analysis tasks are obtained through cross evaluation information of the operation fault hidden danger item analysis tasks of different time nodes, where the cross evaluation information is obtained through text-related range operations of the operation fault hidden danger item analysis tasks.

In the embodiment of the application, after obtaining a plurality of hidden danger analysis decision scores and a plurality of upstream and downstream viscosity indexes corresponding to a plurality of operation fault hidden danger item analysis tasks, the AI fault detection analysis cloud platform determines a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks through the plurality of hidden danger analysis decision scores and the plurality of upstream and downstream viscosity indexes.

In the embodiment of the present application, the task deterministic parameter (task trust weight) may be obtained by integrally summing (e.g., weight strengthening) the hidden danger analysis decision score and the upstream and downstream viscosity indexes, for example, transmission_weighted 1=q transmission_weighted 2+ (1-Q) transmission_weighted 3.

The transmission_weight 1 is a task certainty parameter of running a fault hidden danger item analysis task, the transmission_weight 2 is an upstream and downstream viscosity index, the transmission_weight 3 is a hidden danger analysis decision score, and the Q is a balance coefficient.

STEP102, determining a deterministic inhibition coefficient between every two operation fault hidden trouble item analysis tasks in a plurality of operation fault hidden trouble item analysis tasks by selecting an operation report text of the charging pile system.

In the embodiment of the application, after an AI fault detection analysis cloud platform determines a plurality of operation fault hidden danger item analysis tasks through a selected charging pile system operation report text and a front-end charging pile system operation report text with upstream and downstream association, the AI fault detection analysis cloud platform carries out binary group arrangement on the operation fault hidden danger item analysis tasks in the plurality of operation fault hidden danger item analysis tasks to obtain a plurality of arrangement rules, and the AI fault detection analysis cloud platform carries out calculation on certainty inhibition coefficients (noise coefficients and certainty reduction coefficients) between two operation fault hidden danger item analysis tasks in various arrangement rules (integration strategies) through the selected charging pile system operation report text.

In the embodiment of the application, the AI fault detection analysis cloud platform determines the task chain matching degree (which can be understood as the similarity of task areas) and the hidden danger detail matching degree (which can be understood as the similarity of hidden danger details) between every two operation fault hidden danger item analysis tasks in the operation report text of the selected charging pile system.

In the embodiment of the application, the AI fault detection analysis cloud platform is combined with the AI mining component to determine the fault hidden danger trend characteristics included in the operation fault hidden danger item analysis task, then the AI fault detection analysis cloud platform determines the characteristic distance between the two fault hidden danger trend characteristics, and determines the characteristic distance (such as cosine similarity) as the hidden danger detail matching degree between the two operation fault hidden danger item analysis tasks corresponding to the two fault hidden danger trend characteristics.

In the embodiment of the application, an AI fault detection analysis cloud platform respectively acquires an analysis report text set of a first operation fault hidden danger item analysis task and an analysis report text set of a second operation fault hidden danger item analysis task, wherein the first operation fault hidden danger item analysis task and the second operation fault hidden danger item analysis task are every two operation fault hidden danger item analysis tasks; and then, the AI fault detection analysis cloud platform determines the task chain matching degree between the first operation fault hidden danger item analysis task and the second operation fault hidden danger item analysis task through an analysis report text set of the first operation fault hidden danger item analysis task and an analysis report text set of the second operation fault hidden danger item analysis task so as to determine the task chain matching degree between every two operation fault hidden danger item analysis tasks.

In some examples, the analysis report text set may be a text region that occupies 1/2 of a first text box boundary value and a second text box boundary value of the operation fault hidden danger item analysis task, the AI fault detection analysis cloud platform determines a task cross index between the analysis report text set of the first operation fault hidden danger item analysis task and the analysis report text set of the second operation fault hidden danger item analysis task, and the AI fault detection analysis cloud platform determines a task chain matching degree (task region similarity) between the first operation fault hidden danger item analysis task and the second operation fault hidden danger item analysis task through the analysis report text set of the first operation fault hidden danger item analysis task, the analysis report text set of the second operation fault hidden danger item analysis task, and the task cross index. Further, the task cross index may be understood as a degree of repetition between the analysis report text set of the first operation fault potential term analysis task and the analysis report text set of the second operation fault potential term analysis task.

In the embodiment of the application, the AI fault detection analysis cloud platform determines the deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks through the task chain matching degree and hidden danger detail matching degree.

STEP103, determining a target operation fault hidden danger item analysis task from a plurality of operation fault hidden danger item analysis tasks through a plurality of task deterministic parameters and deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item.

In the embodiment of the application, after the AI fault detection analysis cloud platform respectively determines a plurality of task deterministic parameters corresponding to a plurality of operation fault hidden danger item analysis tasks and deterministic inhibition coefficients between every two operation fault hidden danger item analysis tasks, the AI fault detection analysis cloud platform determines a target operation fault hidden danger item analysis task from the plurality of operation fault hidden danger item analysis tasks through the plurality of task deterministic parameters and the deterministic inhibition coefficients, and determines a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item. The fault hidden danger dynamic monitoring can be understood as fault hidden danger directional monitoring or continuous monitoring, and is used for detecting real-time performance and predictability of a fault hidden danger layer.

In the embodiment of the application, an AI fault detection analysis cloud platform takes a plurality of task deterministic parameters as a task analysis suppression matrix (a fault detection induction vector set possibly causing deviation of fault detection) and the contribution degree (importance) of task analysis suppression elements (matrix units or matrix members); and the deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks is used as the contribution degree of the query characteristics between the two task analysis inhibition elements corresponding to every two operation fault hidden danger item analysis tasks, so that the AI fault detection analysis cloud platform generates a comprehensive and accurate task analysis inhibition matrix corresponding to a plurality of operation fault hidden danger item analysis tasks.

In the embodiment of the application, an AI fault detection analysis cloud platform determines at least one inhibition matrix block in a task analysis inhibition matrix, and determines a first inhibition matrix block from the at least one inhibition matrix block through task deterministic parameters and deterministic inhibition coefficients included in the at least one inhibition matrix block; and determining the operation fault hidden danger item analysis task included in the first inhibition matrix block as a target operation fault hidden danger item analysis task.

In the embodiment of the application, an AI fault detection analysis cloud platform sequentially analyzes task analysis inhibition matrices, sequentially determines all possible inhibition matrix blocks of task links in the task analysis inhibition matrices, and determines all possible inhibition matrix blocks as at least one inhibition matrix block, wherein the task links comprise task analysis inhibition elements and query features.

In the embodiment of the application, an AI fault detection analysis cloud platform determines a report text with the maximum deduction weight from each inhibition matrix block in at least one inhibition matrix block, and determines a set of report texts with the maximum deduction weight in each inhibition matrix block as a first inhibition matrix block of a task analysis inhibition matrix, wherein in the application process, the AI fault detection analysis cloud platform respectively determines at least one group of report texts corresponding to the at least one inhibition matrix block, each inhibition matrix block in the at least one inhibition matrix block corresponds to one group of report texts, and the group of report texts comprises at least one report text; then, determining at least one deduction weight corresponding to each suppression matrix block in at least one suppression matrix block through task deterministic parameters and deterministic suppression coefficients contained in at least one report text by the AI fault detection analysis cloud platform, wherein each report text in the at least one report text corresponds to one deduction weight; determining the target deduction weight with the maximum deduction weight from the least one deduction weight corresponding to each inhibition matrix block until determining the least one target deduction weight corresponding to the least one inhibition matrix block; finally, determining at least one local semantic vector set corresponding to the least one target deduction weight from at least one inhibition matrix block by the AI fault detection analysis cloud platform; and stitching at least one local semantic vector set into a first suppression matrix partition.

In the application process, in view of the fact that the number of task analysis suppression elements and query features in the task analysis suppression matrix is large, the AI fault detection analysis cloud platform divides the task analysis suppression matrix into at least one suppression matrix block, and determines at least one local semantic vector set from the at least one suppression matrix block respectively so as to form the at least one local semantic vector set into a first suppression matrix block, so that timeliness of determining the first suppression matrix block can be improved.

In the embodiment of the application, the deduction weight can be the difference between the task deterministic parameter evaluation index and the number of deterministic inhibition coefficients included in the report text, so that the error induction between the target operation fault hidden danger item analysis tasks determined by the deduction weight through the AI fault detection analysis cloud platform is minimized, and the obtained target operation fault hidden danger item analysis tasks are more reliable.

In the embodiment of the application, an AI fault detection analysis cloud platform acquires task analysis inhibition elements which are included in a first inhibition matrix block, determines an operation fault hidden danger item analysis task corresponding to the task analysis inhibition elements as a target operation fault hidden danger item analysis task, and determines a first target operation fault hidden danger item which is included in the target operation fault hidden danger item analysis task so as to realize fault hidden danger detection guiding processing of the first target operation fault hidden danger item; and the AI fault detection analysis cloud platform clears the operation fault hidden danger item analysis tasks which are not contained in the first inhibition matrix partition in the operation fault hidden danger item analysis tasks.

Therefore, the AI fault detection analysis cloud platform respectively acquires a plurality of task deterministic parameters of a plurality of operation fault hidden danger item analysis tasks and a plurality of deterministic inhibition coefficients between every two operation fault hidden danger item analysis tasks, generates a task analysis inhibition matrix of the operation fault hidden danger items according to the plurality of task deterministic parameters and the plurality of deterministic inhibition coefficients, clears tasks with false detection and error induction from the plurality of operation fault hidden danger item analysis tasks through the task analysis inhibition matrix, and then determines a target analysis task needing to dynamically monitor the fault hidden danger, so that predictability, accuracy and reliability of dynamic monitoring can be ensured when determining the detection guide of the fault hidden danger, and the unmanned charging pile system is accurately and efficiently subjected to fault detection analysis by utilizing an artificial intelligent technology.

In addition, the embodiment of the application also discloses an AI fault detection and analysis method applied to the charging pile system, and the method can comprise STEP201-STEP206 as follows.

STEP201, through selecting the operation report text of the charging pile system, determines the fault hidden danger trend analysis result corresponding to the first target operation fault hidden danger item and the detection interference trend analysis result corresponding to the fault detection interference item, wherein the fault detection interference item can be the operation item with the highest contact factor with the first target operation fault hidden danger item.

According to the embodiment of the application, after determining the target operation fault hidden danger item analysis task, the AI fault detection analysis cloud platform acquires a first target operation fault hidden danger item in the target operation fault hidden danger item analysis task, determines the first target operation fault hidden danger item and a fault detection interference item most similar to the first target operation fault hidden danger item in the operation report text of the selected charging pile system, and then determines a fault hidden danger trend analysis result of the first target operation fault hidden danger item and a detection interference trend analysis result of the fault detection interference item by utilizing a program capable of realizing single fault hidden danger item processing. Further, the program capable of realizing the processing of the single fault hidden trouble item may be a program composed by means of a single fault hidden trouble item processing network.

In the embodiment of the application, an AI fault detection analysis cloud platform determines a target content set comprising a first target operation fault hidden danger item in an operation report text of a selected charging pile system, and then the AI fault detection analysis cloud platform determines the target operation fault hidden danger item which corresponds to a task cross index of the target content set and meets a preset requirement (such as a discrimination requirement of a text related range) as a fault detection interference item most similar to the first target operation fault hidden danger item.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines a fault hidden danger trend analysis result of the first target operation fault hidden danger item in a later charging pile system operation report text and a detection interference trend analysis result of the fault detection interference item in the later charging pile system operation report text according to a single fault hidden danger item processing program. Further, the procedure of single fault hidden trouble item processing includes residual error network and the like.

STEP202, determining a past hidden danger state expression vector set corresponding to a first target operation fault hidden danger item and a past interference state expression vector set corresponding to a fault detection interference item through selecting a front charging pile system operation report text set before a charging pile system operation report text.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines a first target operation fault hidden danger item and a fault detection interference item most similar to the first target operation fault hidden danger item through selecting a preamble charging pile system operation report text set before a charging pile system operation report text, and then determines a past hidden danger state expression vector set of the first target operation fault hidden danger item and a past interference state expression vector set of the fault detection interference item by combining an electronic locking fault item secondary analysis rule.

In the embodiment of the application, the AI fault detection analysis cloud platform acquires a plurality of continuous groups of report texts before the operation report text of the selected charging pile system as a preface charging pile system operation report text set, and determines a past hidden danger state expression vector set of a first target operation fault hidden danger item and a past interference state expression vector set of a fault detection interference item according to a secondary analysis rule capable of realizing the electronic locking fault item.

In the embodiment of the application, the number of vectors in the past hidden danger state expression vector set and the number of vectors in the past interference state expression vector set are matched with the number of groups of the operation report text set of the preamble charging pile system one by one.

For some examples, the achievable electronic lock-up fault term secondary analysis rules may utilize a network of electronic lock-up fault term secondary analysis rules. Further, the electronic lock-up fault term secondary analysis rule includes a recurrent neural network RNN. For some examples, the number of first target operational fault potential items is a plurality.

STEP203, determining a current fault hidden danger trend locating identifier and a current hidden danger state expression vector corresponding to a second target fault hidden danger item through a later charging pile system operation report text which is related with the selected charging pile system operation report text in the upstream-downstream direction, wherein the second target fault hidden danger item is a target operation fault hidden danger item included in a target operation fault hidden danger item analysis task of the later charging pile system operation report text.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines the second target operation fault hidden danger item, the current fault hidden danger trend locating identification corresponding to the second target operation fault hidden danger item and the current hidden danger state expression vector through the operation report text of the later charging pile system. The first target operation fault hidden danger item and the second target operation fault hidden danger item are at least partially bound, and it can be understood that at least part of auxiliary decision operation matters in the first target operation fault hidden danger item are bound with at least part of auxiliary decision operation matters in the second target operation fault hidden danger item. The operation fault hidden danger items of the second target operation fault hidden danger items are a plurality of.

STEP204, determining the relevance of the charging fault items between the first target operation fault hidden danger item and the second target operation fault hidden danger item through the fault hidden danger trend analysis result, the past hidden danger state expression vector set, the current fault hidden danger trend positioning identification and the current hidden danger state expression vector.

In the embodiment of the application, the AI fault detection analysis cloud platform determines the matching degree of the target task chain through the fault hidden danger trend analysis result and the current fault hidden danger trend positioning mark; the AI fault detection analysis cloud platform determines a hidden danger state expression matching degree set through a past hidden danger state expression vector set and a current hidden danger state expression vector; then, the AI fault detection analysis cloud platform determines the target task chain matching degree and the hidden danger state representation matching degree set as a charging fault item association (a difference of a controller fault item/a circuit fault item/a physical circuit fault item) between the first target operation fault hidden danger item and the second target operation fault hidden danger item.

In the embodiment of the application, an AI fault detection analysis cloud platform carries out matching degree calculation on a fault hidden danger trend analysis result and a current fault hidden danger trend positioning mark to obtain a target task chain matching degree; and the AI fault detection analysis cloud platform carries out matching degree calculation on the past hidden danger state expression vector set and the current hidden danger state expression vector set to obtain a hidden danger state expression matching degree set.

STEP205, through detecting interference trend analysis result, past interference state expression vector set, current fault hidden danger trend locating identification and current hidden danger state expression vector, confirms the interference project relevance that charges.

In the embodiment of the application, the AI fault detection analysis cloud platform determines the task chain matching degree of fault detection interference items by detecting the interference trend analysis result and the current fault hidden danger trend positioning identification; the AI fault detection analysis cloud platform determines the hidden danger state expression matching degree of the fault detection interference item through a past interference state expression vector set and a current hidden danger state expression vector; and then, determining the task chain matching degree of the fault detection interference item and the hidden danger state representation matching degree of the fault detection interference item as the relevance of the charging interference item by the AI fault detection analysis cloud platform.

In the embodiment of the application, the AI fault detection analysis cloud platform carries out matching degree calculation on the detection interference trend analysis result and the current fault hidden danger trend positioning mark to obtain the task chain matching degree of the fault detection interference item; and the AI fault detection analysis cloud platform carries out matching degree calculation on the past interference state expression vector set and the current hidden danger state expression vector to obtain hidden danger state expression matching degree of the fault detection interference item.

Further, the target task chain matching degree is the quotient of the task intersection index and the task merging index of the target content set, and the hidden danger state representation matching degree set is the hidden danger state representation characteristic difference.

It can be understood that the calculation rule of the task chain matching degree of the fault detection interference item is consistent with the calculation rule of the target task chain matching degree, and the calculation rule of the hidden danger state representation matching degree of the fault detection interference item is consistent with the calculation rule of the hidden danger state representation matching degree set.

STEP206, determining a fault hidden danger detection guide of the first target operation fault hidden danger item through the charge fault item association and the charge interference item association.

According to the embodiment of the application, the AI fault detection analysis cloud platform determines fault hidden danger deduction tracing information (associated hidden danger state expression vector) between a first target operation fault hidden danger item and a second target operation fault hidden danger item through the charge fault item association and the charge interference item association; the AI fault detection analysis cloud platform extracts auxiliary decision operation matters which are related to the first target operation fault hidden danger item from the second target operation fault hidden danger item by utilizing the fault hidden danger deduction tracing information so as to determine fault hidden danger detection guidance (used for indicating the fault hidden danger pre-judging detection of the target operation fault hidden danger item) of the first target operation fault hidden danger item.

In the embodiment of the application, an AI fault detection analysis cloud platform transmits the relevance of the charging fault item and the relevance of the charging interference item into a set artificial intelligent network (such as a decision tree model); then, determining a plurality of decision scores of various fault hidden danger deduction tracing information by setting an artificial intelligent network, wherein the various fault hidden danger deduction tracing information can be fault hidden danger combined analysis between a first target operation fault hidden danger item and a second target operation fault hidden danger item, and the obtained fault hidden danger deduction tracing information; the AI fault detection analysis cloud platform determines fault hidden danger deduction tracing information with highest decision score from various fault hidden danger deduction tracing information, and uses the fault hidden danger deduction tracing information as fault hidden danger deduction tracing information.

In the embodiment of the application, an artificial intelligent network is set to generate decision scores among each associated fault hidden danger item in various fault hidden danger deduction tracing information, then the decision scores in each fault hidden danger deduction tracing information are subjected to statistical processing to obtain the decision scores corresponding to the fault hidden danger deduction tracing information, and a plurality of decision scores of various fault hidden danger deduction tracing information are obtained.

In the embodiment of the application, an AI fault detection analysis cloud platform combines with a set fault hidden danger analysis network to perform fault hidden danger combined analysis on a first target operation fault hidden danger item in a selected charging pile system operation report text and a second target operation fault hidden danger item in a later charging pile system operation report text, so as to obtain various fault hidden danger deduction tracing information between the first target operation fault hidden danger item and the second target operation fault hidden danger item. In the embodiment of the application, the set fault hidden danger analysis network can be a classifier.

Further, when the AI fault detection analysis cloud platform determines fault hidden danger deduction tracing information, the AI fault detection analysis cloud platform determines auxiliary decision operation matters related to a second target operation fault hidden danger item in a first target operation fault hidden danger item in the fault hidden danger deduction tracing information, and when the AI fault detection analysis cloud platform determines a third target operation fault hidden danger item which is not related to the second target operation fault hidden danger item in the first target operation fault hidden danger item in the fault hidden danger deduction tracing information, the AI fault detection analysis cloud platform acquires a fault hidden danger trend analysis result through a decision trust value of the third target operation fault hidden danger item, and then the AI fault detection analysis cloud platform determines fault hidden danger detection guidance of the first target operation fault hidden danger item by utilizing the fault hidden danger deduction tracing information and the fault hidden danger trend analysis result.

For example, when the AI fault detection analysis cloud platform determines a third target operation fault hidden danger item which is not related to the second target operation fault hidden danger item in the first target operation fault hidden danger item, the AI fault detection analysis cloud platform judges that the third target operation fault hidden danger item in the operation report text of the selected charging pile system does not appear in the operation report text of the later charging pile system, at this moment, the AI fault detection analysis cloud platform judges that the third target operation fault hidden danger item does not appear in the operation report text of the later charging pile system, and when the decision trust value of the third target operation fault hidden danger item does not accord with the set decision trust threshold, the AI fault detection analysis cloud platform characterizes the third target operation fault hidden danger item to switch out the operation report text of the later charging pile system; when the decision trust value of the third target operation fault hidden danger item accords with the set decision trust threshold, the third target operation fault hidden danger item is characterized to be influenced by the fault detection interference item in the operation report text of the next charging pile system, and at the moment, the AI fault detection analysis cloud platform estimates the task chain condition of the third target operation fault hidden danger item in the operation report text of the next charging pile system through the fault hidden danger trend analysis result corresponding to the third target operation fault hidden danger item.

Further, the AI fault detection analysis cloud platform determines auxiliary decision operation matters related to the first target operation fault hidden danger item in the second target operation fault hidden danger item in the fault hidden danger deduction tracing information, and when the AI fault detection analysis cloud platform determines a fourth target operation fault hidden danger item which is not related to the first target operation fault hidden danger item in the second target operation fault hidden danger item in the fault hidden danger deduction tracing information, the AI fault detection analysis cloud platform adds the fourth target operation fault hidden danger item into a next round of deduction tracing information, wherein the next round of deduction tracing information is deduction tracing information generated by using a next charging pile system operation report text as a selected charging pile system operation report text.

For example, when the AI fault detection analysis cloud platform determines a fourth target operation fault hidden danger item which is not related to the first target operation fault hidden danger item from the second target operation fault hidden danger items, the AI fault detection analysis cloud platform characterizes the fourth target operation fault hidden danger item as an additional target operation fault hidden danger item, and at this time, performs prospective monitoring of fault hidden danger on the fourth target operation fault hidden danger item.

In the embodiment of the application, in the fault hidden danger deduction tracing information, a hidden danger item binary group is formed by the matched target operation fault hidden danger items in the first target operation fault hidden danger item and the second target operation fault hidden danger item, a single hidden danger item is formed by the unmatched target operation fault hidden danger items in the first target operation fault hidden danger item and the second target operation fault hidden danger item, and the AI fault detection analysis cloud platform searches the target operation fault hidden danger items in the second target operation fault hidden danger item from the single hidden danger item to serve as a fourth target operation fault hidden danger item which is irrelevant to the first target operation fault hidden danger item; the AI fault detection analysis cloud platform searches the single hidden danger item for a target operation fault hidden danger item in the first target operation fault hidden danger item to be used as a third target operation fault hidden danger item which is irrelevant to the second target operation fault hidden danger item.

In the embodiment of the application, an AI fault detection analysis cloud platform utilizes a single fault hidden danger item processing program to respectively determine a decision trust value and a fault hidden danger trend analysis result corresponding to a first target operation fault hidden danger item.

In the embodiment of the application, the AI fault detection analysis cloud platform compares the decision trust value corresponding to the third target operation fault hidden danger item with the set decision trust value, and when the decision trust value corresponding to the third target operation fault hidden danger item reaches the set decision trust value, the AI fault detection analysis cloud platform acquires a fault hidden danger trend analysis result.

It can be understood that the program for processing the single fault hidden danger item, the secondary analysis rule of the electronic locking fault item, the setting artificial intelligent network and the setting fault hidden danger analysis network in the embodiment of the application are all adjustable neural networks.

According to the embodiment of the application, the AI fault detection analysis cloud platform deduces the traceability information from the fault hidden danger, determines the real-time fault hidden danger records of different target operation fault hidden danger items in the operation process of the charging pile system, and further can analyze the target operation fault hidden danger items.

In an independent embodiment, after determining a target operation fault hidden danger item analysis task from the plurality of operation fault hidden danger item analysis tasks according to the plurality of task certainty parameters and the certainty suppression coefficient, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task, so as to dynamically monitor the first target operation fault hidden danger item for a fault hidden danger, the method further includes: obtaining a fault hidden danger dynamic monitoring result of the first target operation fault hidden danger item; and carrying out fault early warning on the charging pile system based on the fault hidden danger dynamic monitoring result.

Therefore, the fault early warning of the charging pile system can be accurately and timely carried out by combining the dynamic monitoring result of the fault hidden danger, so that the scheduling processing of fault repair is carried out in advance.

In an independent embodiment, the performing fault early warning of the charging pile system based on the fault hidden danger dynamic monitoring result includes: acquiring a first fault hidden danger development vector and a first operation feedback error reporting vector of the fault hidden danger dynamic monitoring result, and acquiring a second fault hidden danger development vector and a second operation feedback error reporting vector of the historical dynamic monitoring result; determining whether the trend of the dynamic monitoring result of the hidden trouble and the trend of the historical dynamic monitoring result are matched according to the first hidden trouble development vector and the second hidden trouble development vector, and determining whether the fault performance of the dynamic monitoring result of the hidden trouble and the fault performance of the historical dynamic monitoring result are matched according to the first operation feedback error reporting vector and the second operation feedback error reporting vector; and if the trend of the dynamic monitoring result of the fault hidden danger is similar to the trend of the dynamic monitoring result of the historical potential hazard and the fault performance of the dynamic monitoring result of the fault hidden danger is matched with the fault performance of the dynamic monitoring result of the historical potential hazard, carrying out fault early warning on the charging pile system aiming at the dynamic monitoring result of the fault hidden danger based on the fault early warning record of the dynamic monitoring result of the historical potential hazard.

Therefore, the rapid and accurate fault early warning of the charging pile system can be performed by referring to the historical dynamic monitoring result.

Therefore, the AI fault detection analysis cloud platform determines the detection interference trend analysis result of the fault detection interference item through selecting the operation report text of the charging pile system, determines the past interference state expression vector set of the fault detection interference item through selecting the operation report text of the front-end charging pile system before the operation report text of the charging pile system, and fuses the detection interference trend analysis result and the past interference state expression vector set of the fault detection interference item to determine the fault hidden danger detection guide of the first target operation fault hidden danger item in the operation report text of the selected charging pile system, so that when the prospective monitoring of the fault hidden danger is carried out, the negative influence of the fault detection interference item on the prospective monitoring of the fault hidden danger is restrained by utilizing the detection interference trend analysis result and the past interference state expression vector set of the fault detection interference item, and the accurate and efficient fault detection analysis is carried out on the unmanned charging pile system by utilizing an artificial intelligent technology.

According to the same inventive concept, fig. 2 shows a block diagram of an AI fault detection and analysis apparatus applied to a charging pile system according to an embodiment of the present application, and the AI fault detection and analysis apparatus applied to the charging pile system may include a task determination module 21, a parameter determination module 22, and a dynamic monitoring module 23 that implement the relevant method steps shown in fig. 1.

A task determining module 21, configured to determine a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks by using a selected charging pile system operation report text and a preamble charging pile system operation report text associated with the selected charging pile system operation report text in an upstream-downstream direction; wherein, each operation fault hidden trouble item analysis task in the operation fault hidden trouble item analysis tasks corresponds to a task certainty parameter.

And the parameter determining module 22 is configured to determine a deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks in the plurality of operation fault hidden danger item analysis tasks according to the operation report text of the selected charging pile system.

The dynamic monitoring module 23 is configured to determine a target operation fault hidden danger item analysis task from the plurality of operation fault hidden danger item analysis tasks according to the plurality of task certainty parameters and the certainty suppression coefficient, and determine a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task, so as to dynamically monitor the first target operation fault hidden danger item for a fault hidden danger.

The method comprises the steps that a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks are determined through a selected operation report text of the charging stake system and a preamble charging stake system operation report text which is related to the upstream and the downstream of the selected operation report text of the charging stake system, and each operation fault hidden danger item analysis task in the operation fault hidden danger item analysis tasks corresponds to one task certainty parameter; determining a deterministic inhibition coefficient between every two operation fault hidden danger item analysis tasks in a plurality of operation fault hidden danger item analysis tasks by selecting an operation report text of a charging pile system; determining a target operation fault hidden danger item analysis task from a plurality of operation fault hidden danger item analysis tasks through a plurality of task deterministic parameters and deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor the fault hidden danger of the first target operation fault hidden danger item. The AI fault detection analysis cloud platform is used for respectively acquiring a plurality of task certainty parameters of a plurality of operation fault hidden danger item analysis tasks and a plurality of certainty suppression coefficients between every two operation fault hidden danger item analysis tasks, clearing tasks with false detection and error induction from the plurality of operation fault hidden danger item analysis tasks according to the plurality of task certainty parameters and the plurality of certainty suppression coefficients, and then determining a target analysis task needing to dynamically monitor the fault hidden danger, so that predictability, accuracy and reliability of dynamic monitoring can be ensured when determining the detection guide of the fault hidden danger, and the unmanned charging pile system is subjected to accurate and efficient fault detection analysis by utilizing an artificial intelligence technology.

The foregoing is only a specific embodiment of the present application. Variations and alternatives will occur to those skilled in the art based on the detailed description provided herein and are intended to be included within the scope of the application.

Claims

1. An AI fault detection and analysis method applied to a charging pile system is characterized by being applied to an AI fault detection and analysis cloud platform, and comprises the following steps:

determining a target operation fault hidden danger item analysis task from the operation fault hidden danger item analysis tasks according to the task deterministic parameters and the deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor fault hidden danger of the first target operation fault hidden danger item;

The task deterministic parameter is task trust weight, and the deterministic suppression coefficient is a noise coefficient or deterministic clipping coefficient;

the determining a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks through a selected operation report text of the charging pile system and a preamble charging pile system operation report text which is related to the selected operation report text in the upstream and downstream comprises: determining a plurality of operation fault hidden danger item analysis tasks and a plurality of hidden danger analysis decision scores corresponding to the operation fault hidden danger item analysis tasks according to the operation report text of the selected charging pile system, wherein each operation fault hidden danger item analysis task in the operation fault hidden danger item analysis tasks corresponds to one hidden danger analysis decision score; determining a plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks according to the operation report text of the selected charging pile system and the operation report text of the prefronous charging pile system, wherein each operation fault hidden danger item analysis task corresponds to one upstream and downstream viscosity index; determining a plurality of task certainty parameters corresponding to the plurality of operation fault hidden danger item analysis tasks according to the plurality of hidden danger analysis decision scores and the plurality of upstream and downstream viscosity indexes;

Determining a deterministic inhibition coefficient between every two operation fault hidden trouble item analysis tasks in the operation fault hidden trouble item analysis tasks according to the operation report text of the selected charging pile system, including: determining task chain matching degree and hidden danger detail matching degree between every two operation fault hidden danger item analysis tasks in the operation report text of the selected charging pile system; and determining the certainty factor between every two operation fault hidden danger item analysis tasks according to the task chain matching degree and the hidden danger detail matching degree.

2. The method of claim 1, wherein determining a plurality of upstream and downstream viscosity indexes corresponding to the plurality of operation fault hidden danger item analysis tasks according to the selected charging pile system operation report text and the preamble charging pile system operation report text comprises:

3. The method of claim 1, wherein determining a task chain matching degree between every two operational fault hidden trouble item analysis tasks in the selected charging pile system operational report text comprises:

4. The method of claim 1, wherein determining a target operational fault potential term analysis task from the plurality of operational fault potential term analysis tasks based on the plurality of task deterministic parameters and the deterministic suppression coefficient comprises:

5. The method of claim 4, wherein determining a first suppression matrix segment from the at least one suppression matrix segment based on the task deterministic parameter and the deterministic suppression coefficient included in the at least one suppression matrix segment comprises:

6. The method of claim 1, wherein after the determining the first target operational fault term included in the target operational fault term analysis task, the method further comprises:

determining fault hidden danger detection guidance of the first target operation fault hidden danger item according to the charging fault item association and the charging interference item association;

the determining the fault hidden danger detection guide of the first target operation fault hidden danger item according to the charging fault item association and the charging interference item association includes: determining fault hidden danger deduction tracing information between the first target operation fault hidden danger item and the second target operation fault hidden danger item according to the charging fault item association and the charging interference item association; and extracting auxiliary decision operation matters related to the first target operation fault hidden danger item from the second target operation fault hidden danger item by combining the fault hidden danger deduction tracing information so as to determine fault hidden danger detection guidance of the first target operation fault hidden danger item.

7. The AI fault detection and analysis system is characterized by comprising an AI fault detection and analysis cloud platform and a charging pile system which are communicated with each other; the AI fault detection analysis cloud platform is used for: determining a plurality of operation fault hidden danger item analysis tasks and a plurality of task certainty parameters corresponding to the operation fault hidden danger item analysis tasks through a selected operation report text of the charging stake system and a preamble charging stake system operation report text which is related to the upstream and downstream of the selected operation report text; wherein, each operation fault hidden trouble item analysis task in the plurality of operation fault hidden trouble item analysis tasks corresponds to a task certainty parameter; determining a deterministic inhibition coefficient between every two operation fault hidden trouble item analysis tasks in the plurality of operation fault hidden trouble item analysis tasks according to the operation report text of the selected charging pile system; determining a target operation fault hidden danger item analysis task from the operation fault hidden danger item analysis tasks according to the task deterministic parameters and the deterministic inhibition coefficients, and determining a first target operation fault hidden danger item included in the target operation fault hidden danger item analysis task so as to dynamically monitor fault hidden danger of the first target operation fault hidden danger item;

8. An AI fault detection analysis cloud platform, comprising: a memory and a processor; the memory is coupled to the processor; the memory is used for storing computer program codes, and the computer program codes comprise computer instructions; wherein the computer instructions, when executed by the processor, cause the AI fault detection analysis cloud platform to perform the method of any of claims 1-6.