CN113111659A

CN113111659A - Power inspection work ticket generation method, system, equipment and storage medium

Info

Publication number: CN113111659A
Application number: CN202110395533.4A
Authority: CN
Inventors: 谈元鹏; 焦飞; 欧阳本红; 张中浩; 王芳; 邓显波
Original assignee: State Grid Corp of China SGCC; China Electric Power Research Institute Co Ltd CEPRI
Current assignee: State Grid Corp of China SGCC; China Electric Power Research Institute Co Ltd CEPRI
Priority date: 2021-04-13
Filing date: 2021-04-13
Publication date: 2021-07-13

Abstract

The invention belongs to the technical field of electric power operation inspection, and discloses a method, a system, equipment and a storage medium for generating a power inspection work ticket, which comprises the following steps: acquiring a plurality of work tasks in a work task list and carrying out entity identification to obtain a plurality of entity references; determining target link entities of each work task according to a preset power equipment inspection service knowledge base and a plurality of entity mentions to obtain a plurality of target link entities; obtaining an optimal path among a plurality of target link entities based on a preset power equipment inspection service knowledge graph; and filling the path node information of each path node in the optimal path in a preset work ticket template to obtain the power patrol work ticket. Realize that electric power patrols and examines intelligent generation of work ticket, effectively solved because a ray of maintainer acquires the degree of difficulty to business field knowledge big, and then the maintenance work precision that causes is low, the timeliness is poor scheduling problem, has greatly promoted electric power fortune and has examined business efficiency, reduces electric power fortune and examine the cost.

Description

Power inspection work ticket generation method, system, equipment and storage medium

Technical Field

The invention belongs to the technical field of electric power operation inspection, and relates to a method, a system, equipment and a storage medium for generating a power inspection work ticket.

Background

The electric power inspection work ticket is an important business document for inspecting and overhauling electric power equipment and is a premise for orderly and safely carrying out electric power work. The inspection personnel usually write and form a work task list describing the related power equipment fault and defect state tasks, and then the inspection team leader of the power company divides the work according to the content of the work task list and the duties of the maintainers in the team and carries out the splitting and the deployment of the tasks in the form of hand-copy tickets to form work tickets distributed to people.

At present, power companies already have knowledge bases in the operation and maintenance fields of transmission, transformation and power distribution projects to support inspection personnel to perform inspection, detection and maintenance. However, due to the wide source of power data, the huge data volume, the complex relationship, the manual work change and the like, the manual ticket writing has the problems of heavy task, lack of standardization, difficulty in consulting, easiness in forgetting and the like.

Disclosure of Invention

The invention aims to overcome the defects of heavy task, lack of standardization, difficulty in consulting and easy omission of manual ticket writing in the prior art, and provides a power inspection work ticket generation method, a power inspection work ticket generation system, power inspection work ticket generation equipment and a storage medium.

In order to achieve the purpose, the invention adopts the following technical scheme to realize the purpose:

in a first aspect of the invention, a power inspection work ticket generation method comprises the following steps:

acquiring a plurality of work tasks in a work task list and carrying out entity identification to obtain a plurality of entity references; determining target link entities of each work task according to a preset power equipment inspection service knowledge base and a plurality of entity mentions to obtain a plurality of target link entities; obtaining an optimal path among a plurality of target link entities based on a preset power equipment inspection service knowledge graph; and filling the path node information of each path node in the optimal path in a preset work ticket template to obtain the power patrol work ticket.

The power patrol work ticket generation method is further improved as follows:

the specific method for obtaining the work task list and performing entity identification to obtain a plurality of entities comprises the following steps: and acquiring a work task list, and performing entity identification by adopting a BERT-BilSTM-CRF model to obtain a plurality of entity references.

The specific method for determining the target link entity of each work task in the work task list according to the preset power equipment inspection service knowledge base and the entity mentions comprises the following steps: traversing a plurality of entity mentions, selecting all candidate entities which have affiliated relations with the entity mentions from a preset power equipment inspection service knowledge base, and forming a candidate entity set to obtain a candidate entity set of each entity mention; and traversing each work task, calculating the similarity between the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity, and taking the candidate entity with the highest similarity as the target link entity of the current work task to obtain the target link entity of each work task.

The specific method for calculating the similarity between the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity comprises the following steps: and calculating the similarity of the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity by adopting a BERT-based double encoder.

The specific method for obtaining the optimal path among the target link entities based on the preset power equipment inspection service knowledge graph comprises the following steps: determining a starting target link entity and a terminating target link entity in a plurality of target link entities; determining target nodes corresponding to a plurality of target link entities in a power equipment inspection service knowledge graph; the target node corresponding to the starting target link entity is a starting target node, the target node corresponding to the ending target link entity is an ending target node, and the target nodes corresponding to the other target link entities are intermediate nodes; based on a preset power equipment inspection service knowledge graph, a Dijkstra algorithm is adopted to obtain and determine an optimal path from an initial target node to a termination target node according to a local optimal path from the initial target node to each intermediate node, a local optimal path between each intermediate node and a local optimal path from each intermediate node to the termination target node, and the optimal path is used as an optimal path between a plurality of target link entities.

The specific method for obtaining the local optimal path between the intermediate nodes comprises the following steps: all the intermediate nodes are arranged completely to obtain a plurality of all-arrangement sets; traversing each full-permutation set, taking the first intermediate node of the full-permutation set as a full-permutation starting node, taking the last intermediate node of the full-permutation set as a full-permutation termination starting node, and obtaining a full-permutation optimal path from the full-permutation starting node to the full-permutation termination starting node by adopting a Dijkstra algorithm; and collecting the fully-arranged optimal paths of all the fully-arranged sets to obtain local optimal paths among all the intermediate nodes.

After determining the target nodes corresponding to the target link entities in the inspection service knowledge graph of the power equipment, the method further comprises the following steps: and with the starting target node and the ending target node as a starting node and an ending node, traversing the power equipment inspection service knowledge graph in a depth-first mode, and deleting nodes which are not traversed from the power equipment inspection service knowledge graph.

In a second aspect of the present invention, a power inspection work ticket generating system includes:

the acquisition module is used for acquiring the work task list and carrying out entity identification to obtain a plurality of entity mentions;

the target link entity determining module is used for determining target link entities mentioned by the entities according to a preset power equipment inspection service knowledge base to obtain a plurality of target link entities;

the optimal path determining module is used for obtaining optimal paths among a plurality of target link entities based on a preset power equipment inspection service knowledge graph;

and the work ticket output module is used for filling the path node information of each path node in the optimal path into a preset work ticket template to obtain the electric power inspection work ticket.

In a third aspect of the present invention, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above power patrol work ticket generating method when executing the computer program.

In a fourth aspect of the present invention, a computer-readable storage medium stores a computer program which, when executed by a processor, implements the steps of the power patrol work ticket generating method described above.

Compared with the prior art, the invention has the following beneficial effects:

the method for generating the electric power inspection work ticket extracts a plurality of entity mentions from a work task list through entity identification, determines target link entities mentioned by each entity according to a preset electric power equipment inspection service knowledge base, and then obtains optimal paths among the plurality of target link entities based on a preset electric power equipment inspection service knowledge graph; then, the path node information of each path node in the optimal path is filled in a preset work ticket template, the intelligent generation of the electric power inspection work ticket is realized, the practical production problems of low accuracy, poor timeliness and the like of maintenance work caused by the fact that a line of maintenance personnel has high difficulty in acquiring knowledge in the business field and low firmness in knowledge mastering are effectively solved, the electric power operation and maintenance business efficiency is greatly improved, the labor force of a line of staff is liberated, and the electric power operation and maintenance cost is reduced.

Furthermore, after target nodes corresponding to a plurality of target link entities are determined in the power equipment inspection service knowledge graph, the starting target node and the ending target node are used as a starting node and an ending node, the power equipment inspection service knowledge graph is deeply and preferentially traversed, and nodes which are not traversed are deleted from the power equipment inspection service knowledge graph. Under the condition of not influencing the objective path, the time efficiency of the algorithm is greatly improved by reducing the number of nodes in the power equipment inspection service knowledge graph.

Drawings

FIG. 1 is a block diagram of a process of generating a work ticket for power patrol according to the present invention;

FIG. 2 is a schematic diagram of the logic principle of the power patrol work ticket generation method of the present invention;

FIG. 3 is a schematic diagram of a knowledge graph framework of inspection service for electrical equipment according to the present invention;

fig. 4 is a schematic structural diagram of the power patrol work ticket generation system of the present invention.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

At present, the knowledge graph organizes structured information in a network form, aims to describe concepts, entities, attributes and relationships among the concepts, the entities and the attributes of an objective world by using nodes and edges, has efficient cross-media big data organization, management and cognitive ability, and has superior portability. Due to the difference in coverage and use modes, the knowledge graph is divided into a general knowledge graph and a field knowledge graph, the general knowledge graph mainly comprises a large amount of common knowledge and is high in reusability, the field knowledge graph is constructed facing professional knowledge and has strict and rich data modes, and higher requirements on the depth and accuracy of knowledge are met, so that the application of the field knowledge graph can obtain better system performance and actual effect.

In the actual work of power inspection, the research of current experts and scholars is mainly developed on the basis of knowledge maps in the aspects of intelligent aid decision making and voice assistance, natural language understanding and query statement splicing, power equipment defect retrieval, intelligent question answering and the like, which shows that a large-scale and high-quality knowledge map with scientific structure, clear hierarchy, comprehensive coverage and high correlation is established, so that the professional knowledge is easy to store, fuse, reason, transplant and visualize, and the efficiency of solving the actual problem is improved.

Based on the knowledge, the electric power information in the work task list is borne in the form of the domain knowledge map, is used for knowledge query, knowledge inference and knowledge generation of the work ticket, and has great significance.

The invention is described in further detail below with reference to the accompanying drawings:

referring to fig. 1 and 2, in an embodiment of the present invention, a method for generating a work ticket for power inspection is provided, so as to realize intelligent generation of the work ticket, so as to solve practical production problems of low accuracy of inspection work, poor timeliness and the like caused by difficulty in acquiring knowledge in a business field and low firmness of knowledge mastering by a front-line inspector, and provide technical support for improvement of power operation and inspection business efficiency and liberation of labor force of the front-line employee.

Specifically, the power patrol work ticket generating method comprises the following steps.

S1: and acquiring a plurality of work tasks in the work task list, and performing entity identification to obtain a plurality of entity references.

The method comprises the following steps of obtaining a plurality of work tasks in a work task list and respectively carrying out entity identification, wherein the specific method comprises the following steps: and acquiring a plurality of work tasks in the work task list, and respectively adopting a BERT-BilSTM-CRF model to perform entity identification to obtain a plurality of entity references.

Through entity recognition, all parts, state grades and defect grades of all work tasks in the work task list can be obtained. Furthermore, several entity mentions may be aggregated to obtain an entity mention set M ═ { M1, M2, …, mi }, where i is the number of entity mentions.

Specifically, a BERT-BilSTM-CRF model is adopted for entity identification. Firstly, BIO labeling is carried out on the work content in the work task list, wherein the BIO labeling is to label each Chinese character as 'B-X', 'I-X' or 'O'. Wherein "B-X" indicates that the fragment in which the element is located belongs to X type and the element is at the beginning of the fragment, "I-X" indicates that the fragment in which the element is located belongs to X type and the element is in the middle position of the fragment, and "O" indicates nothing. The job ticket labeling is mainly for the component, part, status level and defect level type of the device, and corresponds to english, which are component, part, status and grade, respectively, as shown in table 1, and a specific labeling example of a certain job content in the job ticket. And after the labeling is finished, dividing the labeled data into a training set, a verification set and a test set, training and testing through a BERT-BilSTM-CRF model, and finally performing entity identification through the tested BERT-BilSTM-CRF model. Wherein the word vectors are encoded using a bert-base-Chinese (Chinese pre-training model). Experiments show that the BERT-BilSTM-CRF model has good effect, and the entity identification accuracy rate reaches 92%.

TABLE 1 concrete example table

Stride with

The more

Climbing device

Ladder with adjustable height

Lack of

Decrease in the thickness of the steel

，

Structure of the organization

Become into

Is free of

Method of

Climbing bar

Climbing device

Is/are as follows

Tightening device

Heavy load

Form of

State of the art

B-part

I-part

O

B-status

I-status

S2: and determining target link entities of each work task according to a preset power equipment inspection service knowledge base and a plurality of entity mentions to obtain a plurality of target link entities.

The method mainly comprises the steps of filtering out most irrelevant entities in the power equipment inspection service knowledge base for entity mention, and selecting a candidate entity which is optimal to the entity mention as a target link entity.

The specific method for determining the target link entity mentioned by each entity comprises the following steps: traversing a plurality of entity mentions, selecting all candidate entities which have affiliated relations with the entity mentions from a preset power equipment inspection service knowledge base, and forming a candidate entity set to obtain a candidate entity set of each entity mention; and traversing each work task, calculating the similarity between the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity, and taking the candidate entity with the highest similarity as the target link entity of the current work task to obtain the target link entity of each work task.

Specifically, firstly, the description of the part or part defect is spoken seriously due to the work content in the work order, for example, the work content in a certain work order: "the task of eliminating the defect: the attention state of QX20180526741, 10kV Wanxu-214, 5# tower and large span tower foot nail distortion (deformation or deformation) is eliminated. In this case, "distortion (deformation) is formulated in a skewed language, and the standard expression thereof is" deformation of the large span peg ", so that the entity identification result of the part or portion in the previous step is required in the generation stage of the candidate entity for the defect of the part or portion.

At this stage, a plurality of candidate entities mentioned by each entity need to be selected from all entities of the preset electric power equipment inspection service knowledge base to construct a candidate entity set for each work task to be disambiguated, so as to reduce the search space of entity links. Specifically, in the candidate entity generation stage, mainly using entity references (components or PART) identified by the entity, in the electric power equipment inspection service knowledge base, that is, a database storing an electric power equipment inspection service knowledge map, DEFECTs OF equipment or PARTs having PART _ detect (location-DEFECT) or detect _ OF (DEFECT-belonging) relationships with the entity references are searched by using the Cypher language, as shown in table 2, a specific example OF a certain PART/location usually has a plurality OF DEFECTs in one PART or location, and the plurality OF DEFECTs constitute a candidate entity set.

Table 2 concrete example table of candidate entity generation stage

And the BERT-based dual encoder performs entity disambiguation by using semantic similarity between the work task in the work task list and the candidate entities in the candidate entity set, and determines corresponding target link entities. Wherein, one work task has only one target link entity. The dual encoder encodes the work tasks and the candidate entities by using two independent BERT encoders, the score of each candidate entity is the dot product of the work tasks and the candidate entity vectors, namely the semantic similarity of the work tasks and the candidate entities, and the higher score is the target link entity.

By the above introduction, the candidate entity set referred by each entity usually includes more than one candidate entity, and specifically, there are generally three cases where an entity in the job task list refers to the corresponding candidate entity.

When the candidate entity | n_iWhen | ═ 0, it is stated that the corresponding target link entity is not found in the power equipment inspection service knowledge base, then the entity refers to m_iConsidered unlinkable, the tag Empty is returned.

When | n_iWhen 1, directly returning the only candidate entity as the target link entity.

When | n_i|>1, ranking the plurality of candidate entities. In this embodiment, the determination of the optimal candidate entity is achieved by ranking and scoring the plurality of candidate entities. Specifically, in the candidate entity ranking stage, similarity scoring and ranking are performed on candidate entities in the candidate entity set and the current entity mention, and since ranking is performed based on semantic similarity probability, the probability is the highest, which means that the confidence is the highest, and the candidate entity can be regarded as the optimal candidate entity, so that the candidate entity with the highest score is taken as the target link entity mentioned by the current entity.

S3: and obtaining an optimal path among the target link entities based on a preset power equipment inspection service knowledge graph.

Referring to fig. 3, a knowledge graph framework of the power equipment inspection service is shown, where the knowledge graph includes knowledge elements related to the work order and relationships between the elements. The inspection service knowledge graph of the power equipment can be constructed in the power system; the method can also be constructed by self, and the method is obtained by constructing from top to bottom according to the business experience of a line of maintainers and the business rules in the standard documents by using the business data such as the overhead transmission line project ledger, the work order, the work ticket, the repair and test record and the like and the maintenance related standard documents.

The power equipment inspection service knowledge graph covers power engineering operation inspection concepts, entities, relations and attributes including information of equipment, components, parts, defects, faults, descriptions, reasons, solutions, fields/stations/lines, units, personnel, ticket holding types and the like.

Each side in the power equipment inspection service knowledge graph is subjected to weight processing in advance, and a mixed weight assignment method based on expert experience and user click rate in an actual scene is adopted in the embodiment. In the initial state of the weight of the power equipment inspection service knowledge map, an initial value is assigned to the weight of the side of the power equipment inspection service knowledge map according to expert experience, then the side weight value is updated according to the click rate of a user or the utilization rate of each node in the power equipment inspection service knowledge map, the more the utilization rate is, the larger the numerical value is after the weight value is updated, and finally the weight value of each side in the power equipment inspection service knowledge map tends to be stable.

Meanwhile, in the embodiment, data in the power equipment inspection service knowledge graph is stored in an adjacent matrix form, and the number in the matrix is the weight value of the edge of the power equipment inspection service knowledge graph.

Among them, Dijkstra's algorithm (Dijkstra's algorithm, which translates Dijkstra's algorithm), proposed in 1956 by netherlands computer scientist ezhel dyxotre, which solved the single-source shortest path problem of weighted directed graphs using breadth-first search, works by preserving the shortest path from source vertex to terminating vertex found so far for each vertex.

Specifically, the implementation of the algorithm comprises the following steps: step 1: and setting up two sets of Y and N, wherein Y is used for storing all nodes waiting for access, and N records all the accessed nodes. Step 2: and accessing a node which is closest to the starting node and is not accessed in the network nodes, and putting the node into Y for waiting for access. And step 3: and finding out the node closest to the starting point from the Y, putting the node into the N, updating the shortest distance from the adjacent node directly connected with the node with edges to the starting node, and adding the adjacent nodes into the Y. And 4, step 4: and repeating the steps 2 and 3 until the Y set is empty and the N set is all nodes in the network, so as to obtain the optimal path between the two nodes.

Specifically, the Dijkstra algorithm can only solve the most basic problem of finding the optimal path between the starting point and the end point, and cannot solve the problem of adding other limiting conditions. In this embodiment, based on the greedy algorithm, the original problem is decomposed into several subproblems that are easy to solve by the Dijkstra algorithm, and local optimal solutions are solved for each subproblem one by one, and then a global optimal solution is solved on the basis. Specifically, a plurality of target link entities are divided into three subsets, namely a starting point set containing a starting point, an intermediate node set containing all intermediate nodes to be passed through and an end point set containing an end point.

And sequentially obtaining a local optimal path from the starting point set to the intermediate node set, a local optimal path from all nodes in the intermediate node set and a local optimal path from the intermediate node set to the end point set by a Dijkstra algorithm, so as to obtain a global optimal path from the starting point to the end point after passing through all specified intermediate nodes. And screening the limited global optimal path to select the global optimal path meeting the requirement.

The specific method for sequentially obtaining the local optimal path from the starting point set to the intermediate node set, the local optimal path from all the nodes in the intermediate node set and the local optimal path from the intermediate node set to the end point set by the Dijkstra algorithm comprises the following steps:

1) assuming that the number of designated intermediate nodes needing to pass through is n, judging the communication condition between the n intermediate nodes, if any two nodes are not communicated, the path meeting the condition does not exist; otherwise, the next step is carried out. The method for judging whether the communication is carried out is as follows: one of the n intermediate nodes is selected as a root node, the power equipment inspection service knowledge graph is traversed in a depth-first mode, and if the other n-1 intermediate nodes are traversed, the n intermediate nodes are communicated; otherwise, it is not connected.

2) The specific method for obtaining the local optimal path between the intermediate nodes comprises the following steps: all the intermediate nodes are arranged completely to obtain a plurality of all-arrangement sets; traversing each full-permutation set, taking the first intermediate node of the full-permutation set as a full-permutation starting node, taking the last intermediate node of the full-permutation set as a full-permutation termination starting node, and obtaining a full-permutation optimal path from the full-permutation starting node to the full-permutation termination starting node by adopting a Dijkstra algorithm; and collecting the fully-arranged optimal paths of all the fully-arranged sets to obtain local optimal paths among all the intermediate nodes.

Specifically, n intermediate nodes are arranged completely, m elements (m is less than or equal to n) are arbitrarily selected from n different elements and are arranged according to a certain sequence, namely, one arrangement of m elements is selected from n different elements. When m is equal to n, all permutation conditions are called full permutation. The starting point of the full alignment is denoted as V1 and the end point is denoted as Vn. An optimal path between V1 and Vn is calculated. In the specific method, for an intermediate node sequence, if any two adjacent nodes have edges which are directly connected, the distance between the two adjacent nodes is the weight value of the corresponding edge, and otherwise, the optimal path between the two adjacent nodes is obtained by using a Dijkstra algorithm. If the path between them passes through a free node, the free node should be saved into the path. Connecting all the neighboring nodes results in the optimal path between V1 and Vn.

3) And (4) solving a local optimal path from the starting target node to V1, namely the optimal path is a single-source-point optimal path, and directly using a Dijkstra algorithm. If the path between the starting target node and V1 passes through the free nodes, the passed free nodes are saved in order into the local path.

4) And solving a local optimal path from the Vn to the termination target node, namely the optimal path is a single-source-point optimal path, and directly using a Dijkstra algorithm. And if the path between the Vn and the termination target node passes through the free nodes, saving the passed free nodes into the local path in sequence.

5) And connecting the 3 paths in sequence to obtain the whole-course optimal path of all the nodes.

S4: and filling the path node information of each path node in the optimal path in a preset work ticket template to obtain the power patrol work ticket.

Specifically, after an optimal path is obtained, node information required by intelligent output of the work ticket is filled in a slot filling type generation technology based on a template, path node information of each path node in the optimal path is filled in a preset sequence and a slot position, and then the power patrol work ticket is obtained and output.

In summary, in the electric power inspection work ticket generating method, the entity identification is used for extracting a plurality of entity mentions from the work task list, the target link entities mentioned by each entity are determined according to the preset electric power equipment inspection service knowledge base, and then the optimal paths among the target link entities are obtained based on the preset electric power equipment inspection service knowledge graph; then, the path node information of each path node in the optimal path is filled in a preset work ticket template, the intelligent generation of the electric power inspection work ticket is realized, the practical production problems of low accuracy, poor timeliness and the like of maintenance work caused by the fact that a line of maintenance personnel has high difficulty in acquiring knowledge in the business field and low firmness in knowledge mastering are effectively solved, the electric power operation and maintenance business efficiency is greatly improved, the labor force of a line of staff is liberated, and the electric power operation and maintenance cost is reduced.

In another embodiment of the present invention, a method for generating a work ticket for power inspection is provided, where the method for generating a work ticket for power inspection includes, in addition to all the contents in the above embodiments, after determining target nodes corresponding to a plurality of target link entities in a knowledge graph of inspection service of power equipment, at least the following steps are also included: and with the starting target node and the ending target node as a starting node and an ending node, traversing the power equipment inspection service knowledge graph in a depth-first mode, and deleting nodes which are not traversed from the power equipment inspection service knowledge graph.

The above steps can be understood as "pruning", that is, nodes which the target path does not pass through are deleted from the original electric power equipment inspection service knowledge graph. The specific method is that starting from a starting point, the power equipment inspection service knowledge graph is traversed in a depth-first mode, nodes which are not traversed can not be found on a target path certainly, and the nodes can be cut off from the original network graph, so that a new power equipment inspection service knowledge graph is generated and replaces the original power equipment inspection service knowledge graph. The purpose of pruning is to greatly improve the time efficiency of the algorithm by reducing the number of nodes in the power equipment inspection service knowledge graph under the condition of not influencing the objective path.

The following are embodiments of the apparatus of the present invention that may be used to perform embodiments of the method of the present invention. For details of non-careless mistakes in the embodiment of the apparatus, please refer to the embodiment of the method of the present invention.

Referring to fig. 4, in a further embodiment of the present invention, a power inspection work ticket generating system is provided, which can be used to implement the above power inspection work ticket generating method.

The acquisition module is used for acquiring a plurality of work tasks in the work task list and carrying out entity identification to obtain a plurality of entity references; the target link entity determining module is used for determining a target link entity of each work task according to a preset power equipment inspection service knowledge base and a plurality of entity mentions to obtain a plurality of target link entities; the optimal path determining module is used for obtaining optimal paths among a plurality of target link entities based on a preset power equipment inspection service knowledge graph; and the work ticket output module is used for filling the path node information of each path node in the optimal path into a preset work ticket template to obtain the electric power inspection work ticket.

In yet another embodiment of the present invention, a computer device is provided that includes a processor and a memory for storing a computer program comprising program instructions, the processor for executing the program instructions stored by the computer storage medium. The Processor may be a Central Processing Unit (CPU), or may be other general-purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf Programmable gate array (FPGA) or other Programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc., which is a computing core and a control core of the terminal, and is specifically adapted to load and execute one or more instructions in a computer storage medium to implement a corresponding method flow or a corresponding function; the processor provided by the embodiment of the invention can be used for the operation of the power inspection work ticket generation method.

In yet another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a Memory device in a computer device and is used for storing programs and data. It is understood that the computer readable storage medium herein can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also, one or more instructions, which may be one or more computer programs (including program code), are stored in the memory space and are adapted to be loaded and executed by the processor. It should be noted that the computer-readable storage medium may be a high-speed RAM memory, or may be a non-volatile memory (non-volatile memory), such as at least one disk memory. One or more instructions stored in the computer-readable storage medium may be loaded and executed by the processor to implement the corresponding steps of the method for generating a work ticket for power patrol in the above-described embodiments.

As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the embodiments of the invention without departing from the spirit and scope of the invention, which is to be covered by the claims.

Claims

1. A power inspection work ticket generation method is characterized by comprising the following steps:

acquiring a plurality of work tasks in a work task list and carrying out entity identification to obtain a plurality of entity references;

determining target link entities of each work task according to a preset power equipment inspection service knowledge base and a plurality of entity mentions to obtain a plurality of target link entities;

obtaining an optimal path among a plurality of target link entities based on a preset power equipment inspection service knowledge graph;

and filling the path node information of each path node in the optimal path in a preset work ticket template to obtain the power patrol work ticket.

2. The power inspection work ticket generating method according to claim 1, wherein the specific method for obtaining the work task list and performing entity identification to obtain a plurality of entities is as follows:

and acquiring a work task list, and performing entity identification by adopting a BERT-BilSTM-CRF model to obtain a plurality of entity references.

3. The power inspection work ticket generating method according to claim 1, wherein the specific method for determining the target link entity of each work task in the work task list according to the preset power equipment inspection business knowledge base and the mention of a plurality of entities comprises the following steps:

traversing a plurality of entity mentions, selecting all candidate entities which have affiliated relations with the entity mentions from a preset power equipment inspection service knowledge base, and forming a candidate entity set to obtain a candidate entity set of each entity mention;

and traversing each work task, calculating the similarity between the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity, and taking the candidate entity with the highest similarity as the target link entity of the current work task to obtain the target link entity of each work task.

4. The power inspection work ticket generating method according to claim 3, wherein the specific method for calculating the similarity between the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity is as follows:

and calculating the similarity of the work task and each candidate entity in the candidate entity set mentioned by the corresponding entity by adopting a BERT-based double encoder.

5. The power inspection work ticket generating method according to claim 1, wherein the specific method for obtaining the optimal path among the plurality of target link entities based on the preset power equipment inspection service knowledge graph is as follows:

determining a starting target link entity and a terminating target link entity in a plurality of target link entities;

determining target nodes corresponding to a plurality of target link entities in a power equipment inspection service knowledge graph; the target node corresponding to the starting target link entity is a starting target node, the target node corresponding to the ending target link entity is an ending target node, and the target nodes corresponding to the other target link entities are intermediate nodes;

based on a preset power equipment inspection service knowledge graph, a Dijkstra algorithm is adopted to obtain and determine an optimal path from an initial target node to a termination target node according to a local optimal path from the initial target node to each intermediate node, a local optimal path between each intermediate node and a local optimal path from each intermediate node to the termination target node, and the optimal path is used as an optimal path between a plurality of target link entities.

6. The power inspection work ticket generating method according to claim 5, wherein the specific method for obtaining the local optimal path between each intermediate node is as follows:

all the intermediate nodes are arranged completely to obtain a plurality of all-arrangement sets;

traversing each full-permutation set, taking the first intermediate node of the full-permutation set as a full-permutation starting node, taking the last intermediate node of the full-permutation set as a full-permutation termination starting node, and obtaining a full-permutation optimal path from the full-permutation starting node to the full-permutation termination starting node by adopting a Dijkstra algorithm;

and collecting the fully-arranged optimal paths of all the fully-arranged sets to obtain local optimal paths among all the intermediate nodes.

7. The power inspection work ticket generating method according to claim 5, wherein after the target nodes corresponding to the target link entities are determined in the power equipment inspection service knowledge graph, the method further comprises the following steps:

and with the starting target node and the ending target node as a starting node and an ending node, traversing the power equipment inspection service knowledge graph in a depth-first mode, and deleting nodes which are not traversed from the power equipment inspection service knowledge graph.

8. The utility model provides a work ticket generation system is patrolled and examined to electric power which characterized in that includes:

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor when executing the computer program implements the steps of the power patrol work ticket generation method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the power patrol work ticket generation method according to any one of claims 1 to 7.