CN106228246A - Based on semantic unattended duty transformer substation monitoring system and method - Google Patents

Abstract

The invention discloses based on semantic unattended duty transformer substation monitoring system and method, based on semantic reasoning technology, the data of multi-Substation carry out convergence analysis, rule of thumb the semantic rule of model extraction sets up Ontology, trains semantic rule storehouse.Carried out the semantic reasoning of data by expert knowledge library and semantic rule storehouse simultaneously, can accurately analyze and predict power equipment situation, immediately give power equipment aging early warning, and continuous complete semantic rule storehouse during data collection so that early warning precision improves constantly.

Description

Semantic-based unattended substation monitoring system and method

technical field

The invention belongs to the field of substation monitoring systems, in particular to a semantic-based unattended substation monitoring system and method.

Background technique

In May 2009, State Grid proposed the development goal of building a strong smart grid. As of 2015, 5,182 new smart substations have been built, and the development of substations has gradually begun to move from analog circuit protection to today's digital intelligence. Compared with traditional substations, which have repetitive collection resources and complex designs, and non-standard and non-standard information, smart substations use advanced, reliable, integrated, low-carbon, and environmentally-friendly smart equipment to digitize the entire station's information, network the communication platform, and standardize information sharing. To meet basic requirements, it can automatically complete basic functions such as information collection, measurement, control, protection, metering and monitoring, and can support advanced kinetic energy such as real-time automatic control, intelligent adjustment, online analysis and decision-making, and collaborative interaction of the power grid as needed, and realize the integration with adjacent substations. , power grid dispatching and other interactive substations. The monitoring methods of substations are becoming more and more diverse, and unmanned duty has become one of the hot research topics, which greatly reduces the danger of artificial on-site monitoring. However, in unattended smart substations, how to break through the information islands between substations and realize comprehensive analysis based on big data has not been resolved.

Contents of the invention

Purpose of the invention: The present invention proposes a semantic-based unattended substation monitoring expert system for the existing unattended detection system, and uses semantic technology to realize reasoning to obtain more accurate equipment status prediction.

SUMMARY OF THE INVENTION: A semantic-based unattended substation monitoring method includes the following steps:

Step 1: Collect the status data of the power equipment through the power equipment monitor, and save the data according to the attributes of different equipment;

Step 2: Determine the scope and objects of the domain ontology through the power equipment status warning knowledge base, and select the ontology formal description language to describe and represent;

Step 3: Based on the information provided by the domain expert database, establish an empirical model of ontology relationship for power equipment domain state prediction;

Step 4: Establish an OWL-based ontology model in the field of electric equipment based on the ontology relationship empirical model predicted in step 3;

Step 5: According to the inference rules provided by the prediction conditions and the ontology relationship empirical model, use the inference engine engine composed of the Jena inference engine to perform semantic reasoning and obtain the prediction results;

Step 6: Connect to the database driver, create a database connection instance, and connect to the database through a servlet;

Step 7: The mobile client accesses the servlet, and converts the obtained data into Json data format in the servlet and returns it to the mobile client.

A semantics-based unattended substation monitoring system, including a mobile client, a query parser, a power equipment status warning knowledge base, an ontology library, and an inference engine engine, and the power equipment status warning knowledge base is used to store power data of power equipment and its The prediction results of parsing and reasoning; the query parser is used to analyze the structure of electric power data of electric equipment and extract valid data from it; the ontology library is used to store the ontology model of electric equipment status and related instance data; reasoning and drawing conclusions.

Beneficial effects: based on the semantic reasoning technology, the present invention integrates and analyzes the data of multiple substations, establishes a semantic ontology according to the semantic rules extracted from the empirical model, and trains a semantic rule base. At the same time, the semantic reasoning of data is carried out through the expert knowledge base and semantic rule base, which can accurately analyze and predict the status of power equipment, give early warning of power equipment aging, and continuously complete the semantic rule base in the process of data collection, so that the accuracy of early warning can be improved. keep improving. Compared with the prior art, the present invention has the following advantages:

(1) The present invention is compatible with existing SCADA data, adopts the underlying data acquisition framework of multiple substations, breaks the original data information islands of each substation, and realizes information interconnection and sharing;

(2) The present invention combines semantics and reasoning technology, which can realize the fusion of multi-source heterogeneous data, and can solve the accuracy problem of equipment monitoring in unmanned duty, and provide accurate data for the data semantic reasoning prediction of electric power equipment Prediction, realizing early warning of equipment aging in unattended substations;

(3) The present invention is based on the monitoring expert system of multiple substations. Compared with the existing stand-alone monitoring systems such as SCADA, the utilization rate of data can be greatly improved, and one system can be used on multiple platforms, thereby obtaining more effective data;

(4) The equipment state prediction of the present invention adopts semantic technology, establishes an expert experience model, converts the model into semantic rules, and based on the semantic rule base trained by multiple substations, as the amount of data continues to increase, the rule base will become more and more complete , the data prediction becomes more and more accurate;

(5) The visual display of the Android interface of the present invention enables users to conveniently inquire and extract data anytime and anywhere, and provides a more convenient handheld management platform for substation managers.

Description of drawings

Fig. 1 is a schematic diagram of the system of the present invention;

Fig. 2 is a frame diagram of establishing a semantic ontology of the present invention;

Fig. 3 is a frame diagram of the present invention.

detailed description

Embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings;

As shown in Figure 1, the entire framework of the semantic-based unattended substation monitoring expert system consists of an Android mobile client, a query parser, a power equipment status warning knowledge base, an ontology database, and an inference engine.

The power equipment status early warning knowledge base stores common power data and the prediction results of its analysis and reasoning; the query parser is mainly responsible for analyzing the structure of the power data of power equipment and extracting effective data from it; the ontology library is mainly used to store the power equipment status ontology Models and associated instance data. The reasoning engine includes Jena reasoning engine and other reasoning engines, which are responsible for information reasoning on the ontology and drawing prediction conclusions.

As shown in FIG. 2 , the semantic ontology establishment framework diagram of the present invention. The establishment of the ontology adopts a seven-step method, (1) determine the professional field and category of the semantic ontology (2) consider the possibility of reusing the existing semantic ontology (3) list the important terms in the semantic ontology (4) Define the class (Class) and the class hierarchy (Hierarchy) (5) define the attributes of the class (6) define the restrictions on the attributes (7) create an instance. Finally, use the prolog language to write the inference rules, and then use the SPARQL query language to verify the rationality of the ontology establishment.

As shown in Figure 3, the frame diagram of the present invention. The databases currently supported by Jena are PostgreSQL, Mysql and Oracle. First, connect to the database driver and create a DBConnection instance. After the ontology file is persistently stored in the database, it can be inferred and queried. The main prediction rules and the inference rules defined by the expert experience knowledge base borrow the Jena toolkit for semantic reasoning to obtain the prediction results, and then output the client display interface.

Proceed as follows:

Step 1.: Collect relevant power equipment status data through the power equipment monitor, and save data according to the attributes of different equipment;

Step 2.: Determine the scope and objects of the domain ontology, select the ontology formal description language to describe and represent;

Step 3: Based on the data and experience provided by the domain expert database, establish an empirical model of ontology relationship for power equipment domain state prediction;

Step 4: Establish an OWL (Web Ontology Language, Network Ontology Language) ontology model in the field of electric equipment according to the empirical model in step 3;

Step 5: According to the prediction conditions and the inference rules provided by the empirical model, use the Jena tool (Jena is a java API, used to support related applications of the Semantic Web) package to perform semantic reasoning to obtain the prediction result;

Step 6: connect to the database driver, create a database connection instance, and use servlet (the full name of servlet is JavaServlet, which is a server-side program written in Java) technology to connect to the database;

Step 7: Design the Android client, access the servlet, and convert the obtained data into Json data format in the servlet and return it to the Android client display interface.

The invention introduces the semantic big data reasoning technology into the substation system. Using meta-model-based semantic technology and implementing an expert system centered on the expert knowledge base, based on multi-substation data sharing, the huge data returned by power equipment in different substations is analyzed according to the rules, and the accurate equipment status is obtained through the expert system Analysis, improve the utilization efficiency of data, improve the accuracy and precision of equipment status prediction, better locate a certain problem or equipment that is about to have a problem, reduce the difficulty of traditional manual investigation, and also reduce the cost of parts replacement , providing better support for the realization of substation monitoring centered on unmanned duty.

Claims (2)

1. a semantic-based unattended substation monitoring method, is characterized in that, comprises the steps: Step 1: Collect the status data of the power equipment through the power equipment monitor, and save the data according to the attributes of different equipment; Step 2: Determine the scope and objects of the domain ontology through the power equipment status warning knowledge base, and select the ontology formal description language to describe and represent; Step 3: Based on the information provided by the domain expert database, establish an empirical model of ontology relationship for power equipment domain state prediction; Step 4: Establish an OWL-based ontology model in the field of electric equipment based on the ontology relationship empirical model predicted in step 3; Step 5: According to the inference rules provided by the prediction conditions and the ontology relationship empirical model, use the inference engine engine composed of the Jena inference engine to perform semantic reasoning and obtain the prediction results; Step 6: Connect to the database driver, create a database connection instance, and connect to the database through a servlet; Step 7: The mobile client accesses the servlet, and converts the obtained data into Json data format in the servlet and returns it to the mobile client. 2. A semantic-based unattended substation monitoring system, characterized in that it includes a mobile client, a query parser, a knowledge base for early warning of power equipment status, an ontology library, and an inference engine, and the knowledge base for early warning of power equipment status is used to store The power data of power equipment and its prediction results of analysis and reasoning; the query parser is used to analyze the structure of power data of power equipment and extract valid data from it; the ontology library is used to store the ontology model of power equipment state and related instance data; the inference engine It is used to carry out information reasoning on the ontology and draw prediction conclusions.