Fuzzy neural network data mining method based on particle swarm optimization

Abstract

The present invention discloses a fuzzy neural network data mining method based on particle swarm optimization. The method comprises the following steps: cleaning and selecting power data; carrying out enhancement processing on the selected data, and converting the processed data into a form suitable for a neural network model; randomly dividing the obtained data into three data sets: a training data set, a test data set and a confirmation data set, wherein the first two data sets are used to train the neural network, and test the accuracy of the network to construct the neural network model,and the confirmation data set tests the network independently; determining the fuzzy neural network learning algorithm and training the fuzzy neural network; extracting rules from the trained fuzzy neural network; and evaluating the extracted rules. The method is more efficient in mining data in the cockpit of the power system, ensures to obtain optimal power system data mining results, provides decision support for dispatchers and managers, and ensures safe, stable, high-quality and cost-effective running of the power system.

Description

technical field [0001] The invention belongs to the field of power system data analysis and power system monitoring, and relates to a data mining method for an electric cockpit, in particular to a fuzzy neural network data mining method using particle swarm optimization. Background technique [0002] The power system cockpit is proposed under the condition of the rapid development of computer technology, information technology and automation technology and the real-time data in the power system has shown an explosive growth trend. The electric cockpit is a human-computer interaction mode and a "one-stop" decision support system. It simulates the cockpit of a car, an airplane, etc., and the dispatcher can check the key operating indicators of the system according to the real-time operating status of the power grid. Monitoring, early warning, comprehensive analysis, and auxiliary decision-making, visualize the collected power grid data after data mining, and display it intuiti...

AI Technical Summary

Problems solved by technology

Although the BP algorithm provides a basic way of network learning, according to a large number of researches, the BP algorithm also has its essential limitations, that is, the BP algorithm is easy to fall into local convergence, which may lead to prolonged network learning time.

[0008] In view of the deficiencies of existing data mining technology, mainly the problems of network complexity and learning time process in the process of fuzzy neural network data mining, this case arises from this

Images