Research On Power Behavior Monitoring Technology Based On Neural Network

Not only monitoring and statistics the use of electricity can provide users with an effective security alerts,but also supply the governments,manufacturers of electrical appliances reliable data for analysis.Research on electrical behavior,to establish an effective scheme and algorithm for electricity behavior data acquisition,accurate identification and information monitoring,has practical and long-term significance to promote the safety of electrical consumption and the industrial innovation.The research on the technology of power load monitoring is studied,and a monitoring system based on BP neural network is proposed.First of all,study the common techniques in load identification,and analyzed the load identification technique based on load power characteristics,current-voltage track method,current transient characteristics and steady state characteristics,then propose a monitoring method based on neural network after study the characteristics of neural network model and the idea of BP algorithm in detail.Analyzed further the basic model characteristic of the load,studied the extractive technique of electricity behavior data and the pretreatment technology of electricity data characteristics,and discussed specific implement method that maintain current harmonics as input layer data of neural network to match the load category and after that by the means of Euclidean distance algorithm proceed load accurately identify to acquire the electricity behavior data.In addition,the system is implemented and tested using Web Java related technologies.In this paper,the following aspects are analysis and researched:(1)the feature of different kinds of load models in the circuit and the non-intrusive extraction and the feature extraction of the current harmonic data are studied.(2)Obtain the characteristics of load by analyzing the model of different kinds of load,and discuss the feasibility of using current harmonic characteristics for different types of load.(3)The process of applying BP neural network theory to the monitoring of electrical behavior is analyzed and studied.The current harmonic data is used as input samples of neural network to train the load class classification system,and the performance of BP neural network is verified by MATLAB simulation technology.(4)Research on the concrete technology of the electric behavior monitoring system implementation.The whole application platform using Java and related technologies to implementation,including data acquisition terminal technology,reliable data transmission,data storage,electric behavior recognition system and data monitoring platform.Through the electricity behavior monitoring system based on neural network performance in the living area of University,the application of the system promote the management of the Department of intelligent appliances supervision,improve the timeliness and accuracy of the information,which has a certain reference value for further research and application of electrical behavior.