Research And Implementation Of Non-Intrusive Identification Method Based On Transient Characteristics

With the rapid development of the global economy today,the demand of energy for every country is constantly increasing,thus resulting in environmental pollution becoming nore severe.In order to ensure the friendly and harmoniously development of resources and environment,and supporting the long-term and effective development of the national economy,one of the most effective methods on the rational use of electric energy is Non-intrusive load identification.It measures and analyzes load characteristics such as voltage,current,and harmonics at the entrance of the bus for energy usage information of the device.The acquisition of electric energy load information not only helps the electric power company to understand the composition of the user load,analyzes the usage rules of the equipment,evaluates the equipment operating modality,and then strengthens the load side management,but also guiding the user to reasonably arrange the use of the load,improving the electric energy use and achieve the purpose of adjusting the peak-to-valley difference and reducing network loss effectively.Non-intrusive load identification method has a wide prospect,it not only serves as a key technology for the global energy Internet,with the increasing demand for electrical energy,environmental issues are gradually deteriorating,the public's awareness of environmental protection and energy conservation is further strengthened,but also plays an extremely important role on real time monitoring of the state and mode of electrical equipment.In this paper,based on the actual needs of the energy Internet project,the research on a non-intrusive load identification system based on transient information is developed for the non-intrusive identification of electrical equipment in operational mode edge detection and load status identification.Emphasis was placed on improving the edge detection method and completing the design of the entire system.The work done in this paper mainly includes:(1)Analyze and study the related technologies of non-intrusive load identification,including load denoising technology,load transient feature matching technology,and edge detection technology of electrical equipment.(2)Conduct the deeply study of the overall framework and various modules of a transient-based non-intrusive load identification system.It focuses on the establishment of feature database,transient feature matching,denoising filtering,and edge detection algorithm.Combining the transient characteristics of load variations,providing the experimental basis for effective management and monitoring on user-side load.(3)The non-intrusive load identification method was completed.Firstly,the overall architecture platform of the system was built under the RT-LAB environment,and then realizing the functions of each module.The great importance is attached upon edge detection and feature matching algorithms.Aiming at the problem of low accuracy of the load edge detection method in non-intrusive identification of current user-side power equipment,an edge detection method based on Gaussian filter and industrial detection CUSUM(cumulative sum)algorithm and a feature matching method based on DTW algorithm are proposed.Combining the senstive detection on variations of state and mode in CUSUM algorithm and the high stability of the DTW algorithm,the effectiveness and accuracy of the equipment modality are improved.By establishing a non-intrusive load identification platform,the proposed method is simulated and verified by experiments,which shows that the proposed method can effectively improve the accuracy of electrical equipment.