Inverse Detecting Faulty Insulators And Optimization Algorithm Research

With the expansion of power grid scale and the development of UHV technology, increasing attention to running safety of power networks has been paid by people, and much stricter safety standard of power networks has been commanded by the state. The preventative detection of insulators became an urgent problem need to solve in view of the importance of insulators in power networks. The method of inverse detecting faulty insulators realizes the detection of faulty insulators by inverse calculation of voltage distribution on the surface of insulators, taking electric field as input data. The analysis of relation between field and source near the insulator string and the research of optimization algorithms are the key problems, the research works are as follows:â‘ The basic principle of simulation charge method and basic concept of electrotechnical inverse problem have been discussed at first. The basic approach of detection method using inverse problem principle has been formed on the basis of analysis of traditional detecting method of insulators and solving method of inverse problems.â‘¡The simulation charge method has been used to establish forward problem calculation model for power frequency electric field of insulators and Genetic Algorithm has been used to optimal allocation of simulation charges, optimized the quantity of simulation charges on the premise of fixed-site. The fast multipole method has been embedded in the calculation process to improve the product efficiency between the vector and matrix. It was also verified the effectiveness and advantage in computational efficiency aspect through the model bar-plane. On the basis of this, the simulation calculation for forward problem of 110kV line insulators has been analyzed. Reliable accuracy and engineering practicability of electric field forward problem can be clearly demonstrated, so the forward calculation for electric field of insulators is practical value for the detection of faulty insulators.â‘¢The basic theory of inverse problem has been discussed, and the causes of ill-conditioning of inverse problem have been analyzed. It was also discussed the condition number method in ill-conditioning analysis. The calculation model and basic equation for inverse problem of insulators has been established by simulation charge method. On the basis of this, Genetic Algorithm, Particle Swarm Optimization and Tikhonov Regularization have been selected to calculate the inverse problem for power frequency electric field of insulators. It was also discussed the basic principle and calculation flow of the three algorithms and comparative analysis.â‘£In view of strong engineering practicability of Genetic Algorithm and the advantages of Tikhonov Regularization in the treating process of ill-posed problem, the combination of the two algorithms has been used to improve previous algorithm. Genetic Algorithm has been used to optimize the regularization parameter. Example calculation shows that this improved algorithm is significantly better stability than the previous algorithm.Though this paper only talks about the insulator string of 110kV, but the principle and the method is also suit to other insulator strings under different voltage level.