The Research Of Grounding Grids Fault Diagnosis Based On Chemical Reaction Optimization Algorithm

Grounding grids of power plants and substations are a significant part of power systems as they help to maintain the stable operation of power systems and ensure the safety of power workers. They are usually buried in the ground and under terrible environments all the year round. As a result, grounding conductors are prone to corrosion, fracture, and poor electrical condition, which can seriously affect reliable operation of power systems. Therefore, it is essential to find a novel method for fault diagnosis of grounding grids without excavation and cutting off the power, in order to recognize and exclude the fault timely, and ensure safety and reliability of power system.In this thesis, the research status and problems of grounding grid fault diagnosis are summarized. Moreover, the existing fault diagnosis methods for ground grid based on the theory of electric network are analyzed. In addition, the effect of ground lead in measuring process and the corresponding measures to reduce the impact are discussed.On the basis of previous theory, a constrained optimization model of grounding grid fault diagnosis is established. Considering the difficulty of choosing penalty function, interior penalty function is introduced, and then a novel grounding grid fault diagnosis method based on interior penalty function and artificial bee colony algorithm is proposed. Owning to interior penalty function, the method can balance the conflict of minimizing the objective function and meeting the constraints. The penalty function is defined within the feasible region, which helps algorithm to access the feasible solution in the early stage. Simulation results demonstrate the strong diagnostic capacity of this method.Furthermore, chemical reaction algorithm is introduced into grounding grid fault diagnosis. However, the algorithm has defects in convergence rate and accuracy. By implementing elitism strategy and chaos Logistic mapping, the chemical reaction algorithm has been improved. This paper presents an approach for grounding grid fault diagnosis, integrating the theory of electric network, virtual molecular theory, and chemical reaction algorithm, which overcomes the low accuracy of existing methods. In this method, according to the topology structure of grounding grid, the whole network is split into several sub-networks and free branches firstly. Then the multi-objective optimization model of fault diagnosis is established, and finally solved by utilizing chemical reaction optimization algorithm. Simulation results show that the proposed method is efficient and accurate when solving the problem of grounding grid fault diagnosis.