Electric power system fault diagnosis method based on improved particle swarm optimization algorithm

Abstract

The invention discloses an electric power system fault diagnosis method based on an improved particle swarm optimization algorithm. The particle swarm optimization algorithm is initialized to a group of random particles, then an optimal solution is found through iteration, and in the iteration of each time, the particles update themselves by tracking two "extreme values", wherein the first extreme value is an optimal solution that can be found by the particles themselves and is called an individual extreme value pbest, and the other extreme value is an optimal solution which is currently found by the whole population and is called a global extreme value gbest. An evolution speed factor H is introduced in the iteration process of the improved particle swarm optimization algorithm, when a minimum value is sought, the H is made a global optimal value of last iteration / a global optimal value of current iteration, the evolution speeds of the particles are calculated, and when the H is lower than delta, a negative disturbance item is added to a speed updating process so as to change the motion direction of the particles. Compared to a conventional genetic algorithm, the particle swarm optimization algorithm has the advantages of high stability, good convergence characteristics and fast operation speed.

Description

technical field [0001] The invention relates to a power detection, in particular to a power system fault diagnosis method based on an improved particle swarm optimization algorithm. Background technique [0002] The scale, capacity and coverage of modern power grids are getting bigger and bigger, and the power system fault diagnosis system provides a strong guarantee for the reliability and continuity of power supply. The power system fault diagnosis system uses the collected information to identify faulty components and malfunctioning protection and circuit breakers as much as possible, and the identification of faulty components is the key issue. Therefore, research work at home and abroad mainly focuses on the identification of faulty components. At present, the methods of fault diagnosis include logic processing method, expert system method (expert system method based on Petri net), artificial neural network method and method based on optimization technology, etc. Amon...

AI Technical Summary

Problems solved by technology

Among these methods, the expert system method is intuitive and has strong explanatory ability, but it is difficult to obtain a complete knowledge base, has no learning ability, and has poor fault tolerance

The performance of the fault diagnosis of the artificial neural network method depends on whether the sample set is complete. It is extremely difficult to form a complete sample set for large-scale power systems, so the correctness of the diagnosis results cannot be guaranteed in principle.

In addition, for the situation that different fault elements cause the same protection and circuit breaker action, this method can only give one of the solutions, which has great limitations

The general fuzzy system adopts a structure similar to the expert system, so it also has some inherent advantages and disadvantages of the expert system, but increases the fault tolerance

Genetic Algorithm (GA) can basically solve the problem of fault diagnosis from the perspective of optimization, especially in the case of multiple faults or the existence of protection and malfunction of circuit breakers, it can give the global optimal or local optimal multiple possible diagnostic results, but poor stability and slow speed

Images