Research on data self-diagnosis technology based on multi-column neural network

Abstract

In the field of data self-diagnosis, traditional particle swarm optimization algorithms tend to fall into local optimal solutions, and the search efficiency is not high. To solve this problem, a multi-swarm particle swarm optimization algorithm based on population relations and repulsion factors is proposed. SRB-PSO. According to the current The search results define three relationships among populations, namely, dominance, reciprocity, and dominance. The repulsion factor is introduced to ensure the diversity of searches between populations, and the search efficiency of the algorithm is improved through the dominance and dominance relationships, thereby improving the global search of the algorithm. The performance is improved while the quality of the solution is improved. The algorithm is compared with several other mainstream particle swarm optimization and improved algorithms on the standard test set. The experimental results prove that the SRB- PSO algorithm can better maintain particle diversity and has strong global search capabilities. The performance when solving multimodal functions is better than that of other mainstream particle swarm optimization and improvement algorithms.