篇名 | Application of Improved PSO-ELM Algorithm in Optimizing the Path of Robot |
---|---|
卷期 | 29:4 |
作者 | Hong-Ge Ren 、 Rui Yin 、 Tao Shi 、 Fu-Jin Li |
頁次 | 031-038 |
關鍵字 | extreme learning machine 、 higher accuracy 、 improved particle swarm optimization 、 inertia and learning factors 、 optimized path 、 EI 、 MEDLINE 、 Scopus |
出刊日期 | 201808 |
DOI | 10.3966/199115992018082904003 |
In view of the problem of poor stability, large errors in optimizing the parameters of extreme learning machine network (ELM network) by traditional particle swarm optimization algorithm (PSO algorithm), an improved PSO-ELM algorithm (IPSO-ELM algorithm) was proposed in this paper. This algorithm is designed to adjust the inertia factor and learning factors of the PSO algorithm. It chooses the appropriate learning factors and the dynamic inertia factor to improve the optimization performance of PSO algorithm. The core of IPSO-ELM algorithm is to improve the certainty of initial weights and thresholds that belonged to ELM neural network and then train the simples by using ELM neural network for enhancing the generalization ability and stability of system. The simulation experimental results show that the proposed IPSO-ELM algorithm outperforms other similar algorithms with faster convergence speed, better robustness, lower errors, and higher accuracy.