Multi-parameter Bayes Fusion Prediction Method for Electrical Performance of Relay
Aiming to the problems of lack information with single condition parameter and redundancy information with condition multi-parameters for the relay electrical performance prediction, this paper presents a novel method for electrical performance prediction of relays based on Bayes phase space reconstruction with condition multi-parameters.After reconstructing the relay locus for each condition parameter in the same high dimension state space, the proposed method optically fuses the high dimension phase points according to the Bayes theory, which makes the fused locus approximate more closely the real condition state locus.Then the proposed method obtained the prediction model by using the extreme neural network to fit the fused locus for its fast learning ability.At last, the electrical performance is predicted with contact resistance and arc energy parameters as an example, and the prediction model parameters are analyzed for the impacts on the prediction effects.The experimental results show that the proposed method is feasible and effective.
relay Bayes theory phase space reconstruction prediction
Li Chen Shengxue Tang Jingqin Wang
Province-Ministry Joint Key Laboratory of Electromagnetic Field and Electrical Apparatus Reliability Hebei University of Technology Tianjin, China
国际会议
温州
英文
402-406
2014-11-16（万方平台首次上网日期，不代表论文的发表时间）