Data Fusion Using Chaotic Immune Wavelet Networks for Dissolved Gases Analysis of Power Transformer Oil

摘要

In this paper, a new multi-sensor array data fusion technique based on the immune wavelet networks is presented to compensate the effect of poor selectivity of semiconductor gas sensor. The initial structure of the wavelet networks is derived from the training set, and those redundant wavelet basis functions are reduced from the wavelet framework by stepwise orthogonal selection method. Then the network is optimized by the artificial immune algorithm. During the optimization process, a new self-adaptive chaos mutation operator is used to enhance the searching ability. The experimental results show that this method is practical and effective.