Data fusion of wireless sensor network for prognosis and diagnosis of mechanical systems

With the promotion of the latest technologies and the new requirement of humanitarian, the wireless multi-sensor system is applied broadly. This paper studies the data fusion of the industrial wireless sensor networks (IWSNs), in order to acquire more thoughtful data for the prognosis and diagnosis of the monitored device. These authors propose a combination of back propagation neural network (BP NN) and Wavelet Packet algorithm for data fusion. This proposed algorithm is based on each cluster head, which is modelled with a three layers NN. A case study using the ball bearing test data, which is from the Bearing Data Center of the Case Western Reserve University, and to verify the effectiveness of the proposed algorithm. With MATLAB 2016b version, the raw data feature is extracted by the Wavelet Packet and the feature fusion is based on BP NN at sink node. The simulation results show that the proposed algorithm is effective in fault diagnosis of wind turbine.