Data mining research in wireless sensor network based on genetic BP algorithm

This paper associated the advantages of genetic and BP algorithm, and studied the application of genetic BP algorithm in wireless sensor network (WSN) data mining. It built a three-layer BP network based on the rapid random search ability of genetic and the mapping capability of BP algorithm, so the redundant history data in WSN could be mined and reused. The tests to 1000 groups of chlorine concentration data from chlorine gas sensor network had been conducted. And according to the comparison between predicted concentration and actual concentration, the maximum relative error was 11.8%, and the maximum average error was 6.96%. So the conclusion could be obtained that the algorithm model established by the paper is feasible and reliable under the condition of no explosive concentration leakage.