Boiler tube leakage location optimization via adaptive Gaussian mutation

Efficient and reliable operation is the main requirement of the modern power plant, hence it´s significant that the boiler tube leakage source is localized precisely. The four-element acoustic array was exploited in the furnace, a set of hyperboloidal equations are established and the sensors distributed for capture the leakage signal is given. In order to avoiding initial guess and distant source assumption, adaptive Guassian mutation operator implemented in GA is employed to give an explicit solution. Comparing GA to quasi-Newton method, it indicates that GA is more robust and accuracy, but cost more time to converge. Time delay estimator error has great influence on localization accuracy, the conclusion that error level is 0.01us,the four-element array localization error is in the permitted range (smaller than 1m) is draw by numerical calculation.