In order to overcome the shortcomings of traditional back propagation (BP) and single genetic algorithm (GA), a method based on quantum GA (QGA) is proposed to optimize the BP neural network for fault detection of liquid rocket engines. In this QGA-BP method, a dynamic improvement strategy is adopted to adjust the rotation angle according to the evolution situation, and a quantum catastrophe strategy is used as an operation criterion during evolution. Then, the improved QGA is used to optimize the weight and threshold of the BP neural network from multiple spots. This method is applied to a typical fault detection process of a liquid rocket engine. Representative history test data of engine state is used to verify this method, and the results show that the convergence speed, the evolution generation, and the accuracy of fault detection of the QGA-BP model are all improved compared with the traditional BP neural network and the single GA.

Engines; Sociology; Statistics; Fault detection; Convergence; Quantum computing; Biological cells; Back propagation (BP); fault detection; liquid rocket engines; neural network; quantum genetic algorithm (QGA)

TIANJIN UNIVERSITY OF TECHNOLOGY