Research on multi-sensor information fusion algorithm of fan detection robot based on improved BP neural network

Abstract

The functional structure of the fan operating system is complex, and the condition detection of a single signal source will inevitably result in errors and false alarms. The diagnostic method of information fusion can make full use of more information. Thus the problem can be avoided: the Fan Detection robot is equipped with multiple sensors, and these sensors are fused by a reasonable information fusion algorithm. The fused sensor information can obtain a more accurate statement of the fan. The purpose of saving cost and making the fan run stably aims to reduce man-made overhauls. BP neural network has the capability of non-linear mapping, self-learning and self-adaptation, and the fusion performance is good, the application of a wide range. Therefore, BP Neural Network algorithm can be chosen to carry out sensor information fusion.