Research on the Fault Diagnosis Method for Hoisting Machinery Based on Multi-source Information Fusion and BPNN

Abstract

In this paper, a fault diagnosis method for hoisting machinery based on multi-source information fusion and BPNN that has a fast training time and a high accuracy rate and can be converted to on-line monitoring system easily is provided. This method can be used to help people to real-time monitoring equipment and components and trace hazards. Compared with traditional methods currently used, the method has higher diagnostic accuracy and wider diagnostic range. Keywords-multi-source information fusion; back propagation neural network; fault diagnosis; hoisting machinery introduction With the advancement of China's modernization process, the hoisting machinery is widely used in infrastructure construction. The hoisting machinery plays an important role in today's society. Once the related accident occurs, it will cause great bodily injury and property damage. Therefore, it is necessary to research the fault diagnosis method for the hoisting machinery to reduce the occurrence of accidents (1). However, the hoisting machinery actually is a kind of special equipment with most danger factors and biggest accident probability. The machine itself, the environment and the operators are potential hazards that may cause huge personal and property risk. Due to the special structure and movement forms, single-sensor cannot guarantee data acquisition normal. Only when multi-sensor information fusion is used, the reliable assessment of the equipment status can be ensured (2). To solve the problem, a fault diagnosis method for the hoisting machinery based on multi-source information fusion is proposed to realize hazard identification and get the most objective and true security information. The specific algorithm uses the back propagation neural network as the basis algorithm and be realized using Matlab. The modeling method and algorithms are detailed in Section 2 and the implementation method is detailed in Section 3. I. METHODS