Application Of Fuzzy Neural Network In Fault Diagnosis Of Shearer

With science technology rapidly developed, automation is higher, engineering equipment is more complex, reliability and safety of engineering equipment is paid gradually attention to. Large and complex devices can not be made a real-time, accurate diagnosis by conventional methods, which requires the use of intelligent fault diagnosis methods and theories.Mine safety is the basic guarantee of coal resource development. Because coal mine accidents continue to occur. How to improve coal mine safety and reduce the occurrence of coal mine accidents is the main problem that people concern about at present. Troubleshooting of the shearer gradually gians people’s attention. How to ensure the normal operation of Shearer and real-time understanding of Shearer’s work status is currently to be resolved direction.This paper discusses the fuzzy logic systems applied in the diagnosis system of shearer, which is a great progress in coal industry. This paper has detailedly discoursed the basic structure of the fuzzy logic system and analyzed the concept, contents and methods of neural networks in detail. In addition, the intellectual development of the neural network is also discussed. Because of the slow learning speed and poor generalization performance of the traditional gradient algorithm, we have to apply a new learning algorithm named ELM learning algorithm which makes up the differences of generalization performance, and can choose any input weights and thresholds.By establishing fuzzy neural network model, and analyzing Shearer sample failure, according to real-time fault data testing and verifying the model, which proves the system model feasible and effective.With the designing and methods of intelligent fault diagnosis system of the Shearer analysis and discussion. The overall structure of the system and body composition and transmission of data for each module are described, how the system software and hardware achieve is introduced, the diagnosis of the fault simulation and real-time demonstration are made.