The Simulation and Fault Intelligent Analysis of Remote Control System
|School||Wuhan University of Technology|
|Keywords||Remote Control System Simulation Fault Diagnosis Artificial Neutral Networks Fuzzy Self-learning|
Marine simulator is using the computer simulation method to display the control casing, physical dish sets and other simulation equipment, simulate the equipment in ship engine room and its function of operational control, as well as train the manipulative skill and evaluate the skill level for operators. Marine simulator is a necessary facility for modern maritime education. The work of the modern marine engine is reappeared faithfully for students, and can do operations, watch, and troubleshooting training. In order to meet the needs of the work in engine room and the development of marine engine technology, the function of marine engine is needed to improve and expand. Among them, the simulation of marine engine pneumatic control system is a subject of great practical value.This thesis is the sub-topic of the marine simulator, the object of study is the pneumatic device of MAM B&W 12K90MC type remote control system in "New-Dalian" container ship, which belongs to China Shipping Corporation, the primary target is to carry out computer simulation and fault diagnosis theory.According to the pneumatic device in MAM B&W 12K90MC type remote control system, this article undertake a detailed quantitative study about the structure and composition of the pneumatic device, as well as the working principle of each function blocks, describe the logic action process of system components (mainly the valves).This thesis analyzes part of the malfunctions about the MAN B&W type pneumatic device. In order to carry on the fault feature extraction in remote control system, data mining will be applied to the fault diagnosis. The fault tree has been established, and qualitative analysis was carried out. The fault diagnosis expert system has been established, it is able to carry out effective reasoning, through some fault phenomenon, the cause of the fault can be inferred.Based on the Artificial Neural Networks, the fault of the remote control system has been simulated and realized, self-learning mechanism has been established, the diagnose knowledge in repository can be optimized, and according to the effectiveness of diagnosis result, the knowledge in repository can be self adapting correction, the accuracy of the results and effectiveness of treatment has been enhanced. Through the self-learning of the system, the required knowledge of diagnostic system can be obtained unceasingly, in the fault fiagnosis expert system, the difficulties of knowledge acquisition has been solved effectively.Using the method of fuzzy self-learning fault diagnosis, the fault diagnosis of Remote Control System can be realized. Using the sample which has appeared, we can carry on the check result, the rate of accuracy is 90%.