Research On Expert System Of The Hydraulic System Fault Diagnosis Of Offshore Crane

The location of the offshore platform is generally far from land, if a hydraulic fault cannot be removed in time, a downtime of the facility has a direct impact on the production activities of the offshore platform and causes great economic losses.The paper studied the hydraulic system principle of offshore platform crane. Combined with the characteristics of this hydraulic system, a summary of common faults and key components diagnosis in hydraulic system was included. Base on TS model, a fuzzy fault tree of the hydraulic system is built, and the fault probability of each component in hydraulic system is estimated.The pressure and flow curves under load and no-load condition of the lifting hydraulic system were achieved via test. The parameter in the simulation model was determined according to the relative parameter of the hydraulic system, pressure and flow curves can be simulated from the work. The reliability of the model was proved by comparing the two sorts of curves. Adding fault information into the simulation model, the system was applied with fault analysis, which provided curves of main parameter in the system under the condition of pump and valve leakage and concluded diagnosis rule. The diagnosis information of the fault analysis can offer theoretical base for hydraulic fault diagnosis.The fault diagnosis expert system combined expert system and neural network together organically, the characteristic data under normal and abnormal condition which was obtained by AMESim simulation trained the neural net and made it the inference mechanism of the expert system. The expert system can be more adaptable by changing the control parameter of AMESim according to the scene. Weighting factor based on fuzzy mathematics to establish the fault decision matrix combined with the right to determine the fault search matrix, in order to determine the fault search order to reduce the unnecessary removal of troubleshooting procedures, fault diagnosis of other types of search strategies and decision-making has certain significance.In the process of fault diagnosis, the data of the hydraulic system needed to be detected and collected. The offshore crane fault diagnosis expert system mixed the neural network and expert system together. The neural network rule can be upgraded by using Matlab tool function to train the diagnosis data base, according to the fault simulated by AMESim. With the help of Visual Basic, Matlab function can be invoked for online fault diagnosis, and the result was added to the data base.This study is based on offshore crane hydraulic system fault diagnosis project, which is cooperated with Bohai Equipment Service Cooperation subordinated to China National Offshore Oil Corporation (CNOOC). The project was aiming at timely solving the hydraulic failure of the offshore platform crane which is far from land. Monitoring system with the function of real-time detection, combined with neural network was added to the fault diagnosis expert system, which operated practical, convenient and reliable. The system will be also applied to other types of cranes for offshore platform, which has a certain guiding significance for the expert system of other hydraulic system.