Development And Research Of Fault Diagnosis For Hydraulic System Of TAMROCK Jumbo

In this paper, by research and development of fault diagnosis for hydraulic system of TAMROCK Jumbo Ranger6002, the method of combining neural network with expert system is applied to fault diagnosis for hydraulic system of heavy machinery successfully. The object-oriented process and method of modeling real-time fault diagnosis is discussed in this paper, and the technology of database, neural network and expert system are integrated to program in Visual C++6. 0. The practicable frame of real-time fault diagnosis for heavy machinery is built in this paper, for the meantime, the model of fault database and the model of fault diagnosis are done too. The interface of diagnosis system and user programmed in the operating system of Windows XP, which is based on THC560 hydraulic system object, provides friendly communication for update of fault samples database, training of fault samples and fault diagnosis, eases maintenance of fault samples database and fault diagnosis for users.During the course of compiling fault samples, multi-method of knowledge representation, such as production rule , frame and neural network are synthesized to organize and express the fault samples of THC560 hydraulic system. Before being added to the fault sample database, the sample data are disposed to ensure sample integrity and no redundancy.During the course of developing fault diagnostic method, the influence to the training circle times with network structure, learning rate, initial weight value and door value are discussed. By comprehensive analyses and comparing, the comparatively rational value is adopted to be network' s value. According to imitation, if there are enough samples, this software can diagnosis THC560 hydraulic system fault exactly, and provide fault reason and dissolving methods. It is obvious that the software is helpful for preventing fault occurrence and ensure machinery system running safely and reliably.