Road Construction Machinery Fault Diagnosis Expert System Based On Fuzzy Neural Network

Road construction is a collaborative process involving various large machineries such as asphalt paver, road roller, conveyor and so on. It is vital to ensure all construction machineries can work normally in poor working conditions. With the new technology of "convey-pave", asphalt paver and conveyor become core machineries of the whole construction machinery group. To monitor, diagnose and handle the faults of these two kinds of machineries in time will guarantee construction progress. Consequently, doing deep research on fault diagnosis of road construction machineries makes an important practical sense.By deeply investigating a lot of intelligent fault diagnosing methods and theories, this thesis generalizes fault diagnosis methods based on fuzzy theory and neural network. In order to make up for each other's deficiencies, fuzzy theory and neural network are combined to build a fault diagnosis model based on fuzzy neural network.Aiming at fault diagnosing requirement analysis of asphalt and conveyor these two important road construction machineries, a road construction machinery fault diagnosis expert system, which embeds multiple reasoning mechanisms, is designed and implemented. This expert system is composed of diagnosing system and diagnosis information base management system. According to different fault symptoms, diagnosing system will identify fault causes either by fuzzy associate memory neural network based reasoning or by hybrid rule-based reasoning, and provide corresponding processing advice. Knowledge is the motive power of the expert system, so diagnosis information base management system will acquire and improve the knowledge both by directly absorbing field experts' experience and by doing self-learning based on neural network.