Study On Health And Usage Monitoring System For Gearbox Of Military Helicopter

Helicopter dynamic system is the heart of the helicopter whose failure will directly affect the helicopter's flight safety. It is of important significant to carry out dynamic system HUMS research for improving the security of dynamic system. As an important part of dynamic system, gearbox works in the hostile environment the failure of which often leads to serious accidents. Based on this, helicopter gearbox is chosen as a typical component to carry out Health and Usage Monitoring System, so as to lay the foundation for achieving dynamic systems HUMS.To improve fault diagnosis of helicopter gearbox, the intermediate gearbox of some helicopter is taken as diagnostic object. The application of wavelet packet analysis is conducted in-depth study as feature extraction methods. By using the genetic algorithm to optimize BP neural network approach, the accuracy and convergence rate are effectively improved in fault diagnosis of gearbox. On this basis, the helicopter gearbox health and usage monitoring and prototype system is constructed. The detailed contents and innovative work are shown in the following sections.(1) Helicopter gearbox fault type and vibration mechanism is analyzed, including the fault type and vibration mechanism analysis of gear and bearing. And then the fault monitoring methods for the gearbox are presented.(2) The gearbox feature extraction method based on wavelet packet analysis is researched and the validity of the method is verified by experiment.(3) Research based on genetic algorithm optimized BP network gear fault diagnosis method is proposed. By using genetic algorithms to optimize the BP network to overcome the limitations and shortcomings of BP network, the gearbox fault diagnosis model is established based on wavelet packet analysis and genetic BP network. Experimental verification shows that the method has the characteristics of quicker convergence,more stability and higher diagnosis accuracy compared with the traditional method of BP neural network.(4) The helicopter HUMS software architecture is designed including layered architecture for on-board HUMS and OSA-CBM based ground HUMS architecture and the function modules are described in detail. Then a prototype health and usage monitoring system for helicopter gearbox is established and introduced. The system is divided into on-board monitoring system and ground diagnostic system.The thesis research is of great significant for the engineering design of HUMS for helicopter dynamical systems.