Research On Fault Diagnosis Of All Hydraulic Drilling Machine Based On RS-SVM Data Fusion

Drilling rig is a complex control system of mechanical and electrical integration. Full hydraulic drilling rig in operation process prone to spalling, wear, loosing, fatigue, impact, the leakage failure, when fault occurs will lead to changes in some physical parameters of drilling rig, such as dynamic signal non stationarity, the pressure in the hydraulic circuit and flow fluctuation, because of this, no matter what kind of failure, and the physical parameters of the system has a direct or indirect contact, as long as the master of the characteristic information of the rig monitoring parameters, will be able to achieve fault diagnosis of drilling rig.For full hydraulic tunnel drilling rig hydraulic system and deceleration gear fault, from mechanical and electrical equipment of the real-time condition monitoring and fault diagnosis two important directions of research design and research of the fault diagnosis of drilling rig. Drilling rig failure mechanism, MEMS vibration measurement node, multi-sensor data fusion method and drilling rig fault simulation experiments, and other four aspects.Firstly, the failure mechanism of the fully hydraulic drill is studied. Multi sensor fault signal acquisition scheme. Aiming at the vibration measurement of the rotating machinery monitoring difficult area, a wireless vibration measurement node based on MEMS accelerometer is developed to realize the hardware and software design of the node. According to the characteristics of the fault of the rig, through the study of wavelet packet energy eigenvector extraction, rough set attribute reduction and support vector machine(SVM) and other theories, and RS-SVM based on multi sensor data fusion method and its application in the fault diagnosis of drilling rig, improve the accuracy of the drilling machine fault diagnosis and classification is proposed. Finally, the vibration calibration test bench is set up, and the calibration experiment is carried out. And using broken gear, throttle valve leakage branch of the reducer gear reducer gear tooth, hydraulic Madane leakage fault, drilling rig fault simulation experiment.To sum up, the research of the vibration real-time condition monitoring node and multi sensor data fusion fault diagnosis method has a certain reference and application value.