Fault Diagnosis Using DSP And Information Fusion For Shearer

The subject comes from major science and technology projects in Shanxi Province"coal mining equipment running real-time monitoring of the health warning and faultmanagement system."Mining machinery and equipment due to manufacturing errors and improperassembly of the working conditions (such as the load is too large, poor lubrication),making it easy to malfunction. Shearer equipment failure had a mechanical failure,electrical failure and hydraulic failure, but failure mostly in mechanical failure, andmechanical failure focused on transmission parts, so troubleshooting shearer cuttingthe Department of great significance. We establish a DSP platform fault diagnosissystem to achieve real-time detection of Shearer and diagnostic functions, includingfault vibration signal acquisition, digital signal processing, signal acquisition andstorage diagnosis, diagnosis and fault signals resulting signal information Fusion.Our cutting unit drive part of the main object of study, analysis of fault vibrationsignal shearer cutting part of the drive gear characteristics, In this paper, due to limitedconditions, the actual experimental conditions underground are not allowed, so onlythe fault of a similar type gears laboratory simulation experiments of this paper tobuild intelligent diagnostic system for testing.Optional hardware to build a platform forthe digital signal processing chip TMS320F28335processor, combine Shearerestablish specific hardware peripherals and interfaces. Software development iscompleted in two parts: one is to use MATLAB neural network toolbox to buildShearer BP neural network fault diagnosis, extract weights and thresholds, after theestablishment of the neural network model in Simulink and ultimately automaticgeneration algorithm, and thus the fault identification; another part of the mainprogram is written in CCS V3.3, including the data collection, data storage,communication, information integration and troubleshooting, by DS evidence theory tomake a decision level information fusion of the neural network diagnostic results,reducing the uncertainty of the results of fault identification, fault identification results raise the support rate.The above findings establish an actual mechanized mining requires real-timediagnostic system that can be applied to today’s electric traction shearer system,auxiliary Shearer stable.