Research On Gear Fault Early Warning Instrument System Based On Time Series Analysis

Equipment fault prediction and m anagement system is a hot res earch topic at home and abroad in recent years. E quipment failure prediction means that the state characteristics of the de vice at the next time is determined according to the runnin g state characteristics of the equipm ent at present time and the running state characteristics after a period of tim e, and then through the analysis of the characteristics to predict health status and remaining service life of the equipment. Gear system is a key drive o r force tr ansmission member of the device, so the research on the condition m onitoring and fa ult prediction of gear equipm ent is of great significance to the good operation of the equipment.Because of the importance of gear fault prediction, this paper mainly studies the following contents: 1) The theory and process of ARMA model construction; 2)The fault mechanism and spectrum characteristics of the gear and the diagnostic criteria of the gear fault and the problem of freque ncy band division in fault analysis; 3) Equipment vibration signal collector and online monitoring software. The research contents focus on the gear vibration signal acquisition, signal analysis and processing, feature extraction of gear state, state identification and state prediction.Firstly, this paper constructs ARMA model which is applied to the prediction of the spectrum structure of gear. Then the ARMA prediction m odel is applied to a cement plant kiln reducer gear spectrum, based on the difference of the performance of the distributed fault and local fault in the gear analyses were cond ucted on th e trend of the distributed fault or local fault of the reducer high-speed side gear. So as to arrive, the gear is in relatively good operating condition but it has a trend of in the occurrence of distributed fault, people should continue to focus its development trend.Then, this paper studies the realization of the distributed intelligent collector and the online monitoring software in the equipm ent fault prediction and m anagement system. Distributed intelligence co llection uses A8 + DSP processor and FPGA control sample, which has 8 vibration channe ls and a channel speed, and it be able to collect the acceleratio n, velocity and di splacement signals of equipm ent; the collector’s processing for vibration signal include isolated DC, filtering, amplification, integration, A/D conversion, etc. In order to be able to achie ve the signal in the case of an exception to the local s torage acquisition device is also equip ped with an extended memory, at the sam e time, in order to obtain good communication and signal transmission efficiency, it has both USB/RS485 interface and network interface, the signal can be trans mitted to the term inal for monitoring and analysis rea l-time. The online monitoring software is made up of monitoring picture, analysis/evaluation, comprehensive report, management system, to achieve the status of th e equipment monitoring, forecasting and m anagement through analy zing and processing the signals received from the collector in re al-time. Intelligent collector and on lin e monitoring system has been applied in e ngineering practice, and they have good engineering applicability...