Research On Data Compression And Data Transmission In Wireless Monitoring On Fans And Pumps

A great of fans and pumps are widely distributed in the large chemicalplant.Once fault occurs, it will cause device shutdown and economic losses tothe enterprise.Worsely, it will cause Heavy accidents and casualty. Real-timeonline monitoring system is a necessary condition to implement to predictrepair and ensure the stable operation of pump equipment long cycle. Studiedfrom previous work, this paper researched vibration data compression,wireless relay node deployment and wireless routing protocol of the keytechnology to build a wireless real-time monitoring system of fans and pumpsin the large chemical plant and obtained the following achievements:(1) After researching all pump fault diagnosis reports in China PetroleumLiaoyang Petrochemical Research Institute of machinery since1990, wesummarize the main information which provided for the on-line monitoringand fault diagnosis by the pump equipment vibration signal. Further, wepropose unique characteristics of the pump equipment vibration signalcompression contrasted with other signal compression and analyze the characteristics of the algorithm which applied to pump equipment vibrationsignal compression.Aiming at the large reconstruction error exists in the current datacompression algorithms under very low transmission rate which easily lead tofalse diagnosis, this paper presents a new compression algorithm combinedwith the orthogonal transform and vector Quantization. Proved by actualsignal, for vibration signals made by rolling bearing, sliding bearing and gearbox, this algorithm can significantly improve the reconstruction snr in thesituation of low rate transmission (less than300bit/s).(2) Through the research of the relationship between reliability ofindustrial monitoring system and deployment problem relay nodes and theresearch of wireless channel in large-scale chemical plant, we analysis thecomplexity of wireless relay node optimal deployment problem in largechemical device. By simplifying the problem using greedy idea, wetranslatethe whole relay nodes deployment problem into several singlegradient levels relay node deployment optimization problem and set up amathematical model based on0-1programming.In view of the contradiction between the reliability of monitoring anddeployment costs in the process of wireless relay node deployment in largechemical plant, this paper presents a new method for wireless relay nodedeployment. The method contains three parts: wireless channel survey,estimation of wireless channel path attenuation factor and wireless relay node deployment.Through an actual deployment in chemical plant, The algorithmcan effectively reduce the number of wireless relays in the situation of meetingthe premise of real-time and reliability of industrial monitoring, and the timecomplexity of the algorithm is significantly lower than the general heuristicalgorithm.(3) By analysing routing strategies of the industrial wireless protocol, weestablished two mathematical models of overhead and transmission delay ofthe routing strategies in industrial wireless protocol. According to actualdemands of pumps group equipment monitoring in large-scale chemical plant,we analyze features of two kinds of main monitoring mode: daily check modeand equipment alarm mode and propose a routing strategy for each monitoringmode.According to problems of too large routing control overhead and highdata transmission delay in current industrial wireless routing protocol, thispaper proposed a new wireless routing protocol combined with routing tableand on-demand routing according to the monitoring mode of fans and pumpsin large chemical plant. The protocol chooses wireless routing in differentways according to different monitoring mode. Simulation results show that, inthe inspection mode, the proposed protocol can reduce the wireless routingoverhead by30%-40%; in the alarm mode, the proposed protocol can reducetransmission delay by10%-50%.Finally, in order to validate the research results, we built an experimental wireless monitoring system for fans and pumps in the1400000tons/year ofreforming plant of CNPC Liaoyang Petrochemical Industries Co. Theeffectiveness of our system is verified through the practical application.