Design Of Fan Fault Diagnosis And Monitoring System Based DSP Combined With ARM

Mine main fan is one of the four important equipments; it is an importantguarantee for the safety production of coal mine. It is not only responsiblefor underground air purification work, but also completes discharge of coal minedust and other toxic substances, so the work of the main fan is directly related to thesafety of mine production.At present, cost for main fan monitoring is very high, it is hard to realize the highspeed sampling for the vibration signal, so this paper designs a method which is basedARM combined with DSP to realize the fan monitoring system. This method uses thesetps introduced below to study the main fan monitoring system. At first, as therelationship between the spectrum distributions of the vibration signal of fan and thefault of fan, this paper does research on the fault diagnosis based on BP NeuralNetworks, but the BP Networks has some bugs, so this paper studies the algorithmoptimized by the Genetic Algorithm.Then hardware design based ARM combined withDSP is studied, it puts coefficient matrix which comes from the upper machine into theplatform which is made of DSP and ARM, then the algorithm is realized by thishardware platform.Through the comparision between two algorithms, Neural Netwoks optimized bythe Genetic Algorithm has advantage over simple Neural Networks; it can overcomethe bug of BP Neural Networks. As the hardware design, througth this structure, wecan put the coefficient matrix into the lower machine directly; it can meet therequirement of fan fault diagnosis and monitoring system. So this structure has thewide application prospect in main fan fault diagnosis and monitoring system.