A Petri Net Model For Wireless Sequence Control System And Fault Diagnosis

Wireless sensor network has a great application in the industrial control system, so the performance analysis and fault diagnosis are of great value for Wireless sequence control system which combines wireless sensor network and sequence control system. This paper proposed the concept of Wireless sequence control system which consists of supervision level, control level and field level, modeled for Wireless sequence control system based on generalized stochastic Petri net from three aspects:endpoint, base station and controller, set the parameters of Petri net according to physical value and integrated above sub models to obtain a whole generalized stochastic Petri net model for Wireless sequence control system.In order to analyze the performance of Wireless sequence control system, firstly, this paper calculated the state transitions of simplified generalized stochastic Petri net model for Wireless sequence control system, secondly, estimated the time delay of device status uploading based on the mathematical relationship between generalized stochastic Petri net and Markov chain, thirdly, proposed the theory of delay compensation, put the calculated delay value into system model, lastly, simulated through software TimeNET to compare the system performance with and without delay compensation. The error with delay compensation is much little than without, which shows delay compensation is helpful for solving the transition delay of wireless sequence control system.In order to diagnose the fault of wireless sequence control system, firstly, this paper analyzed the faults according to construction to obtain the knowledge base of faults propagation, secondly, modeled for wireless sequence control system based on fuzzy Petri net and set the parameters of Petri net according to physical value, thirdly, added the judgment of threshold to make up the shortfall of forward fuzzy reasoning algorithm and considered all "conflict" conditions in backward reasoning algorithm, lastly, diagnosed the system fault respectively from modified forward and backward reasoning algorithm. The final conclusion shows "the probability of controller logic resending is the biggest" and "the greatest possible reason of fault: controller logic resend is wireless DI endpoint failed down".