Research On Wireless Sensor Network Positioning Technology In Underground Coal Mine Based On TDOA

In recently years, with the increasing amount of coal, coal mine accident occursfrequently. Therefore, It becomes a vital issue that how to find out undergroundpersonnel positioning information timely and accurately and implement effectiverescue after mine accident. At the same time, wireless sensor network becomes anemerging technique, and node positioning is a hotspot research. Therefore, it’simportant to research node positioning of wireless sensor network which is suitablefor underground coal mine complicated environment.In this paper, several types of classical algorithm of wireless sensor networknode positioning have been studied systematically, TDOA positioning technique ofwireless sensor network has been studied intensively, and two classic TDOApositioning algorithms--Chan algorithm and Talor series expansion algorithm havebeen introduced in this paper. The objective function is worked out by TDOApositioning mode derivation. Meanwhile, in order to figure out the problem ofdifficult objective function solving, the improved Particle Swarm Optimization (PSO),combining tabu search and PSO algorithm, has been introduced correspondingly. Thisalgorithm conducts searching by making use of PSO: PSO will stop when algorithmiterates several times and meets tabu search starting standards, and the optimalsolution can be partly searched by tabu search algorithm. For the one thing, thisalgorithm can work out the weakness that tabu search algorithm highly depends oninitial solution, for another, it fully exerts the merit that high tabu search algorithm hillclimbing ability and it can makes for that PSO falls into local optimum prematurely.The PSO algorithm integrates organically with the tabu search algorithm. Simulationresult shows that the improved algorithm makes remarkable improvement in rate andprecision of convergence.NLOS transmission is one of the main reasons of restricting positioning accuracyin wireless sensor network node positioning in underground coal mine. On the basisof research on Wylie identification method, this paper puts forward improved NLOSidentification method, which takes ranging variance and mean value as NLOSidentification index, and increases NLOS identification success rate. Meanwhile, thepositioning algorithm, which TDOA measured value reconstitute, restrain andeliminate NLOS errors, identifies NLOS propagation in TDOA measurement resultsby using improved NLOS identification method, estimates mean value and variance in NLOS propagation delay according to its statistical property, correct TDOA measuredvalue, use this measured value derives new positioning target function, then use tabusearch algorithm to optimal position. Simulation results indicate this algorithmproperty.