Hybrid Algorithm With Artificial Fish Swarm And Particle Swarm For Coverage Optimization Research In The WMSNs

With the increasing diversification of environmental monitoring,theinformation is to be aware of people becoming more and more complex, such asimage、video and so on. So the wireless multimedia sensor networks (WirelessMultimedia Sensor Networks, WMSNs) is born, that is a new type of sensornetwork joined by video, audio, image and other multimedia perception basedon the traditional sensor network. Coverage control is a basic problem in sensornetworks, and it is one of the important indicators to measure monitoring theperformance of the sensor network. And also it is important to ensure the qualityof the follow-up study. The main purpose of the coverage problem is to improvethe initial node random distribution of poor coverage and solve the problem ofwaste resources by adjusting the sensor nodes in the target area.In WMSNs, sensor network nodes have direction and angle that is differentfrom wireless sensor network nodes. It is because that the sensor network isusually monitored for unknown environment or inaccessible environment, so the random deployment of network coverage is used.The location of node isdifficult to ensure a reasonable distribution one time with random deployment,that it will produce more overlapping areas and blind areas. So the coverageoptimization algorithm is used to ensure coverage of the target area, and tooptimize network quality. In this thesis, the hybrid optimization algorithm withartificial fish swarm and particle swarm is put forward based on the WMSNs.Particle swarm optimization algorithm has a faster convergence rate, but itwill easily fall into local optimum and cause precocious phenomenon when theparticles are moving to the optimal direction with the same direction in the laterstage. Artificial fish swarm algorithm is able to jump out of the local extremevalue, and possible to search to the other extreme as far as, and ultimately to theglobal extreme closely, but the convergence speed of the algorithm is slow in thelater, that can find a satisfactory solution domain, but it is difficult to get theaccurate optimal solution. In this thesis, two kinds of algorithm are combinedtouse of WMSNs coverage based on the advantages and disadvantages of thesetwo algorithms,the first to use the artificial fish swarm algorithm to adjust thedistribution of sensor nodes, in order to distribute nodes uniform, then use theparticle swarm algorithmto search quicklyof local, and maximizing the coverageof monitoring area by adjusting the position and direction of sensor nodes.Improving the velocity and position update equation of particle swarmoptimization algorithm. Due to the velocity update formulaof the particle swarmalgorithm are includes the three contents that coupling and conflict is often existed. The inertia，historical experience and group experience of the particleare included,that will lead to the deviation of nodes.Therefore, the speed of theparticle is slowlyupdatedand the performance of the algorithm is reduced. So thespeed of the particle which is include the flight of inertia, historical experienceand group experience are calculatedseparately, and try to more location andselect the best location as the new location of particlesform the multiplelocations whenthe particle chose the locationin order to improve theconvergence speed and accuracy, and aim at achieving better optimization ability.Finally, hybrid algorithm with artificial fish swarm and improve particle swarmis used for coverage optimization, through the simulation experiment to provethat the effect of improved hybrid algorithms can improve the network coveragerate.