Updating technique for particle swarm optimization in nonlinear dynamic systems

Dealing with searching and tracking an optimal solution in dynamic environment becomes more frequently nowadays. For dealing with this matter, Particle Swarm Optimization - Random Times Variable Inertia Weight and Acceleration Coefficient (PSO-RTVIWAC) concept, motivated by Particle Swarm Optimization-Time Variable Acceleration Coefficient (PSO-TVAC) and Particle Swarm Optimization-Random Inertia Weight (PSO-RANDIW) was introduced. PSO-RTVIWAC can accomplish an acceptable accuracy in detecting the target with the small number of particle and iteration. This paper will discuss about modifying the fitness value in the update mechanism for determining the local best and global best to improve the accuracy of detecting the target. By adding a constant value to the current stored fitness value, it will give the opportunity to the next fitness value to be the best fitness value. The result from this modifying technique then will be compared with PSO-RTVIWAC to evaluate the performance.