The Research Of Multiple Target Knapsack Problem On Tabu PSO

Multi-objective optimization problems are important categories of the engineering and science research, which is multiple conflicting object. It is the main issues concerned by circles of academic and engineering. How to obtain the optimal solution of the problem in the limited time or time or resource price is the core of problem.With the largest difference of a single objective optimization problem, a improvement of one object may lead to the performance of other object will decline, under the most circumstances of the essence of multiple objective,at the same time it is not possible to reach the best multiple objects, intelligence deal with it between the objectives and eclectic treatment, to obtain the best solution of all the object, and the best solution is composed by a large numbers even infinite Pareto. In present, there are lots of ways to solve the multi objective, but the method of evolutionary intelligent the basic is used alwaysThis paper introduces the particle swarm algorithm, particle swarm optimization algorithm (also known as Particle Swarm Optimization, PSO) is a new intelligent optimization algorithm, and also describes the classic local optimization tabu search algorithm (Tabu Search or Taboo Search, TS). The two algorithm is proposed based on the particle swarm algorithm, proposes taboo-particle swarm optimization (Tabu-Particle Swarm Optimization Algorithm, T-PSOA), on the one hand, the particle swarm evolutionary process at a later stage into the tabu search algorithm based on tabu algorithm, tabu search and local search method to improve the first phase the quality of the solution. Taboo operation to produce new if better than original solution instead of the original solution. It defines a probability according to the elite, elite probability, random selection of particles of a certain number of dimensions or weight of a corresponding portion of the copy, to maintain the target space diversity. On the other hand, put forward to improve the original particle swarm optimization of operating parameters, namely the establishment of inertia weight as the measure of population diversity of adaptive changes in relationships, and into the algorithm. The measure of population diversity can use the diversity of particle swarm and the concentration index of representation. The target0-1knapsack problem as an example to verify the improvement of taboo-particle swarm algorithm compared with the original algorithm in the performance advantages, and through the example of improved particle swarm algorithm in different parameters to do some research.