Abstract: Combining a multiplier method which deals with constraint problems with improved particle swarm optimization algorithm, a new hybrid particle swarm optimization algorithm was proposed for solving non-linear constraint problems. The new algorithm takes advantage of the particle swarm optimization algorithm and the multiplier method, for the non-available particle appearing in the iterative process, using the multiplier method to produce feasible particle, and then search its optimal value by improved particle swarm optimization. Thus it can not only reduce the probability of falling into local minimum, but also can improve the search accuracy. And the numerical tests show that the proposed new algorithm has the characteristics of validity and searching for more precise particle and better robustness.

Key words: Non-linear constraint optimization, Particle swarm optimization algorithm, Multiplier method