Application Of A Modified PSO Algorithm Combining With GA Operators In Control System Design

Nowadays, the industrial controlled object become more and more complicated, and control requirements continuously improved, in order to meet the requirements of continuous improvement of production, ensure the production process become more stable and meet the requirements.How to find the exact parameter model, and on this basis to find the best controller parameter is very necessary and urgent.Particle swarm optimization (PSO) presented in the past decade. It's a swarm intelligence optimization algorithm, which adapted to the complex optimization problem and it can find the best solution of nonlinear equations quickly. Because the principle of the particle swarm optimization algorithm is simple,easy to understand and get the solving results quickly. PSO has developed into an important branch of intelligent optimization research, and applications in many fields. But particle swarm also has its inherent defect, such as premature convergence, the algorithm may get into local minimum value and can't find the global optimization. Aimed at problems of PSO,the particle swarms optimization based on particle evolution is proposed in the paper, make it more suitable for the actual industrial production, that has important theoretic value and practical significance.Because of the problem that the particle swarm optimization is difficult to deal with local minimum and premature, a modified and novel PSO algorithm combining with GA operators is presented in the paper, and joining in inertial factors that take nonlinear reducing mechanisms, get a new kind of GAPSO algorithm.Meanwhile,another PSO combination of Simulated Annealing composition SAPSO, comparison of three different optimization algorithm, and find the GAPSO is better, so choose the GAPSO algorithm in the next research work.Using Matlab and Visual C++ realized, the novel algorithm is apply to tune of PID control system parameters and the controlled object parameter identification, and simulation results show that the modified algorithm is better than the Particle Swarm Optimization Algorithm and Genetic Algorithm, and improves the PSO's convergence properties remarkablely.Because aim at the real industrial object model identification and controller parameters optimization, the real industrial system use Visual C++ to achieve more than language, it integrates calculation, visualization and programming, with a powerful mathematical computations and good visual interface, use it to achieve this algorithm has good advantage. Finally, compared optimization effect of visual C++ and Matlab, and the optimization results very similar, but the former optimization speed is much faster than the latter, which has a higher practical value.At last, the control system is realized by using super multifunction process control training system (SMPT- 1000) as actual controlled object, for the practical application of the PSO algorithm. Through the OPC client completion of data collection, use Visual C++ realize a modified and novel PSO algorithm for the model of heat exchanger's parameter identification, and realized the SAPSO algorithm for boiler superheated steam temperature control optimization,superheated steam temperature pressure to the parameter model identification, the simulation results shows that the improved algorithm is effective and feasibility.