Application of adaptive genetic-simplex algorithm in parameter optimization of nuclear power components

An adaptive genetic-simplex algorithm for parameter optimization in nuclear power plant is proposed by using adaptive crossover and mutation techniques and integrating genetic algorithm with simplex algorithm. The modified algorithm enables the handing of nonlinear constrained optimization problem because of dramatically improved search capability. Performance comparison between the proposed algorithm and original ones is performed by solving the optimization test problem. To implement parameter optimization in nuclear power plant, the mathematical models of electric heating pressurizer and natural circulation steam generator are established. Finally, by using the modified algorithm, the optimization of steam generator and pressurizer aiming at minimizing weight and volume is implemented respectively. It has been found that the modified algorithm finds global optimal solution effectively in all algorithm tests while the original algorithms only make it in some of the tests. For parameter optimization, the weight and volume of steam generator decreases by 18.56% and 18.39% respectively, and the decrements are 16.54% and 18.97% for that of pressurizer. It is demonstrated that the adaptive genetic-simplex algorithm is capable of dealing with the optimization of nuclear power components. The optimization in this study may provide effective guide for engineering design. Keyword: parameter optimization; steam generator; pressurizer; adaptive genetic-simplex algorithm