中子发生器ECR源智能控制的仿真研究

Research of Intelligentized Control for ECR Ion Source of the Neutron Generator

描述了模糊modular神经网络对中子发生器ECR离子源智能控制仿真实验的实现和仿真智能控制的过程。研究了学习过程中规则条数和学习率对理论模型计算结果平均偏差的影响和控制过程中计算束流的偏差随规则条数和学习率的变化。用3种不同的校正因子研究了对仿真实验过程的影响,并在智能控制中取得了满意的仿真实验结果。

The simulation experiments of intelligentized control, the realization of simulation procedure, and the simulation processes of intelligentized control with the theoretical model of fuzzy-neural modular network for ECR ion source of neutron generator are described in great detail. The averaged deviations of computed result of the model in the learning process and the deviation of computed results in the controlling process are reacted changeably with rule number and study rate. The influence of simulation experiment processes of intelligentized control with the correction factor of three different kinds for ECR ion source are researched, and the satisfying results of intelligentized control are obtained through simulation experiments.