Abstract: Aiming at the low accuracy and poor adaptation of variable universe fuzzy control, and when the control function is inherited to the offspring, there usually exists some distortion which lead to the error of the algorithms. A variable-universe fuzzy control method based on double precision search algorithm is put forward. With the global optimal solution of basic gravitational search algorithm, sequential quadratic programming is employed as a local search method to avoid being trapped into local optimum. This paper presents a double precision search algorithm with “global-local” dual search mechanism. Then the method adjusts the universe by the contraction-expansion factor and geometric proportional factors based on variable universe fuzzy control, and optimizes the parameters through double precision search algorithms to reduce the distortion of control function in the control process and improves the control performance. Comparative experiments show that the stability of DPSA is prominent in the parameter optimization. The controller results in desirable convergence speed. The accuracy and effect are even better than those of other control ways.

Key words: variable universe fuzzy control, double-precision search algorithm, contraction-expansion factor, geometric proportional factors

摘要： 针对定式变论域模糊控制精度不高，自适应能力有限，控制函数在遗传到后代时存在畸变而造成算法本身误差等问题，设计了一种基于双精度搜索算法的变论域模糊控制器。在基本万有引力算法全局搜索的同时，采用序列二次规划进行局部搜索避免算法陷入局部最优，提出具有“全局-局部”双重搜索机制的双精度搜索算法。在变论域模糊控制基础上提出了一种利用伸缩因子、等比因子相互协调来调整论域的构想，且通过双精度搜索算法来寻优参数，降低控制过程中的函数畸变，从而进一步改善控制器性能。对比实验表明DPSA在参数寻优中稳定性突出，控制器不但收敛速度快，且与其他控制方式相比，其精度和效果都有所提高。

关键词: 变论域模糊控制, 双精度搜索算法, 伸缩因子, 等比因子