基于改进萤火虫算法的T-S模型辨识

T-S Model Identification Based on Improved Firefly Algorithm

提出改进萤火虫算法的T-S模型辨识方法。针对传统T-S模型辨识方法中将前件参数和后件参数分开辨识而不能全局优化辨识的缺点,应用改进萤火虫算法对前件参数和后件参数整体编码整体辨识。改进萤火虫算法是在原始算法基础上对吸引度系数作自适应变化,目的是增强算法在迭代初期的搜索能力,防止其陷入局部极值点,并降低算法在迭代后期在最优解附近的振荡,以提高解的精度。提出的方法能较好地找到全局最优解,具有较高的辨识精度。仿真示例证明了改进方法的有效性。

In the paper a new T-S fuzzy model identification method based on Firefly Algorithm was proposed. To overcome the shortcoming of the traditional methods which identify premise parameters and consequences parameters separately, the premise parameters and consequences parameters were encoded together and an Improved Firefly Algo- rithm was applied to do the global optimization identification. Based on original algorithm, the Improved Firefly Algo- rithm changes the attractiveness coefficient self-adaptively, and the purpose is to enhance the algorithm＇ s capability to search in the beginning period of iteration to avoid falling into local optimum, and to reduce the vibration around the optimal solution in the later period of the iteration to improve the accuracy of the solution. The proposed method has a strong ability to find the global optimum, and the identification accuracy is very high. The simulation result il- lustrates the effectiveness of the nroposed method.