提出了一种基于模型建立仿射线性Takagi—Sugeno(T-s)模糊模型的方法;并对其局部与全局逼近性能进行了统一讨论;其中,考虑到该方法简单的局部逼近原理,定义了一种非线性测度指标,利用其建立优化问题对T-S模糊模型的前提变量输入空间进行非均匀划分以减少模糊规则数量。然后,面对建立风能转换系统(rvEcs)高糟度控制设计模型的需要,将上述方法应用于WECS本质非线性气动模型的线性化处理,得到了其T-S模糊线性化模型。最后,通过Matlab仿真实验及多种误差评定指标量化对比逼近精度,验证了上述方法的有效性。
A novel Takagi-Sugeno (T-S)fuzzy modeling procedure with affine linear consequent part was proposed based on the nonlinear model. The local and global approximation capabilities were uniformly discussed. Due to the simple treatment of local approximation, an index of nonlinear measure was defined. Through the index, an optimization problem was established to determine the un-uniform partition to the input space of the premise variables in the T-S fuzzy model, by which the fuzzy rules were efficiently reduced. Then, in order to obtain the control design model of Wind Energy Conversion System (WECS) with high accuracy, the T-S fuzzy modeling procedure was used to deal with the inherent nonlinearity of the aerodynamic model in the WECS. Finally, the model validation was executed through Matlab. Moreover, kinds of error evaluation indexes were used to quantify and compare the approximation accuracy which further suggested the effectiveness of the approaches.