T-S模型把一个非线性系统当做多个线性子系统与其权重乘积之和,能够逼近任意非线性系统。提出基于遗传算法和支持向量机的T-S模型全局优化辨识方法,利用遗传算法同时辨识T-S模型的结构和参数,以结构风险最小化作为辨识的评价指标,综合考虑模型复杂度和辨识误差,辨识精度高,泛化能力强,仿真结果证明了算法的有效性。
T-S model is sum of linear subsystems and its weight product, which is able of approaching arbitrary non-linear systems. T-S model global optimal identification scheme is presented based on Genetic Algorithm (GA) and Support Vector Machine (SVM). Structure and parameter identification of T-S model is realized by GA at the same time. Evaluation function is structure risk minimum, which considers model complexity and identification error. The algorithm advantages are high precision and good generalization of identification. The simulation result illustrates the effectiveness of the proposed method.