论文部分内容阅读
为克服模糊规则提取的盲目性和随机性,提出了一种基于新的自适应模糊C-均值聚类(AFCM)算法的T-S模糊建模方法。首先利用减法聚类来确定聚类数目的上限和初始聚类中心,然后采用改进的模糊C-均值聚类(FCM)算法进一步优化聚类中心,最后通过聚类有效性评判方法自适应地确定规则数及聚类中心,同时改进的FCM算法也克服了野点数据对聚类结果的影响;进而利用加权最小二乘法估计模糊模型的结论参数。用于某型陀螺仪漂移趋势预测中,能够自适应地确定模糊规则个数,并取得了较高精度。仿真实验结果验证了该方法的有效性和可行性。
To overcome the blindness and randomness of fuzzy rule extraction, a new T-S fuzzy modeling method based on the new adaptive fuzzy C-means clustering (AFCM) algorithm was proposed. Firstly, subtractive clustering is used to determine the upper limit of the number of clusters and the initial cluster centers. Then, the improved fuzzy C-means clustering (FCM) algorithm is used to further optimize the clustering centers. Finally, the clustering center is adaptively determined by clustering effectiveness evaluation Rule number and clustering center. At the same time, the improved FCM algorithm also overcomes the effect of the data of the field point on the clustering result. And then the conclusion conclusion parameters of the fuzzy model are estimated by weighted least square method. It can be used to predict the number of fuzzy rules adaptively in the prediction of drifting trend of a certain type of gyroscope and achieves higher accuracy. Simulation results verify the effectiveness and feasibility of the proposed method.