针对一般最小二乘支持向量机处理大规模数据集会出现训练速度慢、计算量大、不易在线训练的缺点,将修正后的遗忘因子矩形窗方法与支持向量机相结合,提出一种基于改进的遗忘因子矩形窗算法的在线最小二乘支持向量机回归算法,既突出了当前窗口数据的作用,又考虑了历史数据的影响.所提出的算法可减少计算量,提高在线辨识精度.仿真算例表明了该方法的有效性. For the general least squares support vector machine (LSSVM), large scale data sets will suffer from the disadvantages of slow training speed, large amount of computation and difficulty of online training. Combining the modified forgetting factor rectangle window method and support vector machine, The algorithm of forgetting factor rectangle window algorithm based on least square support vector machine regression not only highlights the function of the current window data but also considers the influence of historical data.The proposed algorithm can reduce the amount of computation and improve the on line recognition accuracy. The effectiveness of this method is demonstrated.