提出了一种带有约束矩阵的Hénon混沌系统的广义预测控制快速算法,首先用带有遗忘因子的递推最小二乘参数辨识法来逼近混沌系统,然后该算法引入了一种约束矩阵,约束矩阵的引入避免了矩阵求逆的计算,占用内存小,计算速度快,并且具有较强的跟踪给定信号和抑制系统参数摄动和随机噪声的能力.仿真结果验证了该方法的有效性. A fast generalized predictive control algorithm for Hénon chaotic systems with constrained matrices is proposed. First, a chaotic system is approximated by recursive least square parameter identification with forgetting factor. Then, the algorithm introduces a constraint matrix The introduction of matrices avoids the calculation of matrices inversions, takes up less memory and has higher computational speed, and has better ability to track given signals and restrain system parameters perturbation and random noise. Simulation results verify the effectiveness of the proposed method.