针对舵桨联合操控水下机器人系统的非线性特点,将预测控制的思想引入经典S面控制中,构造了一种基于支持向量机(SVM)的预测S面控制器,改善了S面控制器的控制效果,增强其自适应性.用支持向量机辨识水下机器人的非线性系统模型,充分发挥了SVM的泛化能力,能准确预测其运动状态.构造二次型性能优化函数以获取S面控制器的最优控制参数,进而获得水下机器人最优控制律.仿真结果表明:基于支持向量机的预测S面控制器具有结构简单、响应速度快、鲁棒性好等优点可行且有效. Aiming at the non-linear characteristic of rudder propeller and joint control of underwater robot system, the idea of ​​predictive control is introduced into classical S-plane control. A predictive S-plane controller based on support vector machine (SVM) And enhances its adaptability.Using SVM to identify the nonlinear system model of underwater robot, it gives full play to the generalization ability of SVM and accurately predicts the motion state of SVM.The quadratic performance optimization function is constructed to obtain S The optimal control parameters of surface controller are obtained and the optimal control law of underwater robot is obtained.The simulation results show that the proposed S-plane controller based on SVM has the advantages of simple structure, fast response, good robustness and so on .