基于滑模变结构控制的RBF神经元网络

针对高精度飞行仿真转台，设计一种基于滑模变结构控制的RBF神经元网络控制器。该控制器根据滑模变结构控制器的特点，将控制律分为等效控制律和到达控制律。等效控制律使系统运动于滑模面附近，由RBFN拟合而成，权值用自适应算法在线修正，确保了实时控制的可能性；到达控制律可使处于状态空间内任意初始位置的系统趋近于滑模面，由滑模控制器的可达性条件推出，其中用到了系统的不确定性参数的上下界。计算机仿真结果表明了该方法的鲁棒性和实际应用的可能性。

Radial Basis Function Sliding Mode Controller

A sliding mode controller based on RBFN is presented for high precision flight simulation bed. The control value is divided into a hitting control law and an equivalent control law according to the moving characters of the sliding mode controller. Equivalent control law is designed to keep the system sliding along the sliding surface. It is approached by a RBFN and the weight of the network is tuned on line using adaptive algorithm to guarantee the probability of the real time control of the system. Hitting control law is applied to drive the representation point everywhere of the state space onto the sliding surface. It is deduced from the hitting condition of the sliding mode controller and the upper and lower bound parameters must be known. The result of computer simulation demonstrates the robustness and the effectiveness of the proposed algorithm.