【摘要】 为实现3自由度单级直齿轮系统的混沌运动有效控制，用集中质量法建立系统的动力学模型，并用4～5阶Runge-Kutta法求解得到参数区间内的周期运动向混沌运动转迁的规律。针对特定参数区域的混沌运动，以控制参数的扰动量为输出，Poincaré截面上点的欧式距离为输入，构建径向基函数神经网络控制器，使用改进局部搜索能力和寻优速度的引力搜索算法优化径向基函数神经网络控制器的参数，实现系统混沌运动向周期运动的有效控制。结果表明径向基函数神经网络控制方法不受系统的Jacobian矩阵和流形的限制更具有工程普适性。更多还原

【Abstract】 In order to realize the effective control of chaotic motion in a 3-DOF single-stage spur gear system, the dynamics model of the system was established by using the concentrated mass method, and the law of transferring the periodic motion to chaotic motion in the parameter interval was obtained by solving the equation with the 4th or 5th order Runge-Kutta method. For the chaotic motion in a specific parameter region, the radial basis function neural network(RBFNN) controller was constructed with the perturbation amount of the control parameters as the output and the Euclidean distance of the points on the Poincaré cross-section as the input, and the parameters of the RBFNN controller were optimized using a gravitational search algorithm that improved the local search capability and the speed of searching for the optimum, so as to realize the effective control of the system’s chaotic-to-periodic motion. The results show that the RBFNN control method is not limited by the Jacobian matrix and flow shape of the system, and is more universal in engineering.更多还原