一类基于计算转矩的机器人控制方法性能分析

PERFORMANCE EVALUATION OF CONTROLLERS BASED ON COMPUTED TORQUE FOR ROBOT MANIPULATORS

比较了五种基于计算转矩控制方法在机器人运动控制中的性能:固定增益的计算转矩控制、模糊调整增益的计算转矩控制器、扇形模糊逻辑控制器、带饱和函数的组合非线性反馈控制器和无饱和函数的组合非线性反馈控制器。这些控制器都是基于模型的控制方法,具有完全的机器人非线性动力学补偿。所有的控制行为均考虑到现实的限制,如驱动器饱和与摩擦扰动。在同样的期望轨迹下,比较各个控制算法的瞬态响应和稳态特性。仿真结果表明,带有饱和函数的组合非线性反馈控制器具有最佳的瞬态性能,扇形模糊逻辑控制器具有最好的稳态性能。

The paper compares the performance of five kinds of controllers based on computed torque with the problem of robot motion control： computed torque control with fixed gains, computed torque control with varying gains tuned via fuzzy approach, sectorial fuzzy controller, composite nonlinear feedback controller with saturation function and composite nonlinear feedback controller without saturation function. All the controllers are model-based methods and have the complete nonlinear dynamics compensation. The physical limitations are considered, such as actuator saturation and friction. The transient response and steady performance are compared among these control algorithms under the same desired trajectory. The simulation results show that the composite nonlinear feedback controller with saturation function has the best transient response, while the sectorial fuzzy controller has the best steady performance.