基于单神经元的水下机器人S面自适应运动控制

Adaptive S plane control algorithm for underwater vehicles based on single neuron cell

从模糊逻辑控制方式出发,借鉴PID控制结构而形成的S面控制自身不具有自调整能力。为了加强机器人运动控制的自主性,改善控制的动态响应特性,探讨了基于单神经元的S面自适应控制学习算法,实现了控制参数的自动调整,增加的积分项改善了机器人的动态响应特性。实验结果和实际应用表明,基于单神经元的S面自适应控制器对于水下机器人的运动非线性控制具有响应速度快、超调小、自学习能力强、设计简单等多种优点。

Based on the analysis of fuzzy control, the S plane control that was formed with reference to PID control does not have self-learning ability. In order to improve the adaptability and dynamic response ability of Autonomous Underwater Vehicle （AUV）, the adaptive S plane control algorithm based on single neuron cell was introduced and the control parameters self adjustment was achieved and the added integral item also improved the dynamic response ability of AUV. Experimental results and application prove that the nonlinear control method has the merits of quick response, small overshoot, better self- learning and simple design and so on.