【摘要】 以操作能量作为评价函数进行水下机器人运动规划的结果得到随时间变化的目标运动路径和理论控制输出。本文探讨为实现运动控制过程中能量最小化,在进行目标路径跟踪的前提下作为随机量的实际控制输出与理论控制输出的一致性问题。提出了神经网络控制器的结构和学习方法。水下机器人"Twin-Burger"的实验结果验证了本文所提出方法的有效性和可行性。更多还原

【Abstract】 On the basis of the results of the motion planning of AUV by supposing the operating energy as evaluation function,the objective motion path and theoretical control output varied with time were obtained.On the condition of objective path trail,in order to realize the minimum operating energy,the conformance problem of the random values of the practical and theoretical control output was discussed also.The structure and studying method of neural network controller were proposed.The experimental results of the AUV-Twin-Berger verified that the way proposed in this paper are effective and feasible.更多还原