【摘要】 为了克服传统机器人设计方法存在的局限性 ,提高机器人的自适应能力 ,采用神经网络方法实现了进化机器人避碰、趋近及其组合行为学习 .首先 ,提出了新的机器人模拟环境和机器人模型 ,给出了采用神经网络实现进化学习系统的方法 .其次 ,对具有进化学习机制的机器人基本行为和组合行为学习系统进行了仿真 ,并通过仿真证明了新模型不要求环境知识的完备性 ,机器人具有环境自适应学习能力 ,还具有结构简洁、易扩展等特点 .最后 ,对仿真结果进行分析与讨论 ,并指出了进一步研究方向更多还原

【Abstract】 A method to improve the adaptability of robot is presented, which can overcome the limitation of traditional robot design method. Obstacle avoidance, target approach and their combination behavior learning of evolutionary robot are realized by the use of artificial neural network in this paper. First, a robot learning environment and a robot model are presented and the implementation of evolutionary learning system is discussed. Then, the simulation experiments are carried out for basic and combined behavior learning system of intelligent robot that adopted evolutionary learning mechanism. Finally, the simulation results are analyzed and the future research direction is given.更多还原