Intelligent Control of Robotic Contact Operations

[00:38 . 88]纽曼机器人力控制和神经控制这两个独立领域的最新发展已经发展到这样的地步：这些技术的整合可能会在机器人能力方面产生巨大的进步。目前，机器人接触操作是笨拙和无效的，除非在环境高度可预测的情况下，任务可以单独使用位置控制来完成。然而，对于机器人的未来应用，如天基电力系统的自主装配、国防应用机器人或国内机器人的潜在行业，在非结构化环境中接触操作的能力是成功的关键。探索性项目将评估智能控制算法的使用，并采用适当的机器人/环境接触动力学假设和表示，使用“自然导纳控制”和“虚拟动力学”作为整合高级和低级控制的手段。它将考虑智能控制研究中的四个任务：(1)构建一个自反性损伤避免层，以确保机器人及其环境的安全，并使其具有安全自主学习的能力；(2)通过原子行为识别和触发顺序行为的事件识别，分析和“克隆”人类接触操作策略，以供机器人执行；(3)环境动力学在线识别；(4)在NAC/虚拟动力学背景下，构建基于神经网络的“自适应批评”优化控制算法。

0200388NewmanRecent developments in the separate domains of robot force control and neurocontrolhave advanced to the point where integration of these technologies may yield a dramaticleap forward in robot competence. At present, robot contact operations are clumsy andineffective, except in instances where the environment is highly predictable and the taskscan be accomplished using position control alone. However, for futuristic applications ofrobots, such as autonomous assembly of a space-based power system, robots for defenseapplications, or a potential industry in domestic robotics, competence in contactoperations in unstructured environments is essential for success.The exploratory project will evaluate the use of intelligent control algorithms with appropriate set of assumptions and representations of robot/environment contact dynamics, using "Natural Admittance Control" and "virtual dynamics" as a means to integrate high and low-level controls. It will consider four tasks in intelligent control research: (1) constructing a reflexive damage-avoidance layer to assure the safety of a robot and its environment and enable the capability for safe, autonomous learning; (2) human strategies for contact operations will be analyzed and "cloned" for robot execution, using identification of atomic behaviors and recognition of events that trigger sequential behaviors; (3) on-line identification of environment dynamics; (4) neural-net based optimizing control algorithms employing "adaptive critics" will be constructed in the context of NAC/virtual dynamics.