开颅手术机器人的人-机-环交互机制与协同控制方法研究

开颅手术是神经外科一种典型的治疗手段，由于颅骨解剖结构不均，其下方血管、神经、静脉窦等分布错综复杂，目前完全依赖人工实施开颅手术的传统术式存在手术风险高、长时间负重操作、体力精力消耗严重、安全性稳定性差等诸多问题。手术机器人在提高手术的稳定性、可靠性、安全性以及减轻操作强度等方面提供了一种有效的手段。.基于开颅手术的临床需求，构建由医生、机器人和患者所组成的典型的“人-机-环”"融合系统，突破逆向驱动柔顺交互操作、虚拟约束变阻抗控制、医生在环的人机增强导纳、虚拟约束和力交互自然约束下的混合伺服以及基于强化学习应对不确定性约束信息的控制优化等瓶颈技术，解决未知不确定性环境下的交互机制与协同控制问题，实现人机协同手术操作，提升手术机器人的智能化水平，在有效解决开颅手术稳定、柔顺、安全、省力等临床操作难题的同时，进一步丰富了人机交互协作控制的基础理论，有着重要的科学意义与研究价值。

Craniotomy is a typical treatment in neurosurgery. Because the skull structure is extremely complex, and the blood vessels, nerves, and sinuses underneath it are intricately distributed, the current surgical risk is very high. Doctor works very hard, consume a lot of physical energy, and keep their concentration at all times, which is a big hidden danger to the safety of surgery. Surgery robot provides an effective method in improving the stability, reliability, safety and reducing the intensity of operation. Based on the clinical requirements of craniotomy, a system of "human-robot-environment" composed of doctors, robots, and patients was constructed. The reverse drive for compliant interaction operation, virtual fixtures implemented by impedance control, doctor-in-loop reinforcement admittance in human-robot interaction, hybrid servo under virtual fixtures and natural constraints and controller optimization aiming at uncertain constraint information using reinforcement learning are investigated which are conceived to be key points to achieve human-robot collaborative intelligent operation and to improve the robot intelligence level and to. While efficiently solving the clinical challenge of stability, compliance, safety as well as effort-saving, it can supplement the control theory for human-robot interaction and collaboration, with significant scientific and researchable value.