郑昊辰^1,2　刘成沛¹　翟雪倩²　吴鸿敏²

（1.五邑大学轨道交通学院，广东 江门 529020　

2.广东省科学院智能制造研究所/广东省现代控制技术重点实验室，广东 广州 510070）

摘要：针对机器人与软组织交互过程中存在形变预测难、接触属性未知、动力学模型不连续和异常组织识别精度低等问题，提出一种机器人连续性接触状态估计与软组织异常识别方法。通过结合半隐式积分的扩展卡尔曼滤波器预测接触刚度；利用赫兹接触模型、库伦摩擦模型分别感知接触法向力、接触摩擦力；构建包含误差反馈的机器人导纳控制模型，保障交互的稳定性，从而实现对异常组织的有效识别。通过硅胶异常软组织识别实验与离体猪肝异常组织识别实验，验证了该方法的有效性和鲁棒性。

关键词：状态估计；异常识别；导纳控制；接触刚度；赫兹接触模型；库伦摩擦模型

中图分类号：TP242文献标志码：A 文章编号：1674-2605(2025)05-0004-08

DOI：10.12475/aie.20250504开放获取

A Method for Robotic Continuous Contact State Estimation and Soft Tissue Abnormality Identification

ZHENG Haochen^1,2LIU Chengpei¹ZHAI Xueqian²WU Hongmin²

(1.School of Rail Transportation, Wuyi University, Jiangmen 529020, China

2.Institute of Intelligent Manufacturing, Guangdong Academy of Sciences/  Guangdong Key Laboratory of Modern Control Technology, Guangzhou 510070, China)

Abstract: Aiming at the problems of difficult deformation prediction, unknown contact attributes, discontinuous dynamic models, and low accuracy in abnormal tissue identification during robot-soft tissue interaction, a method for robot continuous contact state estimation and soft tissue abnormality identification is proposed. The contact stiffness is predicted by combining an Extended Kalman Filter with semi-implicit integration. The Hertz contact model and the Coulomb friction model are utilized to perceive the normal contact force and the contact friction force, respectively. An admittance control model incorporating error feedback is constructed to ensure interaction stability, thereby enabling effective identification of abnormal tissues. The effectiveness and robustness of the proposed method were validated through experiments on identifying abnormalities in silicone-based soft tissues and ex vivo porcine liver tissues.

Keywords: state estimation; abnormality identification; admittance control; contact stiffness; Hertz contact model; Coulomb friction model