External Force Estimation Method for Force Sensorless Industrial Robots with High Anti-interference

doi:10.3901/JME.260091

Abstract

Abstract: To address the issues of external force estimation for industrial robots—namely, its dependence on accurate dynamic models, poor disturbance rejection, low estimation accuracy, and insufficient interpretability—an external force estimation method for force sensorless industrial robots with high anti-interference is proposed. Firstly, a dynamic model of a 5-degree-of-freedom industrial robot with a 3T2R configuration is established, considering the effects of joint friction torque and external forces. The nonlinear characteristics of joint friction are modeled using the Stribeck friction–velocity model. Secondly, an adaptive super-twisting sliding mode generalized momentum observer, with an adaptively tuned rate gain parameter, is then employed to estimate joint external torques. This approach improved estimation accuracy, reduced reliance on precise dynamic models, and enhanced interpretability. Thirdly, a mapping relationship between joint external torques and external forces is constructed using the Jacobian matrix, enabling external force estimation. Fourthly, comparative experiments are conducted against the adaptive super-twisting sliding mode observer, the super-twisting sliding mode observer, and the first-order generalized momentum observer. Taking Joint 1 as an example, the proposed method reduced the mean absolute error of joint external torque estimation by 12.9%, 39.9%, and 57.9%, respectively. Based on the established mapping, external forces are further estimated, demonstrating higher accuracy and validating the effectiveness of the proposed method. Finally, under the application of the same random disturbance, the proposed method is compared with the first-order generalized momentum observer. The root mean square error of the estimated joint external torque increased by only 1.5% for the proposed method, whereas it increased by 43.3% for the first-order method, verifying the superior disturbance rejection capability of the proposed method.

Key words: industrial robots, external force estimation, friction torque model, generalized momentum observer, super-twisting sliding mode

CLC Number:

TP242

GUO Wanjin, LI Qianhui, XU Mingkun, HOU Xudong, LIU Xiaoheng, CAO Chuqing, ZHAO Lijun. External Force Estimation Method for Force Sensorless Industrial Robots with High Anti-interference[J]. Journal of Mechanical Engineering, 2026, 62(3): 366-383.

References

[1] GE J M，DENG Z H，WANG S X，et al. Vision sensing-based online correction system for robotic weld grinding[J]. Chinese Journal of Mechanical Engineering，2023，36(1)：125.
[2] KOMMURI S K，HAN S，LEE S. External torque estimation using higher order sliding-mode observer for robot manipulators[J]. IEEE/ASME Transactions on Mechatronics，2021，27(1)：513-523.
[3] WANG G，WANG Z，HUANG B，et al. Active compliance control based on EKF torque fusion for robot manipulators[J]. IEEE Robotics and Automation Letters，2023，8(5)：2668-2675.
[4] 邓建新，袁邦颐，黄秋林，等. 基于工业机器人的复杂曲面磨抛关键技术综述[J]. 机械工程学报，2024，60(7)：1-21. DENG Jianxin，YUAN Bangyi，HUANG Qiulin，et al. Review on the key technologies of complex surfaces polishing based on robots[J]. Journal of Mechanical Engineering，2024，60(7)：1-21.
[5] SHIM J，LEE S，JEON D，et al. Contact force estimation using uncertain torque model and friction models for robot manipulator[J]. IEEE Transactions on Industrial Electronics，2024，71(10)：12634-12644.
[6] HE J，SHEN M J，GAO F，et al. Correction：active compliance control of a position-controlled industrial robot for simulating space operations[J]. Chinese Journal of Mechanical Engineering，2023，36(1)：14.
[7] ZHANG T，LIANG X H. Disturbance observer-based robot end constant contact force-tracking control[J]. Complexity，2019，2019(1)：5802453.
[8] LIANG X，YAN Y C，WANG W Q，et al. Adaptive human-robot interaction torque estimation with high accuracy and strong tracking ability for a lower limb rehabilitation robot[J]. IEEE/ASME Transactions on Mechatronics，2024，29(6)：4814-4825.
[9] 李元，孙智，訾斌，等. 基于改进高阶无迹卡尔曼滤波的膝踝康复矫形器预测控制与实验研究[J]. 机械工程学报，2025，61(11)：57-71. LI Yuan，SUN Zhi，ZI Bin，et al. Predictive control and experimental study of knee ankle rehabilitation orthosis based on improved high-order unscented kalman filters[J]. Journal of Mechanical Engineering，2025，61(11)：57-71.
[10] DONG J W，XU J M，WANG L，et al. External force estimation of the industrial robot based on the error probability model and SWVAKF[J]. IEEE Transactions on Instrumentation and Measurement，2022，71：1-11.
[11] LIU S C，WANG L H，WANG X V. Sensorless force estimation for industrial robots using disturbance observer and neural learning of friction approximation[J]. Robotics and Computer-Integrated Manufacturing，2021，71：102168.
[12] HAN L Y，MAO J L，CAO P F，et al. Toward sensorless interaction force estimation for industrial robots using high-order finite-time observers[J]. IEEE Transactions on Industrial Electronics，2021，69(7)：7275-7284.
[13] LIM D，KIM M J，CHA J，et al. MOB-Net：Limb-modularized uncertainty torque learning of humanoids for sensorless external torque estimation[J]. The International Journal of Robotics Research，2025，44(1)：96-128.
[14] KO D K，LEE K W，LEE D H，et al. Vision-based interaction force estimation for robot grip motion without tactile/force sensor[J]. Expert Systems with Applications，2023，211：118441.
[15] ZHANG Y P，LING Z Q，CAO G Z，et al. A fast calibration method for an sEMG-Based lower limb joint torque estimation model[J]. Biomedical Signal Processing and Control，2024，93：106188.
[16] ZHANG T，LIANG X H，ZOU Y B. Robot peg-in-hole assembly based on contact force estimation compensated by convolutional neural network[J]. Control Engineering Practice，2022，120：105012.
[17] JUNG J，YOU S，KIM D，et al. Variable stiffness control via external torque estimation using LSTM[C]// 2022 International Conference on Robotics and Automation (ICRA). IEEE，2022：8325-8330.
[18] SONG Y S，HUANG H L，LIU F，et al. Torque estimation for robotic joint with harmonic reducer based on deformation calibration[J]. IEEE Sensors Journal，2019，20(2)：991-1002.
[19] PENG G Z，CHEN C L P，HE W，et al. Neural- learning-based force sensorless admittance control for robots with input deadzone[J]. IEEE Transactions on Industrial Electronics，2020，68(6)：5184-5196.
[20] WEI Y R，LYU S，LI W S，et al. Contact force estimation of robot manipulators with imperfect dynamic model：on Gaussian process adaptive disturbance Kalman filter[J]. IEEE Transactions on Automation Science and Engineering，2023，21(3)：3524-3537.
[21] 从永正，程盈盈，胡正. 基于三阶超螺旋滑模观测器算法的机器人力矩估计方法[J]. 控制理论与应用，2023，40(7)：1208-1215. CONG Yongzheng，CHENG Yingying，HU Zheng. Torque estimation algorithm of robot based on third-order super-twisting sliding mode observer[J]. Control Theory & Applications，2023，40(7)：1208-1215.
[22] ZHANG T，LI H C，SHI Y P，et al. A novel joint external torque estimate model of the lightweight robot’s joint based on a BP neural network[J]. Robotica，2025，43(3)：1027-1042.
[23] LI Z Y，WEI H C，LIU C G，et al. A high-order finite-time observer for external force estimation of collaborative robots[J]. International Journal of Robust and Nonlinear Control，2025，35(8)：2952-2967.
[24] LONG S K，DANG X J，SUN S L，et al. A novel sliding mode momentum observer for collaborative robot collision detection[J]. Machines，2022，10(9)：818.
[25] CHEN S Y，XIAO H，QIU L M，et al. Robotic flexible collision detection based on second-order sliding-mode momentum observer[C]// 202410th International Conference on Electrical Engineering，Control and Robotics (EECR). IEEE，2024：1-7.
[26] 郭万金. 复杂形状零部件打磨作业机器人研究[D]. 哈尔滨：哈尔滨工业大学，2017. GUO Wanjin. Research on robot manipulator for complex shaped parts deburring[D]. Harbin：Harbin Institute of Technology，2017.
[27] WU J H，LI W，XIONG Z H. Identification of robot dynamic model and joint frictions using a baseplate force sensor[J]. Science China Technological Sciences，2022，65(1)：30-40.
[28] HUANG Y J，KE J H，ZHANG X M，et al. Dynamic parameter identification of serial robots using a hybrid approach[J]. IEEE Transactions on Robotics，2022，39(2)：1607-1621.
[29] DONG J W，XU J M，ZHOU Q Q，et al. Dynamic identification of industrial robot based on nonlinear friction model and LS-SOS algorithm[J]. IEEE Transactions on Instrumentation and Measurement，2021，70：1-12.
[30] JIANG Y，YU Y F，ZHANG H Y，et al. Research on application of compliant control of smart meter box handling robot[J]. Energy Reports，2023，9：1494-1501.
[31] PENG G Z，YANG C G，HE W，et al. Force sensorless admittance control with neural learning for robots with actuator saturation[J]. IEEE Transactions on Industrial Electronics，2019，67(4)：3138-3148.
[32] GAROFALO G，MANSFELD N，JANKOWSKI J，et al. Sliding mode momentum observers for estimation of external torques and joint acceleration[C]// 2019 International Conference on Robotics and Automation (ICRA). IEEE，2019：6117-6123.
[33] MANZANILLA A，IBARRA E，SALAZAR S，et al. Super-twisting integral sliding mode control for trajectory tracking of an Unmanned Underwater Vehicle[J]. Ocean Engineering，2021，234：109164.