Muscle-like contraction control of tendon-sheath artificial muscle☆
Introduction
The skeletal muscle is a biological actuator that can naturally and safely drive human movement, it has soft structures and high power density. Inspired by skeletal muscles, various artificial actuators with compliant structures have been developed for robotic applications. Many studies have shown that the elastic actuator composed of an elastic mechanism and a motor-reducer actuator has high compliance and energy efficiency. As the elastic mechanism can be placed in series or parallel and can have variable stiffness, the currently developed elastic actuators include series elastic actuators [1], [2], parallel elastic actuators [3], variable stiffness actuators [4], [5], and series–parallel elastic actuators [6]. However, the elastic actuators have poor flexibility as using rigid structures. To achieve a muscle-like compliant structure, many artificial muscles with soft structures, light weight, and small size were proposed in recent years. They are made of various materials and have different contraction mechanisms [7]. The pneumatic artificial muscle (PAM) is one of the most commonly used artificial muscles, usually consisting of a rubber or silicone bladder and a hard braided sheath. With the restraint of the outer braided sheath, the PAM can contract by inflating the inner bladder. PAMs have high power density and soft structures, and have been widely used in many robotic applications [8], [9]. However, the air pumps are too bulky for portable mechanisms.
In recent studies, smart and composite materials have shown their potential in developing artificial muscles, including shape memory alloy (SMA), ionic polymer-metal composite (IPMC), dielectric elastomer, carbon nanotubes, nylon fiber, etc. They can generate a recoverable deformation under thermal, electrical, or chemical stimulation. The SMA can deform to its original shape after being heated [10]. The ions in the IPMC can migrate under a low voltage (1–5 V), resulting in a macroscopic deformation [11]. The dielectric elastomer can produce large deformation once a high voltage is applied (up to thousands of volts) [12], [13]. The twisted carbon nanotube yarns can produce high-speed rotation or large deformation when heated or electrified [14], [15]. Similarly, the highly twisted nylon fibers, such as sewing threads or fishing lines, can generate linear or torsional motion when heated, and have a large power-to-weight ratio [16], [17]. However, these materials have some shortcomings in actuation, including slow response, high excitation voltage, and limited tension and deformation. These shortcomings make them unsuitable for robotic applications that require precision motion control and high power output.
In addition to these mentioned shortages, current artificial muscles also lack muscle contraction property. The unique active contraction property is essential for the biological muscle to be the best natural actuator to adapt and balance the external loads. To improve the safety and comfort of human–robot interaction systems (such as orthotics, exoskeleton and prosthetic devices), an artificial muscle with more muscle-like behaviors should be developed.
Related researches have shown that muscle contraction can be well described by the Hill muscle model [18] which usually consists of three units, i.e., the contraction element (CE), the series elasticity (SE), and the parallel elasticity (PE). The SE and the PE are passive elasticities regulating muscle stiffness and energy storage [19], [20], [21], and have great significance in the efficiency and safety of joint movements [22]. The CE is the core element of the Hill muscle model and can generate active muscle contractions. Muscle contraction is a complex process that converts chemical energy into mechanical energy, its most important property is the relationship between muscle force, length, and velocity. With this regard, many Hill-type muscle models using different equations as the CE to describe the active contraction property.
Inspired by the Hill muscle model, this paper studies the muscle-like contraction strategy on the tendon-sheath artificial muscle proposed in [23]. Similar to the three elements of the Hill muscle model, the tendon-sheath artificial muscle takes a motor-driven tendon-sheath mechanism as the CE to generate active contraction, and uses two springs as the SE and the PE. The tendon-sheath artificial muscle has a flexible structure and resilience similar to biological muscles. In this paper, a muscle-like contraction control method is proposed to enable the tendon-sheath artificial muscle to contract like skeletal muscles. As Fig. 1 shows, the active contraction force of the tendon-sheath artificial muscle is adjusted by the activation , the muscle contraction length , and the contraction velocity . Considering the nonlinear transmission characteristics of the tendon-sheath artificial muscle and the inconvenience of measuring the force and length of the CE, the tendon-sheath artificial muscle transmission model is established to estimate the CE length and compensate for the friction loss in the force transmission.
The paper is organized as follows. First, the transmission characteristics of the tendon-sheath artificial muscle are studied, and the mathematical model is established in Section 2. Section 3 conducts experiments to validate the accuracy of the transmission model. To imitate the muscle contraction behavior, a muscle-like contraction control method based on the transmission model and the Hill muscle model is proposed in Section 4. In Section 5, two typical muscle contraction experiments are conducted to test the muscle-like contraction performance of the tendon-sheath artificial muscle. Finally, conclusions and future work are discussed in Section 6.
Section snippets
Modeling of the tendon-sheath artificial muscle
The tendon-sheath mechanism has severe hysteresis transmission characteristics due to the varied friction between the tendon and the sheath. In addition, the elastic springs further complicate the nonlinearity of the system and bring challenges to the control of the tendon-sheath artificial muscle. In this section, the tendon-sheath artificial muscle modeling is carried out based on a tendon-sheath transmission model.
Experimental setup
To validate the transmission model of the tendon-sheath artificial muscle, an experiment platform is set up as Fig. 4 shows, which is built on the Simulink Real-Time system. A servo motor (Yaskawa SGMAH-04AA41) is used for providing input force and displacement. Two tension sensors (JLBS-M2-30KG) record the input and output force and two laser position sensors (IL-300, KEYENCE) measure the input and output displacement. Control signals of the servo motor are sent by a PCL-726 board, and the
Hill-type muscle model
According to the Hill-type muscle model, the active contraction force of muscle is related to muscle length, contraction velocity, and muscle activation, as Eq. (17) shows. where represents muscle activation and ranges from 0 to 1. means no activation, and represents full activation. is the maximum muscle force, which refers to the isometric contraction force when the muscle is fully activated at its original length. is the normalized muscle
Muscle-like contraction experiment
To verify the muscle-like contraction control method, two representative muscle contraction experiments, i.e., isometric contraction experiment and quick-release experiment, were performed. The muscle parameters of the Hill-type muscle model are shown in Table 2, in which and of the soleus muscle in [44] were used. The experiment platform is presented in Fig. 8, as only the active contraction force is tested here, PE is not used in the experiment. The tendon-sheath transmission
Conclusions
In this paper, the muscle-like contraction control of the tendon-sheath artificial muscle is studied. A mathematical transmission model of the tendon-sheath artificial muscle was established and verified by experiments. The experimental results demonstrate that the model has good accuracy in predicting movement transmission with different movement frequencies or elasticities of the artificial muscle (average force prediction accuracy: 0.9624, RMSE 0.6443 N; average displacement
CRediT authorship contribution statement
Qi Zhang: Conceptualization, Methodology, Validation, Writing - original draft. Mingxing Yang: Methodology, Writing - original draft. Xiaopeng Shen: Validation, Investigation. Mengqian Tian: Supervision, Writing - review & editing. Xingsong Wang: Supervision, Project administration, Funding acquisition.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was supported by the National Natural Science Foundation of China under Grant 51575100 and 52005006; the Scientific Research Foundation of Graduate School of Southeast University, China under Grant YBPY1850; and the Postgraduate Research & Practice Innovation Program of Jiangsu Province, China under Grant KYCX18_0063.
Qi Zhang was born in Jiangsu Province, China, in 1994. She received the B.S. degree in mechanical engineering and automation from Nanjing University of Science and Technology, Nanjing, China, in 2015. She is currently working toward the Ph.D. degree in mechatronics engineering in the School of Mechanical Engineering, Southeast University, Nanjing, China.
Her major research interests include the tendon-sheath transmission theory, the artificial muscle actuator, and the application of limb
References (44)
- et al.
Design and analysis of spring parallel variable stiffness actuator based on antagonistic principle
Mech Mach Theory
(2019) - et al.
Development of continuum manipulator actuated by thin McKibben pneumatic artificial muscle
Mechatronics
(2019) - et al.
Hysteresis modeling and position control of tendon-sheath mechanism in flexible endoscopic systems
Mechatronics
(2014) - et al.
Modeling, online identification, and compensation control of single tendon sheath system with time-varying configuration
Mech Syst Signal Process
(2019) - et al.
Distal-end force prediction of tendon-sheath mechanisms for flexible endoscopic surgical robots using deep learning
Mech Mach Theory
(2019) - et al.
Modeling and inverse control of a compliant single-tendon-sheath artificial tendon actuator with bending angle compensation
Mechatronics
(2019) - et al.
Human–robot interaction control of rehabilitation robots with series elastic actuators
IEEE Trans Robot
(2015) - et al.
A compact, modular series elastic actuator
J Mech Robot
(2016) - Plooij M, Van Nunspeet M, Wisse M, Vallery H. Design and evaluation of the bi-directional clutched parallel elastic...
- et al.
Probabilistic elastic element design for robust natural dynamics of structure-controlled variable stiffness actuators
J Mech Robot
(2018)
Novel control strategy for the +SPEA: A redundant actuator with reconfigurable parallel elements
Mechatronics
Artificial muscles: Mechanisms, applications, and challenges
Adv Mater
Design of a lightweight soft robotic arm using pneumatic artificial muscles and inflatable sleeves
Soft Robot
Fundamentals on the use of shape memory alloys in soft robotics
Theoretical analysis and design for a multilayered ionic polymer metal composite actuator
J Intell Mater Syst Struct
Dielectric elastomer based grippers for soft robotics
Adv Mater
Dynamic modeling and feedforward control of jaw movements driven by viscoelastic artificial muscles
IEEE/ASME Trans Mechatronics
Electrically, chemically, and photonically powered torsional and tensile actuation of hybrid carbon nanotube yarn muscles
Science
Electrochemically powered, energy-conserving carbon nanotube artificial muscles
Adv Mater
Artificial muscles from fishing line and sewing thread
Science
Multidirectional artificial muscles from nylon
Adv Mater
Biomechanics: Mechanical properties of living tissues
Cited by (0)
Qi Zhang was born in Jiangsu Province, China, in 1994. She received the B.S. degree in mechanical engineering and automation from Nanjing University of Science and Technology, Nanjing, China, in 2015. She is currently working toward the Ph.D. degree in mechatronics engineering in the School of Mechanical Engineering, Southeast University, Nanjing, China.
Her major research interests include the tendon-sheath transmission theory, the artificial muscle actuator, and the application of limb rehabilitation exoskeleton.
Mingxing Yang was born in Shandong Province, China, in 1986. He received the M.S. degree in mechanical design and theory from Shandong University of Technology, Zibo, China, in 2014, and the Ph.D. degree from Southeast University, Nanjing, China, in 2020. He is currently an Associate Professor with Anhui University of Technology, Maanshan, China.
His major research interests include servo control of mechatronic systems, robust adaptive control, and powered exoskeleton robots.
Xiaopeng Shen was born in Jiangsu Province, China, in 1994. He received the B.S. degree in mechanical design, manufacturing, and automation from Southeast University, Nanjing, China, in 2016, where he is currently working toward the M.S. degree in mechatronics engineering in the School of Mechanical Engineering.
His major research interests include the control of tendon-sheath transmission and the construction of tendon-sheath artificial muscle.
Mengqian Tian was born in Shanxi Province, China, in 1971. She received the B.S. degree in mechanical engineering and automation from Northwestern Polytechnical University, Xi’an, China, in 1993, the M.S. degree in mechanical design from Tianjin University, Tianjin, China, in 1996, and the Ph.D. degree in Measurement technology and instrument from Southeast University, Nanjing, China, in 2007.
Since 1996, she has been a faculty member in the School of Mechanical Engineering at Southeast University, where she is currently an Associate Professor. Her major research interests includes the robotics and autonomous systems.
Xingsong Wang received the B.S. degree and M.S. degree from Zhejiang University, Hangzhou, China, in 1988 and 1991, respectively, and the Ph.D. degree from Southeast University, Nanjing, China, in 2000, all in mechanical engineering. He was a visiting scientist at Concordia University, Canada (2000.6-2000.12, 2001.9-2002.3) and at Purdue University, USA (2007.9-2008.3), both at School of Mechanical Engineering. Currently, he is a full professor in the School of Mechanical Engineering and head of Department of Mechatronics at Southeast University. He has published over 140 technique papers and 30 Chinese patents in these areas.
His current research interests include control theory with application in precision CNC machine tools, advanced mechatronics with applications in biomedical engineering, tendon-sheath transmission theory with application in rescue robots, and Mecanum-wheels based automatic guided vehicle systems designing.
- ☆
This paper was recommended for publication by Associate Editor Hamid Reza Karimi.