By mimicking the movement of biological muscles, pneumatic artificial muscles (PAMs) are developed as a novel type of bionic actuator known for their compliance and high safety; however, the inherent characteristics of PAMs (e.g., hysteresis and creep) increase the difficulty in modeling and control. Moreover, unmodeled dynamics in PAM-actuated parallel robots are unavoidable, which further complicates the efficient tracking task of PAM-actuated parallel robots. Therefore, we propose an output feedback controller with interval type-2 fuzzy neural networks (IT2FNNs) for PAM-actuated parallel robots to obtain satisfactory tracking results. Specifically, compared with most existing methods using the interval type-1 fuzzy neural network (NN), the IT2FNN used is more beneficial for dealing with unmodeled dynamics and system uncertainties on PAM-actuated parallel robots. Meanwhile, considering that most practical systems are often only equipped with displacement/angle sensors and lack velocity sensors, an observer is designed to estimate unmeasurable velocity signals. Next, based on Lyapunov techniques, the convergence of tracking errors is proven through theoretical analysis. To our knowledge, this article is the first to apply IT2FNNs with observation information to PAM-actuated parallel robots with unknown dynamics and unmeasurable velocity signals, and provides rigorous stability analysis. Further, several experiments are implemented, and the corresponding results illustrate the effectiveness and robustness of the proposed controller.

Pneumatic artificial muscle (PAM) actuators have passive compliance and vibration absorption capabilities, adapting to high-intensity human-robot interaction movements. Unfortunately, the asymmetric hysteresis of PAMs is prone to produce motion delays and control inaccuracy, and anti-disturbance control is not friendly when applied in exoskeleton robots. Also, most of existing studies on active compliant control are overly reliant on bulky force-sensing, which is limited by sampling accuracy and communication rate. Therefore, it is still a challenge to realize compliant motions of PAM systems, while ensuring the rapid convergence of output signals. To this end, a new practical finite-time compliant controller is designed in this paper, which realizes satisfactory tracking control of horizontal PAM systems. Specifically, the external contact force is estimated by an improved adaptive law, instead of sensor feedback, thus decreasing noise effects. Meanwhile, the proposed controller ensures that the output tracking error converges quickly within known finite time, while reducing computational complexity. In particular, the desired trajectory is updated in real time by a modified admittance model, so as to achieve motion compliance and interaction safety. Compared with the literature, it is the first attempt to provide a compliant control solution for horizontal PAM systems without force feedback information, and ensures the practical finite-time convergence of the output tracking error. The rigorous stability analysis is presented, and the reliability of the proposed method is verified by hardware experiments. Note to Practitioners—Owing to light weight and flexibility, pneumatic artificial muscle (PAM) actuators can better meet the growing demands of human-machine cooperation tasks. In areas such as power assist and rehabilitation training equipment, it is necessary to ensure that robots can properly adjust the motion trajectory according to the applied force, so as to replace unbearable “hard contact” with more dexterous “compliant interaction”. Inspired by this, a new compliant control method is designed in this paper for horizontal PAM systems, which realizes the online adjustment of desired motion trajectories according to contact forces, and ensures the rapid convergence of the tracking error within known finite time. In particular, the proposed method improves an efficient contact force estimation solution, avoiding bulky and expensive force-sensing, while reducing noise effects. Compared with existing results, this paper for the first time presents a compliant control method without sensor feedback of horizontal PAM systems, and improves the motion rapidity, accuracy, and stability in practical tracking and training cases. Meanwhile, rigorous stability analysis is provided by Lyapunov techniques, and the effectiveness of the proposed method is verified by experiments on a self-built PAM testbench. In the future, we will try to apply the proposed method to limb functional training scenarios, aiming to expand its practical prospects and increase efficiency.

As a kind of soft actuator with inherent compliance, pneumatic artificial muscles (PAMs) have great application potential in robots. However, some challenging issues, such as high nonlinearities, sensor noises, and external disturbances, inevitably bring enormous difficulties to the accurate control of PAM-driven robots. To this end, this paper proposes an adaptive output feedback fuzzy switching control method for switched-form PAM-driven parallel robot systems, utilizing admittance models to rebuild compliant trajectories. Specifically, based on the nonrecursive high-order sliding mode (HOSM) differentiators with fixed-time convergence, unmeasurable velocity signals can be reconstructed to eliminate the adverse effects of measurement noises, decreasing the time delay of feedback signals. Moreover, a soft switching strategy is designed to flexibly adjust the switching weights and intervals of fuzzy structures, maintaining smooth control commands. Further, by introducing a nonsingular terminal sliding manifold, tracking errors can rapidly converge to a small neighborhood around the origins within a finite time, and all closed-loop variables are proved to be bounded through the Lyapunov stability theory. Finally, several groups of experiments are carried out on a self-built PAM-driven parallel robot to verify the effectiveness of the suggested method.

With strong load capacity and high maneuverability of payload attitude regulation, dual rotary cranes (DRCs) are widely applied for transportation and assembly tasks in infrastructure construction. For DRCs, to achieve safe and accurate control of the payload position and attitude, it is necessary to enhance the motion synchronization of two cranes, under the premise of controlling more state variables with fewer control inputs based on nonlinear coupling dynamics; moreover, the finite-time convergence of positioning errors is also expected to be guaranteed for high efficiency. To this end, this paper proposes an assembly-oriented finite-time coordinated controllerwithoutany linearization to the nonlinear crane dynamics, which realizes accurate and stable regulation of the payload position and attitude through coordinated boom motions. To our knowledge, the proposed controller provides thefirstclosed-loop control solution to realize both horizontal and non-horizontal payload hoisting for DRCs based on practical assembly demands. Theoretically, through elaborate design of the synchronization error and coupling errors, the real-time information exchange between the two cranes is realized for thefirsttime, which enhances the boom motion synchronization while suppressing payload swings. Furthermore, by introducing continuous terminal sliding mode surfaces with a multi-layer nested structure, the finite-time convergence of the boom positioning errors and the synchronization error is ensured with chattering reduction. Additionally, rigorous closed-loop stability analysis is provided based on Lyapunov techniques and Barbalat’s Lemma. Finally, the effectiveness and robustness of the proposed controller are verified by hardware experimental results.Note to Practitioners—This paper is motivated by the coordinated motion control problem of dual rotary cranes (DRCs), which aims to achieve safe and accurate control of the payload position and attitude oriented on practical assembly demands. At present, most control methods for DRCs only realize horizontal payload transportation, which not only ignores the requirements of payload attitude regulation in assembly tasks, but also lacks the guarantee for boom motion coordination and the finite-time convergence of state variables. To address these issues, based on the nonlinear crane dynamicswithoutany linearization, this paper proposes an assembly-oriented finite-time coordinated controller for DRCs, which achieves precise and stable regulation of the payload position and attitude through coordinated boom motions, and simultaneously enhances the system rapidity by ensuring the finite-time convergence of boom positioning errors. Furthermore, the detailed controller design and stability analysis process is provided, and the effectiveness of the proposed method is verified by hardware experiments. In the future research, we will try to apply the proposed method to practical operations of DRCs for complex assembly tasks of large heavy objects.

Pneumatic artificial muscles (PAMs), as a kind of soft actuators, can overcome compliance limitations of traditional rigid actuators to improve the adaptability of robots. However, some inherent strong nonlinearities and time-varying properties of PAMs, e.g., complex hysteresis and creep, may lead to a lot of control problems. In addition, PAM-driven systems are also faced with input constraints (e.g., deadzones, saturations, and unidirectional inputs), unknown/unmodeled dynamics and external disturbances, which badly degrade the control performance and even cause accidents. Therefore, this paper proposes anewhysteresis compensation-based immersion and invariance (I&I) adaptive fuzzy control method for PAM-driven humanoid robot manipulators, which can approximate the unknown functions and estimate the unknown parameters. To our knowledge, this is thefirstmethod for PAM-driven systems that compensates for system nonlinearities (not onlycomplex hysteresis,but alsoinput deadzones) by utilizing thepriorinformation in inverse hysteresis models, andsimultaneouslyestimates the unknown functions and parameters by designing a fuzzy update law and a parameter update law based on I&I methodology, respectively, which increases thecontrol frequencyof systems and improves tracking performance during high-speed motions. Finally, we apply the proposed approach on a self-built PAM-driven humanoid robot manipulator to validate its effectiveness and robustness.Note to Practitioners—Faced with the practical requirements of robots that interact closely with humans, improving the adaptability and compliance of robots by utilizing soft actuators, such as PAMs, can satisfy current demands. Moreover, PAMs also have many expective characteristics (e.g., high power density, light material, low costs, clean power, etc.), which makes PAMs occupy an important status in the field of soft robotics. However, unknown parameters/structures, strong nonlinearities, and input constraints, may badly degrade the control performance of PAMs. Based on the above characteristics, this paper proposes anewhysteresis compensation-based adaptive fuzzy controller for PAM-driven humanoid robot manipulators, which realizesaccurate trackingcontrol during high-speed motions by using inverse hysteresis models to compensate for strong nonlinearities in PAMs, and a fuzzy update law and a parameter update law based on I&I methodology are utilized to estimate unknown parameters/structures. In addition, the proposed controller cansimultaneouslycompensate for input deadzones by utilizing the hysteresis information, which improves thecontrol frequencyof the manipulators, and can rapidly suppress tracking errors. Experimental results are provided to validate the effectiveness of the presented method. In the future, more effective compensation methods, such as rate-dependent hysteresis models, will further be carried out to compensate for system nonlinearities.

With the advantages of natural flexibility, large force-weight ratios, and green cleanliness, pneumatic artificial muscle (PAM) actuators that mimic biological skeletal muscles have attracted much attention. However, the inherent defects of PAMs, such as high nonlinearities, limited contraction lengths and frequencies, and multiple input constraints, pose significant challenges to the motion control of PAM-actuated parallel robots; meanwhile, most existing methods do not take into account motion constraints and working efficiency. To this end, a prescribed-time adaptive fuzzy motion control method is developed in this article, where PAM-actuated parallel robots can accurately achieve prescribed tracking performance within an allowable input pressure range. In particular, regardless of the initial values of target trajectories, the expected tracking accuracy is achieved within the prescribed time by restricting the tracking errors to the improved performance constraints; also, the motion velocities remain within the preset dynamic constraints, thereby improving the working safety and efficiency. To the best of authors' knowledge, this article presents thefirstadaptive fuzzy motion control method for PAM-actuatedparallelrobots, which cansimultaneouslyachieve motion constraints and prescribed tracking performance. Moreover, the stability of all signals is proved through theoretical analysis, and then the effectiveness of the proposed method is fully verified by a series of hardware experiments.

As pneumatic artificial muscles (PAMs) are similar to biological muscles in structure and movement mechanisms, parallel robots actuated by PAMs have development prospects in rehabilitation and industry, with advantages such as compliance, high safety, strong bearing capacity, and satisfactory dynamic performance. However, the parameter uncertainties and model complexity related to inherent characteristics of parallel robots actuated by PAMs (e.g., time-varying, coupling, hysteresis, creep, and high nonlinearity), bring challenges to accurate dynamic modeling and controller design. Therefore, to achieve satisfactory tracking performance, this article presents an adaptive compensation tracking controller with error constraints for parallel robots actuated by PAMs. The proposed controller deals with parameter uncertainties by estimating system parameters to ensure accurate tracking, which is indicated as an effective solution for a combination of PAMs and parallel robots. Furthermore, using desired trajectory signals in the complicated regression matrix, the online computational burden is significantly reduced. Moreover, to improve operation safety further, an auxiliary term with a theoretical demonstration guarantees that the tracking errors are maintained within allowable ranges. Then, the closed-loop stability is demonstrated by Lyapunov techniques. As far as we know, it is the first time that the challenges of parameter uncertainties, computational burdens, and error constraints of parallel robots actuated by PAMs are simultaneously addressed, which has both theoretical significance and practical value. Finally, the hardware experiments are implemented under different scenarios, and the results indicate that the proposed method achieves satisfactory tracking performance.

Pneumatic artificial muscle (PAM) actuators exhibit practical compliance and great payload-to-weight ratios when driving robotic exoskeletons. However, filling with highly compressed gas makes PAMs susceptible to sensor noises, which may degrade the state response and increase control efforts. In addition, most of the existing optimal controllers require linearized operations or complex network calculations. To this end, a supervised learning control method with preassigned-time performance is studied, which achieves satisfactory motion control of the compliant PAM robots. In particular, the utilized dynamic observer with time-varying gains significantly reduces the effect of observation noises, and enhances the state convergence speed by combining with the preassigned-time constraints. Simultaneously, the improved supervised learning algorithm further optimizes input air consumption, which only involves the iterative adjustment of network weights. In contrast to the literature, this article presents a new solution to minimize energy consumption of the compliant PAM robots, while ensuring that the output states converge within the preassigned time, independent of parameter design. Rigorous stability analysis is provided and several experiments validate the tracking efficacy of the proposed method.

Featured with high power density, excellent flexibility, shock absorption capacity, etc., pneumatic muscles (PMs) promote the development of exoskeleton robots and rehabilitation equipment. However, the complex nonlinearities of PMs limit efficiency optimization in closed-loop control, while the force-displacement coupling, soft materials, deficient workspace, etc., make it more difficult to simultaneously increase motion speeds and ensure the safety of multiple PM-actuated (PMA) robots. Although force sensors can currently be replaced by applying state estimation techniques, the amplification effects of measurement noises still compromise control accuracy and stability in practice. To this end, this article proposes a reinforcement learning-based robust motion control method with the prescribed performance, which achieves efficient and satisfactory tracking control for PMA robotic arms. In particular, by elaborately incorporating an integral term, a robust generalized proportional integral observer is used to eliminate measurement noises. Meanwhile, by using an actor–critic network to optimize control performance, an error-transformation-based continuous controller is designed to guarantee the uniformly ultimately boundedness of tracking errors. Compared with most existing methods, this article provides the first solution to restrict the entire transient and steady-state performance of PMA robotic arms, improve the noise suppression capability, and optimize the control efficiency simultaneously. Finally, complete stability analysis based on Lyapunov techniques is provided, and several groups of hardware experiments demonstrate the practicability and robustness of the proposed method.

Recently, the pneumatic artificial muscle (PAM) that can reproduce natural muscle functionalities has become one of the core actuator mechanisms of intelligent interactive soft robots. Unfortunately, some inherent defects (e.g., unignorable nonlinearities, hysteresis, low shrinkage frequencies, etc.) have limited the application progress of humanoid PAM arm robots. Additionally, the input constraints (e.g., saturations, dead zones, unidirectional inputs, etc.), unexpected external disturbances, unidentifiable system parameters, and inevitable unmodeled dynamics are usually complicated, which cannot be easily eliminated through existing adaptive control methods. This article proposes an adaptive fuzzy-sliding mode control method for humanoid PAM arm robotswithoutany information of precise model structures and system parameters, which can suppress the unexpected effects of complicated unknown functions and achieve high performance tracking control,simultaneously. To the best of our knowledge, the proposed controller is thefirstmethod for the humanoid PAM arm robots that considers the nonlinear input constraints including unidirectional conditions, saturations, and dead zones,simultaneously. Next, all the input constraints, system parameter uncertainties, unmodeled dynamics, and external disturbances can be estimated adaptively by utilizing the proposed fuzzy update law. Particularly, a sliding mode control law is designed to compensate possible fuzzy approximation errors, and rigorous Lyapunov-based stability analysis is provided to ensure that the state errors can converge to zero withinfinitetime. Hardware experiments are carried out later tovalidate the effectiveness and robustness of the proposed method.