SDShuzhen Diao

Shuzhen Diao

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Shuzhen Diao has a total of 23 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Double-Loop Fuzzy Neural Network-Based Fixed-Time Robust Control for Antagonistic PM-Actuated Wrist Robots With Motion Constraints

    by Yuexuan Xu, Shuzhen Diao, Tong Yang, Xinlin Zhang, Ming Li, Yakun Gao, David Navarro-Alarcon, Ning Sun

    Antagonistic pneumatic muscle (PM)-actuated wrist robots have great potential in rehabilitation and industrial applications. The antagonistic connection of PMs, which mimics the agonist-antagonist muscle pairs in human joints, provides substantial advantages such as improved joint stability and a better balance of torque disturbances. However, PM-actuated robots exhibit complex nonlinearities, such as hysteresis, creep, input delay, and time-varying parameters, while also confronting challenges such as external disturbances and coupling effects. In this paper, a switching non-singular terminal sliding mode control (NTSMC) method with a double-loop fuzzy neural network (DLFNN) is developed. This method enables the antagonistic PM-actuated wrist robots to achieve fast and precise tracking performance. Specifically, the lumped disturbances are estimated online using the DLFNN, which can adaptively adjust the weight of the inner and outer layers, achieving accurate approximation and robustness. Based on the estimated value of disturbances, a switching NTSMC is implemented to ensure that tracking errors converge to the origin within the fixed time. Switching functions guarantee fast convergence when the sliding surface errors are large. Meanwhile, switching functions ensure non-singularity as the sliding surface errors converge to the origin. Furthermore, joint angles and angular velocities are limited within the specific ranges by designing exponential constraint terms as time-varying proportional-differential gains, rather than traditional barrier functions that may induce excessive control inputs. Both detailed stability analysis and experimental validation demonstrate the effectiveness and adaptability of the proposed method.

  • Output Feedback Control for PAM-Actuated Parallel Robots With Interval Type-2 Fuzzy Neural Networks

    by Qihang Wang, Shuzhen Diao, Gendi Liu, Tong Yang, Yongsheng Wang, Yanding Qin, Ning Sun

    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.


  • Modeling and control for PAM-actuated robots: an overview

    by Xinlin Zhang, Shuzhen Diao, Tong Yang, Yanding Qin, Ning Sun

    With the increasing integration of robots into various fields such as industry, agriculture, specialized industries, and home services, research on human–robot coexistence and collaboration has become more urgent, which is crucial for enhancing the operational capabilities of robots and ensuring personal safety. As a kind of soft actuator with smart structures, pneumatic artificial muscles (PAMs), known for their high power-to-weight ratio, ease of installation, clean energy usage, and resistance to electromagnetic interference, can form a series of PAM-actuated robots combined with rigid robotic architecture, endowing them with the ability to coordinate rigidity and flexibility. Recent advancements in the modeling and control methods of PAM-actuated robots have significantly improved their practical application value. However, there is a lack of systematic and structured review papers that explore the differences and connections among various modeling and control methods for PAM-actuated robots, and analyze the progression of their development. This paper will provide a detailed discussion of the key characteristics of various modeling and controller design for PAM-actuated robots, and offer guidance for the next generation of research directions in this field.


  • Velocity Feedback-Free Synchronous Control for Pneumatic Artificial Muscle-Actuated Parallel Robots With Fractional-Order Hysteresis

    by Shuzhen Diao, Gendi Liu, Xinlin Zhang, Jing Zhao, Yanding Qin, Wei Sun, Ning Sun

    Pneumatic artificial muscles (PAMs) can reduce the weight of robots and endow them with safer flexibility, by leveraging advantages such as large force-weight ratios and controllable stiffness. However, their input-output relationship also exhibits hysteresis, which is detrimental to the control performance of robots. The existing hysteresis models of PAMs usually have complex structures, and the accuracy depends on the number of model parameters. Therefore, by establishing a modified fractional-order Bouc-Wen (MFOBW) model, this article develops a synchronous controller with hysteresis compensation for PAM-actuated parallel robots (PAMPRs), which quickly reconstructs the velocities of master arms and achieves their synchronization and trajectory tracking. Notably, fractional calculus is employed to capture the rate-dependent hysteresis of PAMs, effectively simplifying the model structure, and the added auxiliary term further improves model accuracy; meanwhile, the designed velocity estimation system containing sign functions not only reconstructs unmeasurable velocities, but also accelerates the convergence rate without producing chattering. To our knowledge, this article provides the first MFOBW model for PAMs’ hysteresis, and the first finite-time synchronous controller for PAMPRs without velocity information. Additionally, the closed-loop stability is proven through theoretical analysis, and the MFOBW model’s effectiveness and the proposed method’s feasibility are verified through hardware experiments.


  • Admittance-Based Output Feedback Fuzzy Switching Control for PAM-Driven Parallel Robots via Nonsingular Terminal Sliding Mode

    by Xinlin Zhang, Gendi Liu, Shuzhen Diao, Tong Yang, Yongchun Fang, Ning Sun

    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.


  • LSTM-NN-Enhanced Tracking Control for PAM-Driven Parallel Robot Systems With Guaranteed Performance

    by Xinlin Zhang, Shuzhen Diao, Tong Yang, Yongchun Fang, Ning Sun

    Mechanical systems often face unpredictable surrounding situations in applications, which bring lots of intangible uncertainties into system operations. Further, some robot systems, especially, pneumatic artificial muscle (PAM)-driven robot systems, also have accumulative nonlinearities, such as rate-dependent hysteresis, creep, and periodic/regular time-varying parameters, increasing design difficulties of high-accuracy controllers. This paper develops a long short-term memory neural network (LSTM-NN)-enhanced adaptive controller for PAM-driven parallel robot systems with transient and steady-state performance constraints. Specifically, a continuous-time LSTM-NN structure is introduced to recover unknown lumped dynamics, improving the approximation ability of time-dependent terms with accumulative effects. Moreover, a new two-stage error transformation function is designed to flexibly adjust the desired transient and steady-state performance, facilitating better adaptation to task requirements. To our knowledge, this paper proposes the first solution of utilizing the LSTM-NN-based neuroadaptive method for soft actuator-driven robots to enhance tracking accuracy with transient/steady-state performance improvement. The detailed stability analysis and several groups of experimental results on the self-built platform are provided to verify the feasibility and versatility of the proposed method.


  • Finite-Time Synchronous Motion Control for Pneumatic Muscle-Actuated Parallel Robots in Cartesian Space

    by Shuzhen Diao, Gendi Liu, Tong Yang, Menghua Zhang, Yanding Qin, Wei Sun, Ning Sun

    Pneumatic muscles, being biomimetic actuators with excellent flexibility and high power density, are highly favored in the field of flexible robots. Nevertheless, the accompanying problems, including high nonlinearities, time-varying parameters, hysteresis, etc., also bring a lot of trouble to their motion control. Moreover, for multiple motion chain systems, such as pneumatic muscle-actuated (PMA) parallel robots, most existing results have not considered the motion synchronization problem among motion chains, adversely affecting the overall motion performance. Therefore, this article develops a synchronous motion control (SMC) method for PMA parallel robots in Cartesian space, which not only guarantees the asymptotic tracking performance of end effectors, but also makes tracking errors and synchronization errors converge to near the origin in the adjustable finite time. Specifically, the proposed method compensates for the synchronization errors of end effectors in the coordinate axis directions, significantly improving motion accuracy; meanwhile, the elaborately modified coupling errors can successfully avoid the singularity problem at special positions. For all our knowledge, this article proposes the first finite-time adaptive SMC method analyzed and designed in Cartesian space, to obtain rapid and accurate tracking performance without prior knowledge of model information. Finally, rigorous theoretical analysis is provided through the extended finite-time stability theorem, and adequate hardware experiments are carried out to verify the feasibility of the proposed method.


  • Supervised Learning Control for Compliant Pneumatic Artificial Muscle Robots With Preassigned-Time Performance

    by Gendi Liu, Shuzhen Diao, Tong Yang, Xinlin Zhang, Yongchun Fang, Ning Sun

    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.