Xuefeng Bao's research while affiliated with University of Wisconsin - Milwaukee and other places

Publications (19)

Conference Paper
Full-text available
In this paper, we proposed an assist-as-needed (AAN) control strategy that incorporated the continuous volition ankle joint torque prediction for a cable-driven ankle exoskeleton. The volitional torque was predicted via an sEMG-US imaging-driven Hill-type neuromuscular model (HNM) with the shank muscle activation estimation by fusing sEMG and US im...
Article
Full-text available
A hybrid exoskeleton comprising a powered exoskeleton and functional electrical stimulation (FES) is a promising technology for restoration of standing and walking functions after a neurological injury. Its shared control remains challenging due to the need to optimally distribute joint torques among FES and the powered exoskeleton while compensati...
Article
Full-text available
A hybrid exoskeleton that combines functional electrical stimulation (FES) and a powered exoskeleton is an emerging technology for assisting people with mobility disorders. The cooperative use of FES and the exoskeleton allows active muscle contractions through FES while robustifying torque generation to reduce FES-induced muscle fatigue. In this a...
Article
A hybrid neuroprosthesis system is a promising rehabilitation technology to restore lower limb function in persons with paraplegia. The technology combines functional electrical stimulation (FES) and a powered lower limb exoskeleton to produce movements for walking and standing. The main control challenge in the hybrid neuroprosthesis is to achieve...
Chapter
In this paper a lower-limb powered exoskeleton is combined with functional electrical stimulation of the quadriceps muscle to achieve a standing-up motion. As two actuation mechanisms (FES and the motors) act on the knee joints, it is desirable to optimally coordinate them. A feedback controller that stabilizes the desired standing-up motion is der...
Article
Currently controllers that dynamically modulate functional electrical stimulation (FES) and a powered exoskeleton at the same time during standing-up movements are largely unavailable. In this paper, an optimal shared control of FES and a powered exoskeleton is designed to perform sitting to standing (STS) movements with a hybrid exoskeleton. A hie...
Conference Paper
Full-text available
In this paper, a novel neural network based iterative learning controller for a hybrid exoskeleton is presented. The control allocation between functional electrical stimulation and knee electric motors uses a model predictive control strategy. Further to address modeling uncertainties, the controller identifies the system dynamics and input gain m...
Article
Full-text available
Functional electrical stimulation (FES) has recently been proposed as a supplementary torque assist in lower-limb powered exoskeletons for persons with paraplegia. In the combined system, also known as a hybrid neuroprosthesis, both FES-assist and the exoskeleton act to generate lower-limb torques to achieve standing and walking functions. Due to t...
Article
Functional electrical stimulation (FES) is prescribed as a treatment to restore motor function in individuals with neurological impairments. However, the rapid onset of FES-induced muscle fatigue significantly limits its duration of use and limb movement quality. In this paper, an electric motor assist is proposed to alleviate the fatigue effects b...
Article
In this paper a Lyapunov-based model predictive control (LMPC) method to control knee extension during an input-delayed neuromuscular electrical stimulation is developed. This method incorporates a contractive constraint under a delay compensation control law that achieves system stability despite an unknown constant input delay and imperfectly est...
Article
In this paper, a robust iterative learning switching controller that uses optimal virtual constraint is designed for a hybrid walking exoskeleton that uses functional electrical stimulation and a powered exoskeleton. The synthesis of iterative learning control with sliding-mode control improves tracking performance and accuracy. The motivation for...
Preprint
Equipping approximate dynamic programming (ADP) with input constraints has a tremendous significance. This enables ADP to be applied to the systems with actuator limitations, which is quite common for dynamical systems. In a conventional constrained ADP framework, the optimal control is searched via a policy iteration algorithm, where the value und...
Article
Reinforcement learning based adaptive/approximate dynamic programming (ADP) is a powerful technique to determine an approximate optimal controller for a dynamical system. These methods bypass the need to analytically solve the nonlinear Hamilton-Jacobi-Bellman equation, whose solution is often to difficult to determine but is needed to determine th...
Article
A hybrid neuroprosthesis that combines human muscle power, elicited through functional electrical stimulation (FES), with a powered orthosis may be advantageous over a sole FES or a powered exoskeleton-based rehabilitation system. The hybrid system can conceivably overcome torque reduction due to FESinduced muscle fatigue by complementarily using t...
Article
Introduction: Optimal frequency modulation during functional electrical stimulation (FES) may minimize or delay the onset of FES-induced muscle fatigue. Methods: An offline dynamic optimization method, constrained to a modified Hill-Huxley model, was used to determine the minimum number of pulses that would maintain a constant desired isometric...

Citations

... Direct comparison with prior implementations of hybrid exoskeletons that combine neural stimulation with external motors is difficult due to the variety of outcome measures and activities reported in the literature [14,[40][41][42]; however, some generalizations can be made. Similar to other implementations, the system effectively learned over time, reduced errors [27], and adapted to simulated fatigue represented by a gross decline in muscle force-generating capacities [26]. ...
... First, the velocity response is anticipated to be 10 RPM with insignificant variation in each condition. From (19), the estimated load torque should be equal to zero while m load = 0 kg of case A. Thus, the dashed line presents the theoretical value shown in Figure 20c, whereas the dotted line is the load torque estimated by the proposed disturbance observer in case A. Obviously, the dotted line is approximate to the dashed line. Then, the estimated load torques tested by standard weights in case B (m load = 0.5 kg) and case C (m load = 1 kg) are shown in Figures 21c and 22c, where the estimated load torque (dotted line) is close to the theoretical results (dashed line). ...
... Compared to our previous simulation studies by Molazadeh et al. (2019), Bao et al. (2020b), and Molazadeh et al. (2018a,b), the study presents a more detailed derivation of the controller, improved robustness to modeling uncertainties, and supporting stability analysis. Furthermore, extensive sit-to-stand experiments with a hybrid exoskeleton validated the approach on four participants with no disabilities. ...
... The system then switches between motor or muscle activation to control the joint depending on an estimate of muscle fatigue. This method is extended in [27], where a neural network-based ILC is applied to learn the system dynamics but requires an additional model predictive controller to allocate control effort between the redundant actuators. Major limitations of some of these controller designs include (1) requiring a good model of the system, which is difficult to assemble and changes from user to user; (2) the need for additional system identification, which could be expensive for whole-body systems; and (3) the ILC only being applied to a subsystem instead of being the guiding control architecture. ...
... The motion of the corresponding joints between the robots, that is, the joint angle generated by the motion planning [13] is used as the tracking control target to realize the coordinated motion relationship between the hip and knee joints of the hybrid robot, thereby driving the human body for rehabilitation training. Model-based control methods such as PID control [14], computed torque control [15], robust control [16] and modelbased feedback control [17] are often adopted for the trajectorytracking; hence, the accuracy of the model has a significant influence on the control. However, because of static structural errors, such as machining error and assembly error [18], and uncertainty of relevant dynamical parameters [19], accurate modelling is difficult. ...
... The main challenge, however, is to optimally allocate control effort between FES and the electric motor. In our previous research, an offline optimization method was used to optimize a hybrid walking system that uses FES and a passive orthosis [18] and a hybrid leg extension neuroprosthesis [19]. Motivated to optimize FES control in real time, in [20], a nonlinear model predictive control (NMPC) to elicit knee extension was developed. ...
... Carron et al. applied the combination to a compliant 6-DOF robotic arm, where an inverse dynamics FL was used to obtain a discrete-time linar model, the extended Kalman filter was used to estimate the states, and a discrete-time MPC was incorporated with the model [11]. Bao et al. presented the combination to a hybrid neuroprosthetic system, where an FL was used to reduce computational loads, and then an MPC was applied through a barrier cost function to deal with the nonlinear input constrains, originally converted from linear ones [12]. Chen et al. presented the combination in cascade and applied it to the control of automotive fuel cell oxygen excess ratio, where an FL cascaded with a continuous-time MPC was used to perform anti-disturbance control. ...
... In [26], ILC estimates system dynamics to inform a sliding-mode controller. The system then switches between motor or muscle activation to control the joint depending on an estimate of muscle fatigue. ...
... In Eq. 9, u M,k represents the motor input and u F,k is the FES input. In Eq. 11, U ∈ [0, 1] × [t r , t r + t p ] is the input constraint (Kirsch et al. (2018); Sun et al. (2018)). Subscription r is the receding horizon value, for example, t r shows the time in the rth receding horizon. ...
... As compared to FFNNs, RNN, which is composed of feedforward neural and feedback loops, can acquire more system information [35][36][37][38]. As shown in Figure 5, a RNN, which includes an input layer, a hidden layer, and an output layer, is adopted to compensate the uncertainties in the wind-turbine generation system and described as follows: ...