Recent connections in the adaptive control literature to continuous-time analogs of Nesterov's accelerated gradient method have led to the development of new real-time adaptation laws based on accelerated gradient methods. However, previous results assume that the system's uncertainties are linear-in-the-parameters (LIP). To compensate for non-LIP uncertainties, our preliminary results developed a neural network (NN)-based accelerated gradient adaptive controller to achieve trajectory tracking for nonlinear systems; however, the development and analysis only considered single-hidden-layer NNs. In this article, a generalized deep NN (DNN) architecture with an arbitrary number of hidden layers is considered, and a new DNN-based accelerated gradient adaptation scheme is developed to generate estimates of all the DNN weights in real-time. A nonsmooth Lyapunov-based analysis is used to guarantee the developed accelerated gradient-based DNN adaptation design achieves global asymptotic tracking error convergence for general nonlinear control affine systems subject to unknown (non-LIP) drift dynamics and exogenous disturbances. A comprehensive set of simulation studies are conducted on a two-state nonlinear system, a robotic manipulator, and a complex 20-D nonlinear system to demonstrate the improved performance of the developed method. Our simulation studies demonstrate enhanced tracking and function approximation performance from both DNN architectures and accelerated gradient adaptation.

The rendezvous or position consensus problem is a fundamental topic within the multiagent system (MAS) literature and has numerous engineering applications. The majority of recent results that solve the position consensus problem rely on communication networks and each agent's ability to obtain position information via direct measurements and communication. Distributed coordination strategies for MASs that are network free and have agents that cannot directly measure position information are scarce. In this work, we develop a relay–explorer strategy to achieve position consensus at a common, desired location. In particular, a group of explorer agents, lacking global position sensors, use open-loop estimators of their position to independently dead reckon toward the desired location. To prevent the difference between the estimated and true position of each explorer from growing beyond a user-defined threshold, a mobile information service provider (relay agent) that is capable of measuring its position in the global coordinate frame, intermittently visits each explorer to provide position information as determined by a maximum dwell-time condition. The relay agent has a position estimator for each explorer, and each estimator is synchronized with the corresponding explorer. The relay agent uses these position estimators to locate each explorer, maneuver to them, and provide position feedback. The contribution of this work over our precursory results is the consideration of uncertain explorer agent dynamics, which are estimated online using recurrent neural networks and integral concurrent learning. The estimated dynamics serve as feed-forward model approximations in the position estimators used by the explorers and relay agent, which generate more accurate position estimates once a finite excitation condition is satisfied. The MAS is modeled as a switched system, and a Lyapunov-based analysis is used to derive a maximum dwell-time condition for each explorer, prove the MAS is exponentially regulated to the desired location, and show the error between the estimated and true explorer dynamics is uniformly ultimately bounded. Experiments and multiple simulation examples are provided to verify the theoretical development, explore the scalability and learning performance of the approach, and shed light on future extensions.

A network of distributed agents operating in a GPS-denied environment is tasked with tracking the position of a mobile asset. The agents use time-of-flight (ToF) measurements obtained from transmissions of a beacon signal by the asset to estimate the asset's position, but the estimates are noisy due to clock drift at the agents' local clocks. The clock drift can be reduced by having the asset and network of agents perform a distributed synchronization process; however, synchronization and localization cannot be performed simultaneously because the clock states are not well defined during synchronization and due to constraints on the radios' communication resources. The problem of optimizing when the agents should perform sensing or synchronization is formulated as a partially observed Markov decision process (POMDP) with continuous observations. Table-based Q-learning is used to search for an optimal policy on the belief space of the POMDP, but some form of approximation must be used because the belief space is a high -dimensional continuous space. We compare the performance of two approaches: 1) in replicated Q-learning, we learn a$Q$function that is a linear function of the continuous beliefs; 2) in triple-Q learning, the beliefs are replaced by a model with fewer parameters, and quantization is used on a continuous parameter. Simulations are used to compare the performance of these methods with the approach that is most commonly used, which is using periodic synchronization at an optimized, fixed rate.

This article introduces a Lyapunov-based nonlinear control scheme that mitigates false-data-injection (FDI) attacks in real time for a centralized multiagent system (MAS) with nonlinear Euler–Lagrange dynamics with additive disturbances and input delays. Since the state tracking error is faulty during FDI attacks, the central controller cannot solely rely on the state tracking error to coordinate the MAS. Therefore, a novel feedback signal is designed to address this challenge and to ensure that the MAS achieves the control objective. The proposed controller combines both learning and model-based approaches to estimate the agents’ states and to detect and respond to FDI attacks. A neural network (NN) is used to detect the FDI attacks, where the update laws for the NN weights are designed based on the corresponding stability analysis. Lyapunov–Krasovskii functionals are used in the Lyapunov-based stability analysis to ensure semi-global uniformly ultimately bounded tracking. A simulation of multiagent robots with nonlinear Euler–Lagrange dynamics is provided, demonstrating the promising performance of the developed method to respond to FDI attacks.

Closed-loop control of functional electrical stimulation coupled with motorized assistance to induce cycling is a rehabilitative strategy that can improve the mobility of people with neurological conditions (NCs). However, robust control methods, which are currently pervasive in the cycling literature, have limited effectiveness due to the use of high stimulation intensity leading to accelerated fatigue during cycling protocols. This paper examines the design of a distributed repetitive learning controller (RLC) that commands an independent learning feedforward term to each of the six stimulated lower-limb muscle groups and an electric motor during the tracking of a periodic cadence trajectory. The switched controller activates lower limb muscles during kinematic efficient regions of the crank cycle and provides motorized assistance only when most needed (i.e., during the portions of the crank cycle where muscles evoke a low torque output). The controller exploits the periodicity of the desired cadence trajectory to learn from previous control inputs for each muscle group and electric motor. A Lyapunov-based stability analysis guarantees asymptotic tracking via an invariance-like corollary for nonsmooth systems. The switched distributed RLC was evaluated in experiments with seven able-bodied individuals and five participants with NCs. A mean root-mean-squared cadence error of 3.58 ± 0.43 revolutions per minute (RPM) (0.07 ± 7.35% average error) and 4.26 ± 0.84 RPM (0.1 ± 8.99% average error) was obtained for the healthy and neurologically impaired populations, respectively.

In networked control systems (NCS), agents participating in a network share their data with others to work together. When agents share their data, they can naturally expose the NCS to layers of faults and cyber-attacks, which can contribute to the propagation of error from one agent/area to another within the system. One common type of attack in which adversaries corrupt information within a NCS is called a false data injection (FDI) attack. This article proposes a control scheme, which enables a NCS to detect and mitigate FDI attacks and, at the same time, compensate for measurement noise and process noise. Furthermore, the developed controller is designed to be robust to unknown inputs. The algorithm incorporates a Kalman filter as an observer to estimate agents' states. We also develop a neural network (NN) architecture to detect and respond to any anomalies caused by FDI attacks. The weights of the NN are updated using an extended Kalman filter, which significantly improves the accuracy of FDI detection. A simulation of the results is provided, which illustrates satisfactory performance of the developed method to accurately detect and respond to FDI attacks.

In this article, an infinite-horizon optimal regulation problem is considered for a control-affine nonlinear autonomous agent subject to input constraints in the presence of dynamic avoidance regions. A local model-based approximate dynamic programming method is implemented to approximate the value function in a local neighborhood of the agent. By performing local approximations, prior knowledge of the locations of avoidance regions is not required. To alleviate the a priori knowledge of the number of avoidance regions in the operating domain, an extension is provided that modifies the value function approximation. The developed feedback-based motion planning strategy guarantees uniformly ultimately bounded convergence of the approximated control policy to the optimal policy while also ensuring the agent remains outside avoidance regions. Simulations are included to demonstrate the preliminary development for a kinematic unicycle and generic nonlinear system. Results from three experiments are also presented to illustrate the performance of the developed method, where a quadcopter achieves approximate optimal regulation while avoiding three mobile obstacles. To demonstrate the developed method, known avoidance regions are used in the first experiment, unknown avoidance regions are used in the second experiment, and an unknown time-varying obstacle directed by a remote pilot is included in the third experiment.

We consider the problem of jammer placement to partition a wireless network, where the network nodes and jammers are located in the real plane. In previous research, we found optimal and suboptimal jammer placements by reducing the search space for the jammers to the locations of the network nodes. In this paper, we develop techniques to find optimal jammer placements over all possible jammer placements in the real plane. Our approach finds a set of candidate jammer locations (CJLs) such that a jammer-placement solution using the CJLs achieves the minimum possible cardinality among all possible jammer placements in the real plane. The CJLs can be used directly with the optimal and fast, suboptimal algorithms for jammer placement from our previous work.

A function approximation method is developed which aims to approximate a function in a small neighborhood of a state that travels within a compact set. The method provides a novel approximation strategy for the efficient approximation of nonlinear functions for real-time simulations and experiments. The development is based on the theory of universal reproducing kernel Hilbert spaces over the n -dimensional Euclidean space. Several theorems are introduced which support the development of this state following (StaF) method. In particular, it is shown that there is a bound on the number of kernel functions required for the maintenance of an accurate function approximation as a state moves through a compact set. In addition, a weight update law, based on gradient descent, is introduced where arbitrarily close accuracy can be achieved provided the weight update law is iterated at a sufficient frequency, as detailed in Theorem 4. An experience-based approximation method is presented which utilizes the samples of the estimations of the ideal weights to generate a global approximation of a function. The experience-based approximation interpolates the samples of the weight estimates using radial basis functions. To illustrate the StaF method, the method is utilized for derivative estimation, function approximation, and is applied to an adaptive dynamic programming problem where it is demonstrated that the stability is maintained with a reduced number of basis functions.

Autonomous agents are often tasked with operating in an area where feedback is unavailable. Inspired by such applications, this paper develops a novel switched system-based control method for uncertain nonlinear systems with temporary loss-of-state feedback. To compensate for intermittent feedback, an observer is used while state feedback is available to reduce the estimation error, and a predictor is utilized to propagate the estimates while state feedback is unavailable. Based on the resulting subsystems, maximum and minimum dwell time conditions are developed via a Lyapunov-based switched systems analysis to relax the constraint of maintaining constant feedback. Using the dwell time conditions, a switching trajectory is developed to enter and exit the feedback denied region in a manner that ensures that the overall switched system remains stable. A scheme for designing a switching trajectory with a smooth transition function is provided. Simulation and experimental results are presented to demonstrate the performance of control design.

In this paper, an indirect herding problem is considered for a single herder that is outnumbered by multiple target agents. Indirect herding is a problem that involves a network of one or more controllable herding agents and indirectly controlled target agents (i.e., the target agents can only be controlled through the herder by exploiting herder-target interactions), where the goal is to achieve a network-wide objective. This paper investigates the unique problem where a single herder is required to regulate all the target agents to some desired formation. The problem is further complicated by the fact that the target agents have uncertain nonlinear dynamics, including uncertainties in the herder-target interaction. Neural network function approximation methods are used along with switched systems methods to ensure uniformly ultimately bounded convergence of the agent trajectories provided the developed sufficient dwell-time conditions are satisfied. Simulation and experimental results that involve one herding agent and multiple target agents demonstrate the validity of the designed controller and function approximation scheme for multiple herding objectives.

A decentralized controller is designed for leader-based synchronization of communication-delayed networked agents. The agents have heterogeneous dynamics modeled by uncertain, nonlinear Euler-Lagrange equations of motion affected by heterogeneous, unknown, exogenous disturbances. The developed controller requires only one-hop (delayed) communication from network neighbors and the communication delays are assumed to be heterogeneous, uncertain, and time-varying. Each agent uses an estimate of communication delay to provide feedback of estimated recent tracking error. Simulation results are provided to demonstrate the improved performance of the developed controller over other popular control designs.

An infinite-horizon optimal regulation problem for a control-affine deterministic system is solved online using a local state following (StaF) kernel and a regional model-based reinforcement learning (R-MBRL) method to approximate the value function. Unlike traditional methods such as R-MBRL that aim to approximate the value function over a large compact set, the StaF kernel approach aims to approximate the value function in a local neighborhood of the state that travels within a compact set. In this paper, the value function is approximated using a state-dependent convex combination of the StaF-based and the R-MBRL-based approximations. As the state enters a neighborhood containing the origin, the value function transitions from being approximated by the StaF approach to the R-MBRL approach. Semiglobal uniformly ultimately bounded (SGUUB) convergence of the system states to the origin is established using a Lyapunov-based analysis. Simulation results are provided for two, three, six, and ten-state dynamical systems to demonstrate the scalability and performance of the developed method.

When using a camera to estimate the pose of a moving target, measurements may only be available intermittently, due to feature tracking losses from occlusions or the limited field of view of the camera. Results spanning back to the Kalman filter have demonstrated the utility of using a predictor to update state estimates when measurements are not available, but target velocity measurements or a motion model must be known to implement a predictor for image-based pose estimation. In this paper, a novel estimator and predictor are developed to simultaneously learn a motion model, and estimate the pose, of a moving target from a moving camera. A stability analysis is provided to prove convergence of the state estimates and function approximation without requiring the restrictive persistent excitation condition. Two experiments illustrate the performance of the developed estimator and predictor. One experiment involves a stationary camera observing a mobile robot with sporadic feature tracking losses, and a second experiment involves a quadcopter moving between two mobile robots on a road network.

Online approximation of the optimal station-keeping strategy for a marine craft subject to an irrotational current is considered. An approximate policy that minimizes a user-defined cost function over an infinite time horizon is obtained using an actor-critic-identifier-based adaptive dynamic programming technique. The hydrodynamic drift dynamics are assumed to be unknown; therefore, a concurrent learning-based system identifier is developed to identify the unknown model parameters. The identified model is used to implement an adaptive model-based reinforcement learning technique to estimate the unknown value function. The developed policy guarantees uniformly ultimately bounded convergence of the vehicle to the desired station and uniformly ultimately bounded convergence of the approximated policies to the optimal polices without the requirement of persistence of excitation. The developed strategy is validated using an autonomous underwater vehicle, where the three degrees-of-freedom in the horizontal plane are regulated. The experiments are conducted in a second-magnitude spring located in central Florida.

Wireless communication systems are susceptible to both unintentional interference and intentional jamming attacks. For mesh and ad-hoc networks, interference affects the network topology and can cause the network to partition, which may completely disrupt the applications or missions that depend on the network. Defensive techniques can be applied to try to prevent such disruptions to the network topology. Most previous research in this area is on improving network resilience by adapting the network topology when a jamming attack occurs. In this paper, we consider making a network more robust to jamming attacks before any such attack has happened. We consider a network in which the positions of most of the radios in the network are not under the control of the network operator, but the network operator can position a few “helper nodes” to add robustness against jamming. For instance, most of the nodes are radios on vehicles participating in a mission, and the helper nodes are mounted on mobile robots or UAVs. We develop techniques to determine where to position the helper nodes to maximize the robustness of the network to certain jamming attacks aimed at disrupting the network topology. Using our recent results for quickly determining how to attack a network, we use the harmony search algorithm to find helper node placements that maximize the number of jammers needed to disrupt the network.

We consider the problem of finding fast techniques to solve jammer placement problems, where the goal is to partition a wireless network and the communication and jamming effects can be modeled using graphs. We have previously developed optimal jammer placement techniques based on solving a binary linear program, but the complexity of the optimal solution is an exponential function of the number of nodes in the network. Several computationally efficient techniques are known for finding edge separators of networks, including spectral cuts and multilevel cuts. However, the jammer placement problem is more similar to finding a vertex separator and is complicated by the fact that a jammer can block communications to all of the neighbors of the jammer in a jamming graph. In this paper, we develop computationally efficient, but suboptimal, techniques for jammer placement that first find an edge cut using spectral or multilevel graph cuts and then find the minimum number of jammers required to block communication across all of the edges in the edge cut. The results show that the proposed techniques are orders of magnitude faster than even heuristic approaches to solving the optimum jammer placement problem, while offering different tradeoffs between the number of jammers required and the maximum order of the residual subnetworks. For networks of over 400 nodes, the solution based on multi-level graph partitioning uses fewer jammers, leaves smaller clusters, and requires lower computational complexity than the heuristic optimization of the optimal jammer placement problem.

Cycling induced by automatic control of functional electrical stimulation provides a means of therapeutic exercise and functional restoration for people affected by paralysis. During cycling induced by functional electrical stimulation, various muscle groups are stimulated according to the cycle crank angle; however, because of kinematic constraints on the cycle-rider system, stimulation is typically only applied in a subsection of the crank cycle. Therefore, these systems can be considered as switched control systems with autonomous, state-dependent switching with potentially unstable modes. Previous studies have included an electric motor in the system to provide additional control authority, but no studies have considered the effects of switched control in the stability analysis of the motorized functional electrical stimulation cycling system. In this paper, a model of the motorized cycle-rider system with functional electrical stimulation is developed that includes the effects of a switched control input. A novel switching strategy for the electric motor is designed to only provide assistance in the regions of the crank cycle where the kinematic effectiveness of the rider's muscles is low. A switched sliding-mode controller is designed, and global, exponentially stable tracking of a desired crank trajectory is guaranteed via Lyapunov methods for switched systems, despite parametric uncertainty in the nonlinear model and unknown, time-varying disturbances. Experimental results from five able-bodied, passive riders are presented to validate the control design, and the developed control system achieves an average cadence tracking error of$0.00\pm 2.91$revolutions per minute for a desired trajectory of 50 revolutions per minute.

Neuromuscular electrical stimulation (NMES) induces muscle contractions via electrical stimuli. NMES can be used for rehabilitation and to enable functional movements; however, a fundamental limitation is the early onset of fatigue. Asynchronous stimulation is a method that can reduce fatigue by utilizing multiple stimulation channels to segregate and switch between different sets of recruited motor units. However, switching between stimulation channels is challenging due to each channel's differing response to stimulation. To address this challenge, a switched systems analysis is used in the present work to design a controller that allows for instantaneous switching between stimulation channels. The developed controller yields semi-global exponential tracking of a desired angular trajectory for a person's knee-joint. Experiments were conducted in six able-bodied individuals. Compared to conventional stimulation, the results indicate that asynchronous stimulation with the developed controller yields longer durations of successful tracking despite different responses between the stimulation channels.

This brief paper provides an approximate online adaptive solution to the infinite-horizon optimal tracking problem for control-affine continuous-time nonlinear systems with unknown drift dynamics. To relax the persistence of excitation condition, model-based reinforcement learning is implemented using a concurrent-learning-based system identifier to simulate experience by evaluating the Bellman error over unexplored areas of the state space. Tracking of the desired trajectory and convergence of the developed policy to a neighborhood of the optimal policy are established via Lyapunov-based stability analysis. Simulation results demonstrate the effectiveness of the developed technique.

Switched systems theory is used to analyze the stability of image-based observers for three-dimensional localization of objects in a scene in the presence of intermittent measurements due to occlusions, feature tracking losses, or a limited camera field of view, for example. Generally, observers or filters that are exponentially stable under persistent measurement availability may have unbounded error growth under intermittent measurement loss, even while providing seemingly accurate state estimates. By constructing a framework that utilizes a state predictor during periods when measurements are not available, a class of image-based observers is shown to be exponentially convergent in the presence of intermittent measurements if an average dwell time, and a total unmeasurability time, condition is satisfied. The conditions are developed in a general form, applicable to any observer that is exponentially convergent assuming persistent visibility, and utilizes object motion knowledge to reduce the amount of time measurements must be available to maintain convergence guarantees. Based on the stability results, simulations are provided to show improved performance compared to a zero-order hold approach, where state estimates are held constant when measurements are not available. Experimental results are also included to verify the theoretical results, to demonstrate applicability of the developed observer and predictor design, and to compare against a typical approach using an extended Kalman filter.

An upper motor neuron lesion (UMNL) can be caused by various neurological disorders or trauma and leads to disabilities. Neuromuscular electrical stimulation (NMES) is a technique that is widely used for rehabilitation and restoration of motor function for people suffering from UMNL. Typically, stability analysis for closed-loop NMES ignores the modulated implementation of NMES. However, electrical stimulation must be applied to muscle as a modulated series of pulses. In this paper, a muscle activation model with an amplitude modulated control input is developed to capture the discontinuous nature of muscle activation, and an identification-based closed-loop NMES controller is designed and analyzed for the uncertain amplitude modulated muscle activation model. Semi-global uniformly ultimately bounded tracking is guaranteed. The stability of the closed-loop system is analyzed with Lyapunov-based methods, and a pulse frequency related gain condition is obtained. Experiments are performed with five able-bodied subjects to demonstrate the interplay between the control gains and the pulse frequency, and results are provided which indicate that control gains should be increased to maintain stability if the stimulation pulse frequency is decreased to mitigate muscle fatigue. For the first time, this paper brings together an analysis of the controller and modulation scheme.

An approximate online equilibrium solution is developed for an N -player nonzero-sum game subject to continuous-time nonlinear unknown dynamics and an infinite horizon quadratic cost. A novel actor-critic-identifier structure is used, wherein a robust dynamic neural network is used to asymptotically identify the uncertain system with additive disturbances, and a set of critic and actor NNs are used to approximate the value functions and equilibrium policies, respectively. The weight update laws for the actor neural networks (NNs) are generated using a gradient-descent method, and the critic NNs are generated by least square regression, which are both based on the modified Bellman error that is independent of the system dynamics. A Lyapunov-based stability analysis shows that uniformly ultimately bounded tracking is achieved, and a convergence analysis demonstrates that the approximate control policies converge to a neighborhood of the optimal solutions. The actor, critic, and identifier structures are implemented in real time continuously and simultaneously. Simulations on two and three player games illustrate the performance of the developed method.

We consider the problem of assigning a group of users with different rate requirements to a set of frequency bands, which may have different bandwidths, when the users access the channel via carrier-sense multiple access (CSMA). For example, in systems that employ dynamic spectrum access (DSA), secondary users may interleave their transmissions into space-time-frequency slots left open by primary users. The set of primary users that are active at a given time may leave a set of available channels that have unequal bandwidths. The use of CSMA by the secondary users reduces the need to coordinate transmissions among the users allocated to a particular frequency band, but it also results in potential collisions, which reduce the overall data rate that can be accommodated in the band. This is especially true when the presence of a finite sensing delay is considered. In this paper, we consider the layer problem of allocating users to the available frequency bands to minimize the bandwidth used (to accommodate other groups of secondary users), while taking into account the CSMA interactions of assigning multiple users to a band. We formulate this as a new form of bin packing problem, in which the size of the bin depends on the number of users that are assigned to the bin. A near optimal solution to this problem is found numerically using the Gurobi solver, and the performance is compared with the suboptimal first-fit algorithm, which has complexity that is appropriate for online implementation. Simulation results are provided to compare the optimal and online algorithms in terms of efficiency in allocating the bandwidth to the users and their complexity.

This study focuses on the development of a nonlinear control design for a fully-actuated autonomous underwater vehicle (AUV) using a continuous robust integral of the sign of the error control structure to compensate for system uncertainties and sufficiently smooth bounded exogenous disturbances. A Lyapunov stability analysis is included to prove semiglobal asymptotic tracking. The resulting controller is experimentally validated on an AUV developed at the University of Florida in both controlled and open-water environments.

Neuromuscular electrical stimulation (NMES) is a prescribed treatment for various neuromuscular disorders, where an electrical stimulus is provided to elicit a muscle contraction. Barriers to the development of NMES controllers exist because the muscle response to an electrical stimulation is nonlinear and the muscle model is uncertain. Efforts in this paper focus on the development of an adaptive inverse optimal NMES controller. The controller yields desired limb trajectory tracking while simultaneously minimizing a cost functional that is positive in the error states and stimulation input. The development of this framework allows tradeoffs to be made between tracking performance and control effort by putting different penalties on error states and control input, depending on the clinical goal or functional task. The controller is examined through a Lyapunov-based analysis. Experiments on able-bodied individuals are provided to demonstrate the performance of the developed controller.

Autonomous mobile robots, such as unmanned aerial or ground vehicles, will play important roles in future military and commercial applications. Future systems will require the robots to communicate with each other on a peer-to-peer basis over a wireless link and are able to form a networked system such as an ad hoc network. In this paper, we consider a problem of optimizing the topology of the network for such systems. Our solution consists of two steps. We first develop a novel method to select a network tree topology from arbitrary initial connected network topology. Second, we develop a new optimization algorithm to reconfigure the network, starting from the tree found in the first step, while maintaining the connectivity to minimize the aggregate traffic under the constraints on the total number of hops that the mobile robots may move. We present simulation results to compare the performance of our algorithms.

The method for camera motion estimation is proposed for the moving objects. Whereas the estimation of the structure and motion (SaM) of the moving objects usually involves the constraints on the motion of the camera and the object, the moving camera velocities can be estimated in our work using the images of the moving object from the single camera without any constraint on the camera and object motion. To this end, the dynamics of the partially measurable state are arranged in such a way that the recursive least-squares (RLS) algorithm can be employed for stationary objects and then the nonlinear observer based on RISE (robust integral signed error) method for dynamic objects sequentially. The proposed method has advantages in that when the proposed method and the previously developed SaM algorithms are combined together, we can reconstruct the 3-D structure of the moving objects from 2-D images from a single camera. Simulation results under time-varying velocities of both camera and object are presented to verify the proposed method.

This study introduces a novel visual servo controller that is designed to control the pose of the camera to keep multiple objects in the field of view (FOV) of a mobile camera. In contrast with other visual servo methods, the control objective is not formulated in terms of a goal pose or a goal image. Rather, a set of underdetermined task functions are developed to regulate the mean and variance of a set of image features. Regulating these task functions inhibits feature points from leaving the camera FOV. An additional task function is used to maintain a high level of motion perceptibility, which ensures that desired feature point velocities can be achieved. These task functions are mapped to camera velocity, which serves as the system input. A proof of stability is presented for tracking three or fewer targets. Experiments of tracking eight or more targets have verified the performance of the proposed method.

The control of dynamic systems that undergo an impact collision is both theoretically challenging and of practical importance. An appeal of studying systems that undergo an impact is that short-duration effects such as high stresses, rapid dissipation of energy, and fast acceleration and deceleration may be achieved from low-energy sources. However, colliding systems present a difficult control challenge because the equations of motion are different when the system suddenly transitions from a noncontact state to a contact state. In this paper, an adaptive nonlinear controller is designed to regulate the states of two dynamic systems that collide. The academic example of a planar robot colliding with an unactuated mass-spring system is used to represent a broader class of such systems. The control objective is defined as the desire to command a robot to collide with an unactuated system and regulate the mass to a desired compressed state while compensating for the unknown constant system parameters. Lyapunov-based methods are used to develop a continuous adaptive controller that yields asymptotic regulation of the mass and robot links. It is interesting to note that one controller is responsible for achieving the control objective when the robot is in free motion (i.e., decoupled from the mass-spring system), when the systems collide, and when the system dynamics are coupled.

An optimal extremum seeking approach is developed in this paper to identify frequency and voltage modulation parameters for a neuromuscular electrical stimulation control objective. The control objective is to externally apply optimally varied voltage or frequency modulation parameters to a human quadriceps muscle to generate a desired knee joint angle. Experimental results are provided to illustrate the limb positioning performance of a real-time extremum seeking routine (i.e., Brent's Method).

Previous homography-based visual servo controllers have been developed using an error system that contains a singularity resulting from the representation of the rotation matrix. For some aerospace applications such as visual servo control of satellites or air vehicles, the singularity introduced by the rotation representation may be restrictive. To eliminate this singularity, a homography-based visual servo controller is developed in this paper based on an error system composed of the unit quaternion representation. The proposed adaptive controller regulates a camera to a desired position and orientation that is determined from a desired image. A quaternion-based Lyapunov function is developed to facilitate the control design and the stability analysis

A visual servo tracking controller is developed in this paper for a monocular camera system mounted on an underactuated wheeled mobile robot (WMR) subject to nonholonomic motion constraints (i.e., the camera-in-hand problem). A prerecorded image sequence (e.g., a video) of three target points is used to define a desired trajectory for the WMR. By comparing the target points from a stationary reference image with the corresponding target points in the live image and the prerecorded sequence of images, projective geometric relationships are exploited to construct Euclidean homographies. The information obtained by decomposing the Euclidean homography is used to develop a kinematic controller. A Lyapunov-based analysis is used to develop an adaptive update law to actively compensate for the lack of depth information required for the translation error system. Experimental results are provided to demonstrate the control design.

Several factors must be considered for robotic task execution in the presence of a fault, including: detection, identification, and accommodation for the fault. In this paper, a nonlinear observer is used to identify a class of actuator faults once the fault has been detected by some other method. Advantages of the proposed fault-identification method are that it is based on the nonlinear dynamic model of a robot manipulator (and hence, can be extended to a number of general Euler Lagrange systems), it does not require acceleration measurements, and it is independent from the controller. A Lyapunov-based analysis is provided to prove that the developed fault observer converges to the actual fault. Experimental results are provided to illustrate the performance of the identification method.

A monocular camera-based vision system attached to a mobile robot (i.e., the camera-in-hand configuration) is considered in this paper. By comparing corresponding target points of an object from two different camera images, geometric relationships are exploited to derive a transformation that relates the actual position and orientation of the mobile robot to a reference position and orientation. This transformation is used to synthesize a rotation and translation error system from the current position and orientation to the fixed reference position and orientation. Lyapunov-based techniques are used to construct an adaptive estimate to compensate for a constant, unmeasurable depth parameter, and to prove asymptotic regulation of the mobile robot. The contribution of this paper is that Lyapunov techniques are exploited to craft an adaptive controller that enables mobile robot position and orientation regulation despite the lack of an object model and the lack of depth information. Experimental results are provided to illustrate the performance of the controller.

In this paper, the 3-dimensional (3D) position of a camera held by the end-effector of a cartesian robot manipulator is regulated to a constant desired position despite (i) the lack of depth information of the actual or desired camera position from a target, (ii) the lack of a 3D model of the target object, and (iii) the lack of information of intrinsic and extrinsic camera calibration matrices. Specifically, by fusing 2D image-space and 3D task-space information (i.e., 2.5D visual servoing), an adaptive controller is developed that is proven to ensure asymptotic position regulation of the cartesian robot. The stability of the proposed controller is proven through a rigorous Lyapunov-based stability analysis.

Several factors must be considered for robotic task execution in the presence of a fault, including: detection, identification, and accommodation for the fault. In this paper, a prediction error based dead-zone residual function and nonlinear observers are used to detect and identify a class of actuator faults. Advantages of the proposed fault detection and identification methods are that they are based on the nonlinear dynamic model of a robot manipulator (and hence, can be extended to a number of general Euler Lagrange systems), they do not require acceleration measurements, and they are independent from the controller. A Lyapunov-based analysis is provided to prove that the developed fault observer converges to the actual fault.

In this paper, a homography-based adaptive visual servo controller is described to enable a robot end-effector to track a desired Euclidean space trajectory as determined by a sequence of images for the fixed camera configuration. To achieve the objective, a Lyapunov-based adaptive control strategy is employed to actively compensate for unknown depth measurements and the lack of an object model. The error systems are constructed as a hybrid of pixel information and reconstructed Euclidean variables obtained by comparing the images and decomposing a homographic relationship.

A visual servo tracking controller is developed in this paper for a monocular camera system mounted on an underactuated wheeled robot (WMR) subject nonholonomic motion constraints (i.e., the camera-in-hand problem). A prerecorded image sequence (e.g., a video) of three target points is used to define a desired trajectory for the WMR. By comparing the target points from the prerecorded sequence with the corresponding target points in the live image, projective geometric relationships are exploited to construct a Euclidean homography. The information obtained by decomposing the Euclidean homography is used to develop a kinematic controller. A Lyapunov-based analysis is used to develop an adaptive update law to actively compensate for the lack of depth information required for the translation error system.

Typical visual servoing control objectives are formulated by the desire to position/orient a camera based on a reference image obtained by a priori positioning the camera in the desired location. In this paper, projective geometric techniques are used to formulate a visual servo control problem based on a cooperative camera scheme that provides an alternative to the typical "teach by showing" methodology. Specifically, by using a second camera with zoom capabilities, the proposed "teach by zooming" alternative approach eliminates the need for the camera to be a priori positioned in the desired location. To achieve this result, development is provided that illustrates how cameras with different intrinsic parameters can be used (i.e., camera independent). Control objectives are developed for cases when the intrinsic camera calibration matrices for the camera-in-hand and fixed camera are both known or both unknown.

In this paper, we develop an adaptive torque control input for wheeled mobile robots (WMRs) that can be utilized in a modular manner with parameter estimate update laws to enable unified tracking and regulation, that is, a prediction error based update law ensures the parameter estimate vector is bounded, all of the signals are proven to be bounded. An additional stability analysis is provided to prove that if the adaptive update law is designed such that the prediction error is square-integrable and the estimated inertia matrix is positive-definite, the WMR tracking and regulation errors are globally asymptotically forced to a control term that can be made arbitrarily small.

In this paper, we utilize a Lyapunov-based design approach. to construct a visual servoing controller for a robot manipulator that ensures uniformly ultimately bounded (UUB) end-effector position tracking performance despite parametric uncertainty throughout the entire robot/camera system. The UUB end-effector tracking result exploits information from both a fixed camera and a camera-in-hand. although both cameras contain parametric uncertainty in the calibration parameters (e.g., focal length, image center, scaling factors, and camera position and orientation). The advantages of the cooperative camera configuration are that: (i) the fixed camera can be mounted so that a large robot workspace is visible, (ii) the camera-in hand is mounted so that a high resolution, close-up view of an object is achieved, facilitating the potential for more precise robotic motion. and (iii) the fixed camera provides a mechanism for treating the problem of determining the relative velocity of the robot end-effector with respect to the object for the camera-in-hand object tracking problem when the camera is uncalibrated.

In this paper, the development of a novel piezoelectric-based locomotion design for a meso-scale mobile robot is described. The design exploits a compliant mechanical structure that enables piezoelectric forces to be amplified and transmitted to "legs" that propel the robot through a lift and pull scheme. The lift and pull design contrasts with typical slip/stick approaches that assume the inertia effects of the robot will allow the robot to slip on a smooth surface. By eliminating the slipping assumption, the potential surfaces that the proposed robot can traverse is significantly extended. Discussions are provided regarding the proposed locomotion method, kinematic modeling and system constraints, mechanical optimization, and control of the meso-scale robot.

In this paper, a learning-based feedforward term is developed to solve a general control problem in the presence of unknown nonlinear dynamics with a known period. Since the learning-based feedforward term is generated from a straightforward Lyapunov-like stability analysis, the control designer can utilize other Lyapunov-based design techniques to develop hybrid control schemes that utilize learning-based feedforward terms to compensate for periodic dynamics and other Lyapunov-based approaches (e.g., adaptive-based feedforward terms) to compensate for nonperiodic dynamics. To illustrate this point, a hybrid adaptive/learning control scheme is utilized to achieve global asymptotic link position tracking for a robot manipulator.

A Simulink robotic toolkit (SRTK) for the Puma 560 robot manipulator is developed on the MATLAB/Simulink-based platform. Through the use of the real-time Linux target and the real-time Windows target, the SRTK can be executed on the Linux or Win32-based operating systems in real-time. Moreover, the graphical user-friendly nature of Simulink allows the SRTK to be a flexible tool that can easily be customized to fit the specific needs of the user, that is, based on the layered approach of the SRTK, the user can perform operations such as calibration, joint control, Cartesian control, Cartesian PD control, impedance control, some trajectory generation tasks, and real-time simulation of the Puma 560 through a user-friendly MATLAB-based graphical user interface without writing any code. The SRTK allows a researcher to use the Puma 560 without the burden of the external issues related to the control, interface, and software issues, while providing for the flexibility for easily modifying components for increased functionality.

This paper considers the problem of position/orientation tracking control of wheeled mobile robots via visual servoing in the presence of parametric uncertainty associated with the mechanical dynamics and the camera system. Specifically, we design an adaptive controller that compensates for uncertain camera and mechanical parameters and ensures global asymptotic position/orientation tracking. Simulation and experimental results are included to illustrate the performance of the control law.

We introduce a new approach to fault detection for robot manipulators. The technique, which is based on the isolation of fault signatures via filtered torque prediction error estimates, does not require measurements or estimates of manipulator acceleration as is the case with some of the previously suggested methods. The method is formally demonstrated to be robust under uncertainty in the robot parameters. Furthermore, an adaptive version of the algorithm is introduced, and shown to both improve coverage and significantly reduce detection times. The effectiveness of the approach is demonstrated by experiments with a two-joint manipulator system.

In this paper, we introduce a new approach to fault detection for robot manipulators. The technique, which is based on the isolation of fault signatures via filtered torque prediction error estimates, does not require measurements or estimates of manipulator acceleration as is the case with some previously suggested methods. The method is formally demonstrated to be robust under uncertainty in the robot parameters. Furthermore, an adaptive version of the algorithm is introduced, and shown to both improve coverage and significantly reduce detection times. The effectiveness of the approach is demonstrated by experiments with a two-joint manipulator system.

The original paper of Su and Stepanenko (ibid., vol.14, p.651-5, 1998) presents the design of an adaptive/robust controller for uncertain electrically-driven robots with no velocity measurements. This note shows that the claim that velocity measurements are not required for control implementation is incorrect.

This paper presents the design of a differentiable, kinematic control law that achieves global asymptotic tracking. In addition, we also illustrate how the proposed kinematic controller provides global exponential tracking provided the reference trajectory satisfies a mild persistency of excitation (PE) condition. We also illustrate how the proposed kinematic controller can be slightly modified to provide for global asymptotic regulation of both the position and orientation of the mobile robot. Finally, we embed the differentiable kinematic controller inside of an adaptive controller that fosters global asymptotic tracking despite parametric uncertainty associated with the dynamic model. Experimental results are also provided to illustrate the performance of the proposed adaptive tracking controller.

We present a global, output feedback, tracking controller for rigid-link, flexible-joint robots. Specifically, we design an exact model knowledge, Lyapunov-based controller which provides global asymptotic link position tracking despite the fact that only link position measurements are available.