Directional control of planar human arm movement.

We have proposed two different digital control methods, where the controlled variables are joint torque and joint angular velocity, of space robot manipulators using the transpose of Generalized Jacobian Matrix. These methods are developed in discrete time domain so that the stability of system can be guaranteed when digital computers are used as controllers. The relation between the two methods, however, is not reported, although the basic concepts of these methods ar similar. This paper presents a novel digital control method, which integrates joint torque control with joint velocity control, of nonredundant space robot manipulators using the transpose of Generalized Jacobian Matrix. The merit of this new control law is to be applicable to both joint's torque controller and joint's velocity controller by only tuning the controller gains. Furthermore, computer simulation of a 6-DOF space robot manipulator is performed. Simulation result demonstrates the effectiveness of the proposed method.

Design of a digital adaptive algorithm for torque-controlled robotic joints

Improving feasibility of robotic milling through robot placement optimisation

This is a proposal for a new technique for evaluating spasticity in the upper limbs of hemiplegic patients. Each subject sat on a chair or stood up, and his or her forearm was extended or flexed by a physician. The subject was instructed to relax. The elbow joint angle, torque, and electromyograms (EMGs) of the biceps brachii, triceps brachii, and brachioradialis muscles were measured. The relationship between the elbow joint angle and torque was approximated with a mathematical model, which consisted of elastic components depending on both muscle activities and elbow joint angle, by the least squares method. The inertia and visco-elastic coefficients were obtained. The elbow angle response was then estimated with the obtained inertia and visco-elastic coefficients by the Runge-Kutta method, and the estimated elbow angle was compared to the observed one. The relationships between the elbow angle and torque were approximated well with the model. Next, the average elasticity was calculated and compared to the modified Ashworth scale. The average elasticity had a tendency to increase as the Ash- worth scale increased. In addition, the average elasticity varied depending on the posture of the subjects.

This paper examines the effects of errors in applied force on robot force-controlled task performance. Two configuration-dependent effects of force errors are identified. Kinematic stability errors occur when the robot configuration is such that small inaccuracies in joint torque create relatively large robot link displacements. This sensitivity to torque errors can seriously inhibit task performance in certain configurations. Task performance can also be affected by errors in the endpoint force angle which can be significant for even small changes in joint torque. Performance maps are developed as tools to identify robot configurations sensitive to these errors. The effects of force errors are demonstrated in the performance of a planar peg and hole task by a three link manipulator under joint torque control. >

The application of robot manipulators to batch manufacturing product assembly requires the control and monitoring of both robot force and position. A good force sensing technique is imperative to increase the robot's capabilities in this area. A fast reliable force servo system based on the proper sensing technique is also needed to increase the speed of robot operations. This paper analyzes currently existing force sensing techniques concluding that a joint torque sensing technique will results the best performance in a force servo system. A new, simple, high again, wide bandwidth torque servo system using a joint torque sensor system has been developed and has been verified by using a single joint manipulator. An accurate method of selecting the force servo joints to provide compliance has also been developed. Finally, a complete and consistent algorithm for performing compliance is presented.

The objective of this study is to develop a technique for the quantitative evaluation of spasticity in the upper limbs. The subjects were two healthy volunteers and 19 hemiplegic patients. Each subject was seated or standing in a resting state, and the forearm was flexed five times and extended five times in each posture through a force-and-torque sensor placed on the wrist. The forearms of subjects who could not stand up were flexed and extended only in a seated position. The elbow joint angle, torque, and electromyograms (EMGs) of the biceps brachii muscle, triceps brachii muscle, and brachioradialis muscle during flexion and extension were measured. The inertia of forearms and visco-elastic coefficients were estimated by the least squares method with a mathematical model that consisted of elastic components depending on the elbow joint angle and muscle activities. The elbow joint angle and torque were then estimated with the obtained inertia and visco-elastic coefficients and compared to the observed angle and torque in order to evaluate the performance of the mathematical model. The elbow joint angle and torque were closely estimated by the mathematical model. Next, the maximum value of the elastic coefficient during flexion/extension around the elbow joint (k max) was calculated and compared with the Ashworth scale. The results show that the k max had a tendency to increase as the Ashworth scale increased. In conclusion, the mathematical model could represent the relationship between the elbow joint angle and torque. The k max would be medically and clinically useful for the quantitative evaluation of spasticity in hemiplegic patients.

1. We studied seven subjects with cerebellar lesions and seven control subjects as they made reaching movements in the sagittal plane to a target directly in front of them. Reaches were made under three different conditions: 1) "slow-accurate," 2) "fast-accurate," and 3) "fast as possible." All subjects were videotaped moving in a sagittal plane with markers on the index finger, wrist, elbow, and shoulder. Marker positions were digitized and then used to calculate joint angles. For each of the shoulder, elbow and wrist joints, inverse dynamics equations based on a three-segment limb model were used to estimate the net torque (sum of components) and each of the component torques. The component torques consisted of the torque due to gravity, the dynamic interaction torques induced passively by the movement of the adjacent joint, and the torque produced by the muscles and passive tissue elements (sometimes called "residual" torque). 2. A kinematic analysis of the movement trajectory and the change in joint angles showed that the reaches of subjects with cerebellar lesions were abnormal compared with reaches of control subjects. In both the slow-accurate and fast-accurate conditions the cerebellar subjects made abnormally curved wrist paths; the curvature was greater in the slow-accurate condition. During the slow-accurate condition, cerebellar subjects showed target undershoot and tended to move one joint at a time (decomposition). During the fast-accurate reaches, the cerebellar subjects showed target overshoot. Additionally, in the fast-accurate condition, cerebellar subjects moved the joints at abnormal rates relative to one another, but the movements were less decomposed. Only three subjects were tested in the fast as possible condition; this condition was analyzed only to determine maximal reaching speeds of subjects with cerebellar lesions. Cerebellar subjects moved more slowly than controls in all three conditions. 3. A kinetic analysis of torques generated at each joint during the slow-accurate reaches and the fast-accurate reaches revealed that subjects with cerebellar lesions produced very different torque profiles compared with control subjects. In the slow-accurate condition, the cerebellar subjects produced abnormal elbow muscle torques that prevented the normal elbow extension early in the reach. In the fast-accurate condition, the cerebellar subjects produced inappropriate levels of shoulder muscle torque and also produced elbow muscle torques that did not very appropriately with the dynamic interaction torques that occurred at the elbow. Lack of appropriate muscle torque resulted in excessive contributions of the dynamic interaction torque during the fast-accurate reaches. 4. The inability to produce muscle torques that predict, accommodate, and compensate for the dynamic interaction torques appears to be an important cause of the classic kinematic deficits shown by cerebellar subjects during attempted reaching. These kinematic deficits include incoordination of the shoulder and the elbow joints, a curved trajectory, and overshoot. In the fast-accurate condition, cerebellar subjects often made inappropriate muscle torques relative to the dynamic interaction torques. Because of this, interaction torques often determined the pattern of incoordination of the elbow and shoulder that produced the curved trajectory and target overshoot. In the slow-accurate condition, we reason that the cerebellar subjects may use a decomposition strategy so as to simplify the movement and not have to control both joints simultaneously. From these results, we suggest that a major role of the cerebellum is in generating muscle torques at a joint that will predict the interaction torques being generated by other moving joints and compensate for them as they occur.

This paper presents an efficient method for adaptive control of humanoid robot arms with biarticular muscles, which exhibits multiple beneficial properties. In our approach, sliding control was chosen to get joint torque first and then the joint torque was distributed to muscle forces. The muscle force was computed based on a Jacobian matrix between joint torque space and muscle force space. In addition, internal forces were used to optimize the computed muscle forces - and thus, two important benefits arise: through our proposed method, not only we are making sure that each muscle force stays within its predefined force boundary; but we are also enabling the muscles to work in the middle of their working range, which is considered to be an anti-fatigue state. Yet more benefit derived, comes from the fact that in our method all the dynamic parameters are updated in real-time, and can thus one can account for perturbations and disturbances during operation. Compared with previous work in parameter adaptation, a composite method was proposed which utilizes the prediction error to accelerate parameter convergence speed. Our method was tested for the case of reaching motions. The results clearly illustrate the benefits of the method.

The objective of this study was to understand how people adapt to errors when using a myoelectric control interface. We compared adaptation across 1) non-amputee subjects using joint angle, joint torque, and myoelectric control interfaces, and 2) amputee subjects using myoelectric control interfaces with residual and intact limbs (five total control interface conditions). We measured trial-by-trial adaptation to self-generated errors and random perturbations during a virtual, single degree-of-freedom task with two levels of feedback uncertainty, and evaluated adaptation by fitting a hierarchical Kalman filter model. We have two main results. First, adaptation to random perturbations was similar across all control interfaces, whereas adaptation to self-generated errors differed. These patterns matched predictions of our model, which was fit to each control interface by changing the process noise parameter that represented system variability. Second, in amputee subjects, we found similar adaptation rates and error levels between residual and intact limbs. These results link prosthesis control to broader areas of motor learning and adaptation and provide a useful model of adaptation with myoelectric control. The model of adaptation will help us understand and solve prosthesis control challenges, such as providing additional sensory feedback.

The authors have proposed a torque sensor which utilizes elasticity of harmonic drives. The characteristics of joint torque control were experimentally examined using this torque sensor. Three types of torque control laws were implemented with a one-link arm to find a control method which provides excellent friction reduction and dynamic response in joint torque. The experimental results of the joint torque control showed that the torque sensor was very useful for compensating for the nonlinear friction. The torque control was effective in improving the accuracy of motion control at low velocity and suppressed the vibration caused by the joint flexibility.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Towards stabilization and navigational analysis of humanoids in complex arena using a hybridized fuzzy embedded PID controller approach

BackgroundFunctional electrical stimulation (FES) is a neuroprosthetic technique for restoring lost motor function of spinal cord injured (SCI) patients and motor-impaired subjects by delivering short electrical pulses to their paralyzed muscles or motor nerves. FES induces action potentials respectively on muscles or nerves so that muscle activity can be characterized by the synchronous recruitment of motor units with its compound electromyography (EMG) signal is called M-wave. The recorded evoked EMG (eEMG) can be employed to predict the resultant joint torque, and modeling of FES-induced joint torque based on eEMG is an essential step to provide necessary prediction of the expected muscle response before achieving accurate joint torque control by FES.MethodsPrevious works on FES-induced torque tracking issues were mainly based on offline analysis. However, toward personalized clinical rehabilitation applications, real-time FES systems are essentially required considering the subject-specific muscle responses against electrical stimulation. This paper proposes a wireless portable stimulator used for estimating/predicting joint torque based on real time processing of eEMG. Kalman filter and recurrent neural network (RNN) are embedded into the real-time FES system for identification and estimation.ResultsPrediction results on 3 able-bodied subjects and 3 SCI patients demonstrate promising performances. As estimators, both Kalman filter and RNN approaches show clinically feasible results on estimation/prediction of joint torque with eEMG signals only, moreover RNN requires less computational requirement.ConclusionThe proposed real-time FES system establishes a platform for estimating and assessing the mechanical output, the electromyographic recordings and associated models. It will contribute to open a new modality for personalized portable neuroprosthetic control toward consolidated personal healthcare for motor-impaired patients.

1. Eight subjects performed three series of pointing tasks with the unconstrained arm. Series one and two required subjects to move between two fixed targets as quickly as possible with different weights attached to the wrist. By specifying initial and final positions of the finger tip, the first series was performed by flexion of both shoulder and elbow and the second by shoulder flexion and elbow extension. The third series required flexion at both joints, and subjects were instructed to vary movement speed. We examined how variations in load or intended speed were associated with changes in the amount and timing of the electromyographic (EMG) activity and the net muscle torque production. 2. EMG and torque patterns at the individual joints varied with load and speed according to most of the same rules we have described for single-joint movements. 1) Movements were produced by biphasic torque pulses and biphasic or triphasic EMG bursts at both joints. 2) The accelerating impulse was proportional to the load when the subject moved "as fast and accurately as possible" or to speed if that was intentionally varied. 3) The area of the EMG bursts of agonist muscles varied with the impulse. 4) The rates of rise of the net muscle torques and of the EMG bursts were proportional to intended speed and insensitive to inertial load. 5) The areas of the antagonist muscle EMG bursts were proportional to intended movement speed but showed less dependence on load, which is unlike what is observed during single-joint movements. 3. Comparisons across joints showed that the impulse produced at the shoulder was proportional to that produced at the elbow as both varied together with load and speed. The torques at the two joints varied in close synchrony, achieving maxima and going through zero almost simultaneously. 4. We hypothesize that "coordination" of the elbow and shoulder is by the planning and generation of synchronized, biphasic muscle torque pulses that remain in near linear proportionality to each other throughout most of the movement. This linear synergy produces movements with the commonly observed kinematic properties and that are preserved over changes in speed and load.

The signal measured with an electromyogram (EMG) is the summation of all action potentials of motor units active at a certain time. According to previous literature, one can establish the relationship between torque and EMG/activations in a forward way, i.e., employing EMG of multiple channels to estimate the joint torque. Once the relationship is established, the torque can be predicted with EMG recordings. However, in some applications of neuroprosthetics where we need to make muscle control, it is required to inversely have an insight regarding the muscle activations under a specific motion scenario from the corresponding torque. Motivated by this point, this paper investigates inverse estimation of muscle activations in random contractions at the ankle joint. Local multiple regression is exploited for finding the relationship between muscle activations and torque. Such technique is able to rebuild the relationship between muscle activations and joint torque inversely based on experimental data obtained from five able-bodied subjects, and the resultant optimal weight matrix can indicate each muscle's contribution in the production of the torque. Further cross validation on prediction of muscle activations with joint torque with optimal weights shows that such approach may possess promising performance.