Inverse Dynamics Research Articles

Integration of sEMG-Based Learning and Adaptive Fuzzy Sliding Mode Control for an Exoskeleton Assist-as-Needed Support System

This paper presents an adaptive Fuzzy Sliding Mode Control approach for an Assist-as-Needed (AAN) strategy to achieve effective human–exoskeleton synergy. The proposed strategy employs an adaptive instance-based learning algorithm to estimate muscle effort, based on surface Electromyography (sEMG) signals. To determine and control the inverse dynamics of a highly nonlinear 4-degrees-of-freedom exoskeleton designed for upper-limb therapeutic exercises, a modified Recursive Newton-Euler Algorithm (RNEA) with Sliding Mode Control (SMC) was used. The exoskeleton position error and raw sEMG signal from the bicep’s brachii muscle were used as inputs for a fuzzy inference system to produce an output to adjust the sliding mode control law parameters. The proposed robust control law was simulated using MATLAB-Simulink, and the results showed that it could instantly adjust the necessary support, based on the combined motion of the human–exoskeleton system’s muscle engagement, while keeping the state trajectory errors and input torque bounded within ±5×10−2 rads and ±5 N.m, respectively.

The effect of pathological shoulder rhythm on muscle and joint forces after reverse shoulder arthroplasty, a numerical analysis

BackgroundCompromised abduction ability after reverse shoulder arthroplasty is primarily linked to limited glenohumeral range of motion while scapulothoracic mobility can typically be maintained. Glenohumeral joint forces strongly depend on the resulting scapulohumeral rhythm, however, an association between the acting muscle and joint forces and the subject-specific scapulohumeral rhythm after reverse shoulder arthroplasty has not been established. MethodsEleven reverse shoulder arthroplasty patients were divided into groups of poor and excellent abduction ability. Subject-specific models were developed and scaled for each patient using existing motion capture data in AnyBody™. Shoulder muscle and joint forces were obtained using inverse dynamics calculations during shoulder abduction to 100° in the scapula plane. The scapulohumeral rhythm, the resting abduction angle and internal body forces between the outcome groups were compared using a Mann Whitney U test. FindingsThe mean glenohumeral and scapulothoracic contribution to overall shoulder abduction for the excellent group was on average 9.7% higher and 21.4% lower, respectively, compared to the mean of the poor group. For shoulder abduction angles between 30° and 60°, the excellent group demonstrated on average 25% higher muscle forces in the anterior deltoid which was significantly higher compared to the poor outcome patients. Scapulothoracic muscle activity did not differ significantly between the two functional groups. InterpretationAccordingly, rehabilitation strategies focusing on strengthening the anterior part of the deltoid in particular may improve clinical outcomes.

Predicting ankle and knee sagittal kinematics and kinetics using an ankle-mounted inertial sensor

The purpose of this study was to develop a machine learning model to reconstruct time series kinematic and kinetic profiles of the ankle and knee joint across six different tasks using an ankle-mounted IMU. Four male collegiate basketball players performed repeated tasks, including walking, jogging, running, sidestep cutting, max-height jumping, and stop-jumping, resulting in a total of 102 movements. Ankle and knee flexion-extension angles and moments were estimated using motion capture and inverse dynamics and considered ‘actual data’ for the purpose of model fitting. Synchronous acceleration and angular velocity data were collected from right ankle-mounted IMUs. A time-series feature extraction model was used to determine a set of features used as input to a random forest regression model to predict the ankle and knee kinematics and kinetics. Five-fold cross-validation was performed to verify the model accuracy, and statistical parametric mapping was used to determine the difference between the predicted and experimental time series. The random forest regression model predicted the time-series profiles of the ankle and knee flexion-extension angles and moments with high accuracy (Kinematics: R2 ranged from 0.782 to 0.962, RMSE ranged from 2.19° to 11.58°; Kinetics: R2 ranged from 0.711 to 0.966, RMSE ranged from 0.10 Nm/kg to 0.41 Nm/kg). There were differences between predicted and actual time series for the knee flexion-extension moment during stop-jumping and walking. An appropriately trained feature-based regression model can predict time series knee and ankle joint angles and moments across a wide range of tasks using a single ankle-mounted IMU.

A Recursive Lie-Group Formulation for the Second-Order Time Derivatives of the Inverse Dynamics of Parallel Kinematic Manipulators

Series elastic actuators (SEA) were introduced for serial robotic arms. Their model-based trajectory tracking control requires the second time derivatives of the inverse dynamics solution, for which algorithms were proposed. Trajectory control of parallel kinematics manipulators (PKM) equipped with SEAs has not yet been pursued. Key element for this is the computationally efficient evaluation of the second time derivative of the inverse dynamics solution. This has not been presented in the literature, and is addressed in the present paper for the first time. The special topology of PKM is exploited reusing the recursive algorithms for evaluating the inverse dynamics of serial robots. A Lie group formulation is used and all relations are derived within this framework. Numerical results are presented for a 6- DOF Gough-Stewart platform (as part of an exoskeleton), and for a planar PKM when a flatness-based control scheme is applied.

Optimal Maneuvers Around Binary Asteroids Using Particle Swarm Optimization and Machine Learning

Designing optimal transfer trajectories and reference orbit tracking in binary asteroid systems is both challenging and computationally expensive. This paper proposes a method of bypassing the high computational overhead by leveraging a collection of known techniques. Indeed, the proposed framework is based on the combination of artificial intelligence techniques, such as the particle swarm optimization and neural networks, along with the inverse dynamics and the B-splines approximation of the trajectory. The real irregular shapes of the asteroids are considered in the free dynamics of the system, which are obtained via the mutual polyhedral model. The gravitational accelerations of the single asteroids acting on the spacecraft are approximated by using two single-layer neural networks trained via an extreme learning machine. By using a combination of these techniques, the computational time of the whole optimization is decreased from hours to minutes. The proposed approach is applied to the optimal trajectory design around the binary asteroid system, 1999 KW4, showing the feasibility of the proposed optimization approach, reducing the computational effort and time, and increasing the reliability of the obtained results. It is shown through a Monte Carlo analysis that our optimization strategy yields more accurate solutions than other optimization algorithms, such as the interior point and sequential quadratic programming methods, when a random initial guess is provided. Finally, the proposed optimization approach can be used in combination with other techniques to provide a feasible and reliable initial guess for a better solution refinement.t

Effect of neglecting passive spinal structures: a quantitative investigation using the forward-dynamics and inverse-dynamics musculoskeletal approach.

Purpose: Inverse-dynamics (ID) analysis is an approach widely used for studying spine biomechanics and the estimation of muscle forces. Despite the increasing structural complexity of spine models, ID analysis results substantially rely on accurate kinematic data that most of the current technologies are not capable to provide. For this reason, the model complexity is drastically reduced by assuming three degrees of freedom spherical joints and generic kinematic coupling constraints. Moreover, the majority of current ID spine models neglect the contribution of passive structures. The aim of this ID analysis study was to determine the impact of modelled passive structures (i.e., ligaments and intervertebral discs) on remaining joint forces and torques that muscles must balance in the functional spinal unit. Methods: For this purpose, an existing generic spine model developed for the use in the demoa software environment was transferred into the musculoskeletal modelling platform OpenSim. The thoracolumbar spine model previously used in forward-dynamics (FD) simulations provided a full kinematic description of a flexion-extension movement. By using the obtained in silico kinematics, ID analysis was performed. The individual contribution of passive elements to the generalised net joint forces and torques was evaluated in a step-wise approach increasing the model complexity by adding individual biological structures of the spine. Results: The implementation of intervertebral discs and ligaments has significantly reduced compressive loading and anterior torque that is attributed to the acting net muscle forces by -200% and -75%, respectively. The ID model kinematics and kinetics were cross-validated against the FD simulation results. Conclusion: This study clearly shows the importance of incorporating passive spinal structures on the accurate computation of remaining joint loads. Furthermore, for the first time, a generic spine model was used and cross-validated in two different musculoskeletal modelling platforms, i.e., demoa and OpenSim, respectively. In future, a comparison of neuromuscular control strategies for spinal movement can be investigated using both approaches.

Machine learning-based prediction of joint moments based on kinematics in patients with cerebral palsy

Joint moments during gait provide valuable information for clinical decision-making in patients with cerebral palsy (CP). Joint moments are calculated based on ground reaction forces (GRF) using inverse dynamics models. Obtaining GRF from patients with CP is challenging. Typically developed (TD) individuals' joint moments were predicted from joint angles using machine learning, but no such study has been conducted on patients with CP. Accordingly, we aimed to predict the dorsi-plantar flexion, knee flexion–extension, hip flexion–extension, and hip adduction-abduction moments based on the trunk, pelvis, hip, knee, and ankle kinematics during gait in patients with CP and TD individuals using one-dimensional convolutional neural networks (CNN). The anonymized retrospective gait data of 329 TD (26 years ± 14, mass: 70 kg ± 15, height: 167 cm ± 89) and 917 CP (17 years ± 9, mass:47 kg ± 19, height:153 cm ± 36) individuals were evaluated and after applying inclusion–exclusion criteria, 132 TD and 622 CP patients with spastic diplegia were selected. We trained specific CNN models and evaluated their performance using isolated test subject groups based on normalized root mean square error (nRMSE) and Pearson correlation coefficient (PCC). Joint moments were predicted with nRMSE between 18.02% and 13.58% for the CP and between 12.55% and 8.58% for the TD groups, whereas with PCC between 0.85 and 0.93 for the CP and between 0.94 and 0.98 for the TD groups. Machine learning-based joint moment prediction from kinematics could replace conventional moment calculation in CP patients in the future, but the current level of prediction errors restricts its use for clinical decision-making today.

Unilateral transfemoral osseointegrated prostheses improve joint loading during walking

People with unilateral transfemoral amputation using socket prostheses are at increased risk for developing osteoarthritis in both the residual hip and intact lower-limb joints. Osseointegrated prostheses are a surgical alternative to socket prostheses that directly attach to the residual femur via a bone-anchored implant, however their multi-joint loading effect is largely unknown. Our objective was to establish how osseointegrated prostheses influence joint loading during walking. Motion capture data (kinematics, ground reaction forces) were collected from 12 participants at baseline, with socket prostheses, and 12-months after prosthesis osseointegration during overground walking at self-selected speeds. Subject-specific musculoskeletal models were developed at each timepoint relative to osseointegration. Internal joint moments were calculated using inverse dynamics, muscle and joint reaction forces (JRFs) were estimated with static optimization. Changes in internal joint moments, JRFs, and joint loading-symmetry were compared using statistical parametric mapping (p≤ 0.05) before and after osseointegration. Amputated limb hip flexion moments and anterior JRFs decreased during terminal stance (p = 0.002, <0.001; respectively), while amputated limb hip abduction moments increased during mid-stance (p < 0.001), amputated hip rotation moment changed from internal to external throughout early stance (p < 0.001). Intact limb hip extension and knee flexion moments (p = 0.028, 0.032; respectively), superior and resultant knee JRFs (p = 0.046, 0.049; respectively) decreased during the loading response following prosthesis osseointegration. These results may indicate that the direct loading transmission of these novel prostheses create a more typical mechanical environment in bilateral joints, which is comparable with loading observed in able-bodied individuals and could decrease the risk of development or progression of osteoarthritis.

Subject specific optimal control of powered knee exoskeleton to assist human lifting tasks under controlled environment

AbstractWearable robots, sometimes known as exoskeletons, are incredible devices for improving human strength, reducing fatigue, and restoring impaired mobility. The control of powered exoskeletons, on the other hand, is still a challenge. This necessitates the development of a technique to simulate exoskeleton–wearer interaction. This study uses a two-dimensional human skeletal model with a powered knee exoskeleton to predict the optimal lifting motion and assistive torque. For lifting motion prediction, an inverse dynamics optimization formulation is utilized. In addition, the electromechanical dynamics of the exoskeleton DC motor are modeled in the lifting optimization formulation. The design variables are human joint angle profiles and exoskeleton motor current profiles. The human joint torque square is minimized subject to physical and lifting task constraints. Then, the lifting optimization problem is solved by the gradient-based sparse nonlinear optimizer (SNOPT). Furthermore, the optimal exoskeleton torque is implemented through a two-phase control strategy to provide optimal assistance in lifting. Experimental validations of the optimal control with 6 Nm and 16 Nm maximum assistive torque are presented. Both 6 Nm and 16 Nm maximum optimal assistance of the exoskeletons reduce the mean values of vastus lateralis, biceps femoris, and latissimus dorsi muscle activations compared to the lifting without the exoskeleton. However, the mean value of the vastus medialis activation is increased by a small amount for the exoskeleton case, although its peak value is reduced. Finally, the experimental results demonstrate that the proposed lifting optimization formulation and control strategy are promising for powered knee exoskeleton for lifting tasks.

Robot learning from human demonstrations with inconsistent contexts

Visual imitation learning is a promising approach that promotes robots to learn skills from visual demonstrations. However, current visual imitation learning approaches introduce unreasonable assumptions that the contexts of the visual demonstrations and the robot observations are consistent, which affects the flexibility and scalability of the approaches. It is a key challenge for robots to learn from visual demonstrations with inconsistent contexts. Inconsistent contexts may cause a serious difference in the pixel distribution of the operator and the environment, which makes vision-based control policies hardly effective. In this paper, we propose a novel imitation learning framework to enable robots to reproduce behavior by watching human demonstrations with inconsistent contexts, such as different viewpoints, operators, backgrounds, object appearances and positions. Specifically, our framework consists of three networks: flow-based viewpoint transformation network (FVTrans), robot2human alignment network (RANet) and inverse dynamics network (IDNet). First, FVTrans transforms various third-person demonstrations into the fixed robot execution view. With a meta learning strategy, FVTrans can quickly adapt to novel contexts with few samples. Then, RANet aligns the human and the robot at the feature level. Therefore, the demonstration feature can be used as a subgoal of the current moment. Finally, IDNet predicts the joint angles of the robot. We collect a multi-context dataset on the real robot (UR5) for three tasks, including grasping cups, sweeping garbage and placing objects. We empirically demonstrate that our framework can perform three tasks with a high success rate and be effectively generalized to different contexts.

Research on Adaptive Cruise Systems Based on Adjacent Vehicle Trajectory Prediction

Vehicles in the adjacent lane making abrupt lane changes is a common and frequent action during traffic movement. Being aware of adjacent vehicles ahead of time, determining their cut-in intention, monitoring their cut-in trajectory in real time, and actively adjusting following speed are all critical for adaptive cruise systems for vehicles. This study proposes a flexible following-factor-calculation approach that considers the driver’s willingness to take risks for the purpose of identifying cut-in intent, predicting trajectory, and narrowing the window for following cruise speed adjustment to improve passenger ride comfort. To begin, a lane-change trajectory prediction algorithm based on driver adventitious factor correction is proposed in order to correctly predict the lane-change trajectory of adjacent vehicles in urban road traffic scenarios. Second, the flexible following factor and the flexible switching factor of the following target are constructed to overcome the influence of the uncertainty caused by internal and external disturbances on the vehicle following the motion process, and to reduce the impact of cut-in events on passenger comfort. An anti-disturbance rejection control and an adaptive cruise controller based on the vehicle’s longitudinal inverse dynamics model are proposed in order to compensate for and suppress the internal perturbations caused by the vehicle’s internal parameter changes and the random disturbances caused by external road environment changes. The results of simulation and real-world testing showed an average of 28% improvement in passenger comfort.

Subject-specific tribo-contact conditions in total knee replacements: a simulation framework across scales.

Fundamental knowledge about invivo kinematics and contact conditions at the articulating interfaces of total knee replacements are essential for predicting and optimizing their behavior and durability. However, the prevailing motions and contact stresses in total knee replacements cannot be precisely determined using conventional invivo measurement methods. Insilico modeling, in turn, allows for a prediction of the loads, velocities, deformations, stress, and lubrication conditions across the scales during gait. Within the scope of this paper, we therefore combine musculoskeletal modeling with tribo-contact modeling. In the first step, we compute contact forces and sliding velocities by means of inverse dynamics approach and force-dependent kinematic solver based upon experimental gait data, revealing contact forces during healthy/physiological gait of young subjects. In a second step, the derived data are employed as input data for an elastohydrodynamic model based upon the finite element method full-system approach taking into account elastic deformation, the synovial fluid's hydrodynamics as well as mixed lubrication to predict and discuss the subject-specific pressure and lubrication conditions.

High-order inverse dynamics of serial robots based on projective geometric algebra

High-order inverse dynamics of serial robots based on projective geometric algebra

Estimating 3D ground reaction forces in running using three inertial measurement units.

To understand the mechanisms causing running injuries, it is crucial to get insights into biomechanical loading in the runners' environment. Ground reaction forces (GRFs) describe the external forces on the body during running, however, measuring these forces is usually only possible in a gait laboratory. Previous studies show that it is possible to use inertial measurement units (IMUs) to estimate vertical forces, however, forces in anterior-posterior direction play an important role in the push-off. Furthermore, to perform an inverse dynamics approach, for modelling tissue specific loads, 3D GRFs are needed as input. Therefore, the goal of this work was to estimate 3D GRFs using three inertial measurement units. Twelve rear foot strike runners did nine trials at three different velocities (10, 12 and 14 km/h) and three stride frequencies (preferred and preferred ± 10%) on an instrumented treadmill. Then, data from IMUs placed on the pelvis and lower legs were used as input for artificial neural networks (ANNs) to estimate 3D GRFs. Additionally, estimated vertical GRF from a physical model was used as input to create a hybrid machine learning model. Using different splits in validation and training data, different ANNs were fitted and assembled into an ensemble model. Leave-one-subject-out cross-validation was used to validate the models. Performance of the machine learning, hybrid machine learning and a physical model were compared. The estimated vs. measured GRF for the hybrid model had a RMSE normalized over the full range of values of 10.8, 7.8 and 6.8% and a Pearson correlation coefficient of 0.58, 0.91, 0.97 for the mediolateral direction, posterior-anterior and vertical direction respectively. Performance for the three compared models was similar. The ensemble models showed higher model accuracy compared to the ensemble-members. This study is the first to estimate 3D GRF during continuous running from IMUs and shows that it is possible to estimate GRF in posterior-anterior and vertical direction, making it possible to estimate these forces in the outdoor setting. This step towards quantification of biomechanical load in the runners' environment is helpful to gain a better understanding of the development of running injuries.

Computationally Efficient Inverse Dynamics of a Spatial Parallel Mechanism Constrained Directly by the Base at Two Point-Contact Higher Kinematic Pairs

Computationally Efficient Inverse Dynamics of a Spatial Parallel Mechanism Constrained Directly by the Base at Two Point-Contact Higher Kinematic Pairs

Proposing a Graphic Simulator for an Upper Limb Exoskeleton Robot

In this study, a graphic simulator that is used to simulate problems related to kinematics and dynamics for an exoskeleton robot arm with 5 degrees of freedom (DoF) was presented. The graphic simulator utilized the advantages of design software SolidWorks, Catia, and the computing and simulation power of SimMechanics Toolbox in Matlab. The core of the proposed graphic simulator is algorithm to solve the kinematics and dynamic problems of a developing upper-limb rehabilitation robot. The study used the proposed optimization-based algorithm to solve the inverse kinematics (IK) problem for the redundant robot model. Endpoint trajectories were imported from measurement data. The joints variable solutions obtained before entering the dynamics problem were smoothed to ensure feasibility in the later calculation process. A process to solve the inverse dynamics problem using physical model by combining the power of two software SolidWorks and SimMechanics was also proposed. This process ensured that the Robot’s design could be changed and updated to the kinematics calculation fast and easily. To evaluate this procedure, we also compared these dynamics results with results when applying the Lagrange–Euler formulation. All these calculation and simulation processes have been integrated into the graphic simulator software to show efficiency and user-friendliness.

Transient Open‐Closed Loop Experimental Validation of a Nonlinear Two‐Phase Flow Distributed System

AbstractThe oil well drilling process is a nonlinear system with transient nature. Conventional drilling is unable to assure safe and cost‐effective operation for fractured, cavernous, and highly permeable carbonate reservoirs, which contain the largest oil reserves worldwide. Concerning drilling technologies, Pressurized Mud Cap Drilling (PMCD) is suitable for the challenging scenario previously mentioned. According to PMCD technique, a sacrificial fluid is injected through the drill string and a light annular mud is pumped in countercurrent through the annulus region (bullheading), without surface return, forcing gas and drilled cuttings back to formation. A two‐phase flow distributed model (Drift Flux Model – DFM) is developed to properly describe the complex nature of the system. Also, an experimental facility, presenting field similarity, is employed to validate the open – closed loop schemes. The main objective of the controller (control reconfiguration with gain scheduling) is to regulate annulus pressure, handling gas kick, drilling fluid losses and inverse response dynamics. Besides, gas injection, migration and bullheading are studied. The simulations, validated through experimental data, highlight the methodology usefulness for field applications.

High-efficiency inverse dynamics modeling of parallel posture alignment mechanism with actuation redundancy

AbstractThe analytical expression of driving force is helpful to quickly plan the kinematic trajectory of parallel mechanism for automatic drilling and riveting. For parallel posture alignment mechanism, because of its closed-loop characteristics, the inverse dynamic solution is more complex, especially for parallel bracket with actuation redundancy. Considering that the telescopic rods are actually flexible parts, the dynamic analytical modeling is carried out with deformation supplementary equation. Taking the force at the spherical joint as the intermediate variable and the driving force of each active prismatic pair are analytically analyzed by vector cross-product. The modeling was verified by experiment. Compared with previous research methods, the analytical method proposed improves the solution accuracy of driving force slightly and reduces the driving force solution time by 56.28%, which is high efficiency. The maximum error percentage is 1.61%, and the experimental results show that the method of inverse dynamics modeling is practical. This paper can be used for driving force analysis of parallel posture alignment mechanism based on positioner in the field of aircraft assembly.

Electromechanical model of exoskeleton with three mobile links

The article considers an electromechanical model of an exoskeleton with three links, for which a system of differential equations of motion is written and inverse and direct problems of dynamics are solved. The angles between the links that specify anthropoid motion are figured out analytically. The electric drive controlling torques are found as a result of solving the inverse dynamics problem. The found torques are approximated with stepwise piecewise-constant functions simulating the impulse control of the exoskeleton motion. The solution of the direct problem of dynamics is carried out and the link rotation angles as functions of time are found. The results of the numerical solution of the system of differential equations are compared with the initial motion of the links. It has been found that the results of simulation with impulse control are in good agreement with the original motion. The total energy expenditures have been calculated. The exoskeleton simulation, taking into account the electric drive impacts also has been carried out. For this purpose, a system of differential equations of motion has been compiled; a numerical solution of the Cauchy problem for the compiled system has been carried out. The significance of the electric drive impact on the dynamics of the mechanism has been confirmed.

Dynamics Combined With Hill Model for Functional Electrical Stimulation Ankle Angle Prediction.

Musculoskeletal models play an essential role in ankle rehabilitation research. The majority of the existing models have established the relationship between EMG and joint torque. However, EMG signal acquisition requires higher clinical conditions, such as sensitivity to external circumstances, motion artifacts and electrode position. To solve the nonlinear and time-varying nature of joint movement, a Functional Electrical Stimulation (FES) model was proposed in this study to simulate the whole process of ankle dorsiflexion. The model is combined with muscle contraction dynamics based on Hill model and ankle inverse dynamics to connect FES parameters, torques, and ankle angles. In addition, the extended Kalman filter (EKF) algorithm was applied to identify the unknown parameters of the model. Model validation experiment was performed by acquiring the actual data of healthy volunteers. Results showed that the root mean square error (RMSE) and normalized root mean square error (NRMSE) of this model were 11.93%±0.53% and 1.39°±0.26°, respectively, which means it can effectively predict the output variation of ankle joint angle while changing electrical stimulation parameters. Therefore, the proposed mode is essential for developing closed-loop feedback control of electrical stimulation and has the potential to help patients to conduct rehabilitation training.