Inverse Dynamic Model Research Articles

Design of Dynamic Inversion and Explicit Model Following Flight Control Laws for Quadrotor UAS

The objectives of this paper are to advance dynamic inversion (DI) and explicit model following (EMF) flight control laws for quadrotor unmanned aerial systems (UAS) and to develop an efficient strategy to compute the stability and performance robustness statistics of such control laws given parametric model uncertainty. For this purpose, a parametric model of a quadrotor is identified from flight-test data. The identified model is validated both in frequency and time domains. Next, DI and EMF flight control laws are designed for both inner attitude and outer velocity loops. Finally, a novel approach based on an unscented transform is used to evaluate the statistics of the controller's performance based on the statistics of the uncertain model parameters.

Mechanical Impedance Control of Cooperative Robot During Object Manipulation Based on External Force Estimation Using Recurrent Neural Network

This paper proposes a compliant motion control for human-cooperative robots to absorb collision force when persons accidentally touch the robots even while the robot is manipulating an object. In the proposed method, an external force estimator, which can distinguish the net external force from the object manipulation force, is realized using an inverse dynamics model acquired by a recurrent neural network (RNN). By implementing a mechanical impedance control to the estimated external force, the robot can quickly and precisely carry the object keeping the mechanical impedance control functioned and can generate a compliant motion to the net external force only when the person touches it during manipulation. Since the proposed method estimates the external force from the generalized force based on the learned inverse dynamics, it is not necessary to install any sensors on the manipulated object to measure the external force. This allows the robot to detect the collision even when the person touches anywhere on the manipulated object. The RNN inverse dynamics model is evaluated by the leave-one-out cross-validation and it was found that it works well for unknown trajectories excluded from the learning process. Although the details were omitted due to the limitation of the page length, similar to the simulations, the RNN inverse dynamics model was evaluated using unknown trajectories in the six degree-of-freedom experiments, and it has been verified that it functions properly even for the unknown trajectories. Finally, the validity of the proposed method has been confirmed by experiments in which a person touches a robot while it is manipulating an object with six degrees of freedom.

A Case Study on Influence of Utilizing Hill-Type Muscles on Mechanical Efficiency of Biped Running Gait

The presence of compliant elements in biped running mechanisms generates a smoother motion and decreases the impact forces. Biological creatures that have a complicated actuation system with parallel and series elastic elements in their muscles demonstrate very efficient and robust bipedal gaits. The main difficulty of implementing these systems is duplicating their complicated dynamics and control. This paper studies the effects of an actuation system, including Hill-type muscles on the running efficiency of a kneed biped robot model with point feet. In this research, we implement arbitrary trajectories compatible with the initial condition of the robot, and we calculate the necessary muscle forces using an analytical inverse dynamics model. To verify the results, we execute the direct dynamics of the robot with the calculated control inputs to generate the robot’s trajectory. Finally, we calculate the contractile element force of the muscles and its cost of transport, and we investigate the effects of the muscles’ elements on reducing or increasing the cost of transport of the gait and maximum actuating forces.

Semiparametric Deep Learning Manipulator Inverse Dynamics Modeling Method for Smart City and Industrial Applications

In smart cities and factories, robotic applications require high accuracy and security, which depends on precise inverse dynamics modeling. However, the physical modeling methods cannot include the nondeterministic factors of the manipulator, such as flexibility, joint clearance, and friction. In this paper, the Semiparametric Deep Learning (SDL) method is proposed to model robot inverse dynamics. SDL is a type of deep learning framework, designed for optimal inference, combining the Rigid Body Dynamics (RBD) model and Nonparametric Deep Learning (NDL) model. The SDL model takes advantage of the global characteristics of classic RBD and the powerful fitting capabilities of the deep learning approach. Moreover, the parametric and nonparametric parts of the SDL model can be optimized at the same time instead of being optimized separately. The proposed method is validated using experiments, performed on a UR5 robotic platform. The results show that the performance of SDL model is better than that of RBD model and NDL model. SDL can always provide relatively accurate joint torque prediction, even when the RBD or NDL model is not accurate.

The effect of modifying foot progression angle on the knee loading parameters in healthy participants with different static foot postures

BackgroundStudies have found that toe-in gait reduced the peak knee adduction moment (KAM) during early stance, while toe-out gait reduced the peak KAM during late stance. However, some other studies found that toe-in or toe-out gait could reduce the KAM throughout stance phase. There is still a divergence of opinion on the use of toe-in or toe-out gait for reducing the KAM. Research questionThis study aimed to investigate whether static foot posture affected participants’ biomechanical responses to three self-selected foot progression angles (FPA): neutral, toe-out and toe-in. MethodsTwenty-seven healthy participants were recruited for this FPA gait modification experiment and classified into three groups: neutral (n = 8), supination (n = 9) and pronation (n = 10), based on the Foot Posture Index (FPI). The kinematic and kinetic data were recorded with Vicon motion capture system and three force plates. The knee adduction moment and ankle eversion moment were calculated using an inverse dynamics model. The effect of the FPA modification on the knee loading parameters was analysed by the Friedman non-parametric test. ResultsThe KAM results in the neutral group showed that the toe-in gait modification reduced the first peak of the KAM (KAM1), while the KAM1 was increased in the supination group. The effect of the FPA modification on the KAM1 did not reach significance in the pronation group. The toe-out gait modification reduced the second peak (KAM2) regardless of the static posture. SignificanceDifferent static foot postures were correlated with different peak KAM during the early stance phase due to FPA modification. These data suggest that the assessment of static foot posture provides a reference on how to offer adequate FPA modification for knee OA patients with different foot postures.

Forward and inverse dynamics modeling of human shoulder-arm musculoskeletal system with scapulothoracic constraint

In this work, we extend the modeling techniques of human shoulder-arm musculoskeletal dynamics by 1) proposing an extended model with 12 joint degrees of freedom and 27 muscles (modeled as 42 musculotendinous actuators) that is capable of most physiologically and anatomically possible movements, 2) proposing a forward dynamics model driven by muscle activation, where the scapulothoracic constraint is formulated as an anatomically consistent force field, and 3) applying the state-of-the-art inverse dynamics solution on this model. We experimentally validate it against electromyograms for 10 activities of daily living. This validated shoulder-arm musculoskeletal model may e.g., serve as a reference plant model in studying human motor control or as part of a human simulator in the future.

Accounting for dynamic losses in the Jiles-Atherton model of magnetic hysteresis

When designing semiconductor pulse converters of power systems operating in peak load conditions, it is necessary to calculate their characteristics and parameters, taking into account the magnetization reversal processes in the magnetic circuits of transformers and chokes. Magnetic materials of magnetic cores exhibit nonlinear, hysteretic, and dynamic properties due to the presence of classical vortex and additional losses. Various dynamic models, based on the classical Jiles–Atherton (JA) approach, are presented in the literature, however, they are unstable for textured materials and cannot be explained from a physical point of view. These models show certain non-physical situations expressed by the deformation of the hysteresis loop in the saturation region. In this paper, the direct and inverse dynamic hysteresis models are presented, based on the application of the theory of field separation. This approach explains some of the energy aspects of the JA model.

Semi-active control of a swing phase dynamic model of transfemoral prosthetic device based on inverse dynamic model

Improving the gait of transfemoral amputees and making it biomimetic and stable has always been a major effort. A dynamic model of the prosthetic device can predict the kinetic and kinematic performances, when incorporated with a musculoskeletal model. In this regard, a dynamic model of a recent trend of variable damping technology will help a great deal in evaluating the performance of the prosthetic device and also in studying the effect of various parameters on the prostheses. The current paper presents the dynamic model of a single axis two segmental prosthetic knee implemented with a magneto-rheological (MR) damper as a variable damping element. The MR damper is modeled mathematically using Bouc–Wen model with model parameters evaluated by minimizing the error norms for time, displacement and velocity between the experimental and the model-generated results using a genetic algorithm. Two different experimental data sets are used, one for mathematical modeling and other to assess the accuracy of the fit model. A Proportional Derivative plus Controlled Torque controller is employed, and the parameters are tuned to minimize the error between the desired and control input torques. Further, an inverse dynamic model using Bouc–Wen model variables is assumed and validated later. This model predicts the current directly and avoids the necessity of solving any quadratic equation, which is required in the case of inverse models based on modified Bouc–Wen. The dynamic model of the prosthesis is analyzed for the swing phase alone, and the results show that the model traces the desired knee angle and also the shank reaches full knee extension at the end of this phase with terminal velocity small enough to be handled by an extension stop.

Estimating upper extremity joint loads of persons with spinal cord injury walking with a lower extremity powered exoskeleton and forearm crutches

Lower extremity powered exoskeletons with crutch support can provide upright mobility to persons with complete spinal cord injury (SCI); however, crutch use for balance and weight transfer may increase upper extremity (UE) joint loads and injury risk. This research presented the first exoskeleton-human musculoskeletal model to estimate upper extremity biomechanics, driven by 3D motion data of persons with complete SCI walking with an exoskeleton and crutch assistance. Forearm crutches instrumented with strain gauges, force plates, and a 3D motion capture system were used to collect kinematic and kinetic data from five persons with complete SCI while walking with the ARKE exoskeleton. Model output estimated participant upper extremity kinematics, kinetics, and crutch forces. Compared to inverse dynamic biomechanical crutch model studies of persons with incomplete SCI, exoskeleton users walked with more anterior trunk tilt and twice the shoulder flexion angle. Anterior tilt increased forces and moments at the crutch, shoulder, and elbow. Crutch floor contact periods were 30–40% longer, resulting in upper extremity joint impulses 5 to 12 times greater than previously reported. Reducing UE joint loading is important to reduce overuse injuries associated with ambulatory assistive devices. Incorporating a variable assist ankle joint or more experience with exoskeleton walking may reduce UE joint loads, and minimise injury risk. Study outcomes provide a quantitative understanding of UE dynamics during exoskeleton walking that can be used to improve device design, training, and rehabilitation.

Trajectory-tracking of 6-RSS Stewart-Gough manipulator by feedback-linearization control using a novel inverse dynamic model based on the force distribution algorithm

ABSTRACT 6-RSS Stewart-Gough parallel manipulator contains six crank-rod limbs connecting the base and moving platforms to each other, forming a 6DOF manipulator. In this paper, we introduce a novel decoupled inverse dynamic model for this manipulator based on the Force Distribution Algorithm. The performance of the proposed model was evaluated in tracking a complex trajectory (of multiple segments with simultaneous translational and rotational motions) using feedback-linearization control in the joint space and compared with that of the Lagrangian inverse dynamic model. Results showed that this model leads to a better performance in feedback-linearization control, especially when the reference trajectory is quantized, and with less calculation burden in comparison with the Lagrangian model. The control system employing both models showed robustness against payload uncertainty on the moving platform (150% of the moving platform’s mass). The performance assessment and the robustness approval were performed in simulation using a Simscape model specifically built for this purpose in the Simulink environment.

Evaluation of musculoskeletal modelling parameters of the shoulder complex during humeral abduction above 90°

Based on electromyographic data and force measurements within the shoulder joint, there is an indication that muscle and resulting joint reaction forces keep increasing over an abduction angle of 90°. In inverse dynamics models, no single parameter could be attributed to simulate this force behaviour accordingly. The aim of this work is to implement kinematic, kinetic and muscle model modifications to an existing model of the shoulder (AnyBody™) and assess their single and combined effects during abduction up to 140° humeral elevation. The kinematics and the EMG activity of 10 test subjects were measured during humeral abduction. Six modifications were implemented in the model: alternative wrapping of the virtual deltoid muscle elements, utilization of a three element Hill model, strength scaling, motion capture driven clavicle elevation/protraction, translation of the GH joint in dependency of the acting forces and an alteration of the scapula/clavicle rhythm. From the six modifications, 16 different combinations were considered. Parameter combinations with the Hill model changed the resultant GH joint reaction force and led to an increase in force during abduction of the humerus above 90°. Under the premise of muscle activities and forces within the GH joint rising after 90° of humeral abduction, we propose that the Hill type muscle model is a crucial parameter for accurately modelling the shoulder. Furthermore, the outcome of this study indicates that the Hill model induces the co-contraction of the muscles of the shoulder without the need of an additional stability criterion for an inverse dynamics approach.

Mechanical Design and Performance Analysis of a Novel Parallel Robot for Ankle Rehabilitation

Abstract As the population ages, increasingly more individuals experience ankle disabilities caused by stroke and cerebral palsy. Studies on parallel robots for ankle rehabilitation have been conducted under this circumstance. This paper presents a novel parallel ankle rehabilitation robot with the key features of a simple configuration and actuator nonredundancy. The mechanical design is determined, and a prototype is built. Additionally, inverse position solution is addressed to calculate the workspace of the parallel robot. Jacobian matrices mapping the velocity and force from the active joint space to the task space are derived, and kinetostatic performance indices, namely, motion isotropy, force transfer ratio, and force isotropic radius are defined. Moreover, the inverse dynamic model is presented using the Newton–Euler formulation. Dynamic evaluation index, i.e., dynamic uniformity, is proposed according to the derived Jacobian matrix and inertia matrix. Based on the workspace analysis, the parallel robot demonstrates a sufficient workspace for ankle rehabilitation compared with measured range of motion of human ankle joint complex. The results of the kinetostatic and dynamic performance analysis indicate that the parallel robot possesses good motion isotropy, high force transfer ratio, large force isotropic radius, and relatively uniform dynamic dexterity within most of the workspace, especially in the central part. A numerical example is presented to simulate the rehabilitation process and verify the correctness of the inverse dynamic model. The simplicity and the performance of the proposed robot indicate that it has the potential to be widely used for ankle rehabilitation.

Effects of human stature and muscle strength on the standing strategies: A computational biomechanical study

It has been hypothesized that the muscular efforts exerted during standing may be altered by changes in personal factors, such as the body stature and muscular strength. The goal of this work was to assess the contribution of leg muscles using a biomechanical model in different physical conditions and various initial postures. An optimized inverse dynamics model was employed to find the maximum muscular effort in 23,040 postures. The simulation results showed that mid-range knee flexion could help the healthy and strong individuals maintain balance, but those with weaker muscle strength required more knee flexion. Individuals of weak muscular constitution as well as those with tall stature are at the highest risk of imbalance/falling. The number of imbalanced postures due to deficits in the calf and hamstring muscles was reduced by 7.5 times by strengthening the whole body musculature. The calf and the hamstring muscles play a key role in balance regardless of stature.

Effect of suprahyoid muscles on mouth opening with and without lateral pterygoid muscle: 3D inverse dynamic model analysis

ABSTRACT Objective: To evaluate the contribution of suprahyoid muscles to mouth opening in different hyoid bone positions. Methods: The jaw-opening and closing movements were imposed on the 3D inverse dynamic jaw model with and without the lateral pterygoid muscle (LPM). The activation of the muscles was evaluated for eight different positions of the hyoid bone. Results: The muscles with 100% activation provided maximum mouth opening (MMO). When the hyoid bone was replaced, the muscles could achieve MMO at the 135º, 180º, and 225º points with the LPM. Mouth opening was estimated to be 36.8 mm in the absence of the LPM. A jaw opening greater than 36.8 mm was seen when the hyoid bone was repositioned at the 90º, 180º, and 135º points. Discussion: The contribution of suprahyoid muscles to mouth opening varies in different hyoid bone positions, with the inferior and anterior positions having a positive impact.

A new solution for machining with RA-PKMs: Modelling, control and experiments

In this paper, a novel 5-Degree of Freedom (DOF) Redundantly Actuated (RA) Parallel Kinematic Manipulator (PKM) called SPIDER4 is presented. The main purpose of this manipulator is to perform machining tasks such as drilling and milling. All the mathematical models including the forward and inverse kinematic models, as well as the inverse dynamic model were developed. Owing to machining tasks require high precision, a RISE Feedforward controller is proposed for desired trajectory tracking. To show the performance and effectiveness of the proposed control scheme, real-time experiments were performed. The obtained results of the proposed controller compared to the standard RISE controller are presented and discussed. They confirm that the proposed controller outperforms the standard one.

The Effect of Meteorological Factors, Seasonal Factors and Air Pollutions on the Formation of Particulate Matter

To investigate the effect of meteorological factors, seasonal factors and Air Pollutions on the formation of particulate matter (PM2.5) in Jinan, China. Nonlinear dynamic inversion model was established to analyze the effect of meteorological factors, seasonal factors and CO, PM10, SO2, NO2, O3 on PM2.5 formations in different seasons. Temperature has a great influence on PM2.5 concentration variation. Precipitation exacerbate the formation of PM2.5 in Winter. Wind speed make a little contribution to PM2.5 formation in Jinan during different season. The formation of PM2.5 was influenced by confounding factors.

Receding-Horizon Vision Guidance with Smooth Trajectory Blending in the Field of View of Mobile Robots

Applying computer vision to mobile robot navigation has been studied for over two decades. One of the most challenging problems for a vision-based mobile robot involves accurately and stably tracking a guide path in the robot limited field of view under high-speed manoeuvres. Pure pursuit controllers are a prevalent class of path tracking algorithms for mobile robots, while their performance is rather limited to relatively low speeds. In order to cope with the demands of high-speed manoeuvres, a multi-loop receding-horizon control framework, including path tracking, robot control, and drive control, is proposed in this paper. This is done within the vision guidance of differential-driving wheeled mobile robots (DWMRs). Lamé curves are used to synthesize a trajectory with G 2 -continuity in the field of view of the mobile robot for path tracking, from its current posture towards the guide path. The platform twist—point velocity and angular velocity—is calculated according to the curvature of the Lamé-curve trajectory, then transformed into actuated joint rates by means of the inverse-kinematics model; finally, the motor torques needed by the driving wheels are obtained based on the inverse-dynamics model. The whole multi-loop control process, initiated from Lamé-curve blending to computational torque control, is conducted iteratively by means of receding-horizon guidance to robustly drive the mobile robot manoeuvring close to the guide path. The results of numerical simulation show the effectiveness of our approach.

Real-Time Impedance Control Based on Learned Inverse Dynamics

Impedance control of robotic manipulators allows them to perform interaction task where safety is the foremost requirement. A major hurdle in the implementation of impedance control has always been the requirement of availability and online real-time computation of dynamic model. This paper presents a complete data-driven, machine learning approach to impedance control in real time of an industrial manipulator. The technique used here to learn the inverse dynamic model of an industrial robot is based on an incremental nonparametric statistical learning approach and is called locally weighted projection regression. Unlike traditional model-based control schemes, the proposed control strategy requires less a priori knowledge of the system being modeled, is computationally efficient to run in hard real time, and can be updated during online operation of the arm. The main contribution of this paper is the development of learning-based impedance control scheme and its implementation in hard real time (at 500 Hz control loop rate) on an industrial robot (Barrett WAM arm). To validate performance of the proposed scheme, its comparison with a controller based on an analytically derived model (computed online in a numeric recursive way using Newton–Euler approach) has also been presented through experimentation in a laboratory environment.

Dynamic Hysteresis Model and Control Methodology for Force Output Using Piezoelectric Actuator Driving

In the application of piezoelectric jet dispensing, piezoelectric ceramics and flexure hinges cooperate to form a piezoelectric actuator. The dispensing quality of the piezoelectric jet is closely associated with the output force of the piezoelectric actuator. In this study, a corresponding piezoelectric actuator output force experiment platform is built, and the voltage-force hysteresis curve of the piezoelectric actuator is analyzed. The voltage-force hysteresis curve exhibits memory characteristics and rate dependence. A modeling method for the dynamic hysteresis output force model is proposed, and the numerical method of the model is presented. The corresponding output force model of the piezoelectric actuator is established based on experimental data to predict the output force under different input voltages. According to the dynamic hysteresis model, the feedforward control methodology of the dynamic hysteresis inverse model is proposed, and experimental tests are performed. Experimental results show that the proposed model can express the complex nonlinear characteristics of piezoelectric actuators, and the output force feedforward control method based on the dynamic hysteresis inverse model effectively decreases the nonlinear characteristics of the piezoelectric actuator.

Procena otpornosti kolena primenom kompjuterskih metoda

In tennis, the complex serving motions produce high mechanical stresses on player's musculoskeletal, tendon and ligament joints. In this paper, different cognitive methods have been integrated in order to non-invasively assess the knee's bone and cartilage resistance at the maximum power tennis serve. The proposed methodology is based on the creation of patientspecific biomechanical model, as well as on the tracking the knee's kinematics, ground force measurement, inverse dynamics modelling and analysis of the knee using the Finite Element Method with aim to assess the knee resistance of a tennis player, considering acute deformations and potential injuries. The main objective of this paper is development of the optimised computational technology and creation of practical diagnostic tool for non-invasive assessment of the knee function during specific moves and motions in tennis. It is expected that this approach can provide prediction and injury prevention in training and competitive tennis to a significant extent.