Dynamic Analysis of Flexible Multibody Systems for High-Rise Building Window Cleaning Gondola

A virtual prototype of high-rise building window cleaning gondola based on multibody system dynamics software MS（2. ADAMS is presented. The rigid bodies are modeled by CAD software and flexible bodies are modeled by discrete beam method. The whole machine＇s natural characteristics are analyzed and changed to frequency field. According to the results, the dangerous frequencies are avoided and the design can be optimized and the performance can be improved.

Computational dynamics of soft machines

Soft machine refers to a kind of mechanical system made of soft materials to complete sophisticated missions, such as handling a fragile object and crawling along a narrow tunnel corner, under low cost control and actuation. Hence, soft machines have raised great challenges to computational dynamics. In this review article, recent studies of the authors on the dynamic modeling, numerical simulation, and experimental validation of soft machines are summarized in the framework of multibody system dynamics. The dynamic modeling approaches are presented first for the geometric nonlinearities of coupled overall motions and large deformations of a soft component, the physical nonlinearities of a soft component made of hyperelastic or elastoplastic materials, and the frictional contacts/impacts of soft components, respectively. Then the computation approach is outlined for the dynamic simulation of soft machines governed by a set of differential-algebraic equations of very high dimensions, with an emphasis on the efficient computations of the nonlinear elastic force vector of finite elements. The validations of the proposed approaches are given via three case studies, including the locomotion of a soft quadrupedal robot, the spinning deployment of a solar sail of a spacecraft, and the deployment of a mesh reflector of a satellite antenna, as well as the corresponding experimental studies. Finally, some remarks are made for future studies.

A Geometrically Exact Triangular Shell Element Based on Reproducing Kernel DMS-Splines

Tomodel amultibody systemcomposed of shell components,a geometrically exact Kirchho-Love triangular shell element is proposed.The middle surface of the shell element is described by using the DMS-splines,which can exactly represent arbitrary topology piecewise polynomial triangular surfaces.The proposed shell element employs only nodal displacement and can automatically maintain C1 continuity properties at the element boundaries.A reproducing DMS-spline kernel skill is also introduced to improve computation stability and accuracy.The proposed triangular shell element based on reproducing kernel DMS-splines can achieve an almost optimal convergent rate.Finally,the proposed shell element is validated via three static problems of shells and the dynamic simulation of aexible multibody system undergoing both overall motions and large deformations.

A multiscale differential-algebraic neural network-based method for learning dynamical systems

The objective of dynamical system learning tasks is to forecast the future behavior of a system by leveraging observed data.However,such systems can sometimes exhibit rigidity due to significant variations in component parameters or the presence of slow and fast variables,leading to challenges in learning.To overcome this limitation,we propose a multiscale differential-algebraic neural network(MDANN)method that utilizes Lagrangian mechanics and incorporates multiscale information for dynamical system learning.The MDANN method consists of two main components:the Lagrangian mechanics module and the multiscale module.The Lagrangian mechanics module embeds the system in Cartesian coordinates,adopts a differential-algebraic equation format,and uses Lagrange multipliers to impose constraints explicitly,simplifying the learning problem.The multiscale module converts high-frequency components into low-frequency components using radial scaling to learn subprocesses with large differences in velocity.Experimental results demonstrate that the proposed MDANN method effectively improves the learning of dynamical systems under rigid conditions.

Multibody system transfer matrix method:The past, the present, and the future

The multibody system transfer matrix method(MSTMM),a novel dynamics approach developed during the past three decades,has several advantages compared to conventional dynamics methods.Some of these advantages include avoiding global dynamics equations with a system inertia matrix,utilizing low‐order matrices independent of system degree of freedom,high computational speed,and simplicity of computer implementation.MSTMM has been widely used in computer modeling,simulations,and performance evaluation of approximately 150 different complex mechanical systems.In this paper,the following aspects regarding MSTMM are reviewed:basic theory,algorithms,simulation and design software,and applications.Future research directions and generalization to more applications in various fields of science,technology,and engineering are discussed.

Reduced multibody system transfer matrix method using decoupled hinge equations

In the multibody system transfer matrix method(MSTMM),the transfer matrix of body elements may be directly obtained from kinematic and kinetic equations.However,regarding the transfer matrices of hinge elements,typically information of their outboard body is involved complicating modeling and even resulting in combinatorial problems w.r.t.various types of outboard body's output links.This problem may be resolved by formulating decoupled hinge equations and introducing the Riccati transformation in the new version of MSTMM called the reduced multibody system transfer matrix method in this paper.Systematic procedures for chain,tree,closed-loop,and arbitrary general systems are defined,respectively,to generate the overall system equations satisfying the boundary conditions of the system during the entire computational process.As a result,accumulation errors are avoided and computational stability is guaranteed even for huge systems with long chains as demonstrated by examples and comparison with commercial software automatic dynamic analysis of the mechanical system.

Robust Machine Learning Mapping of sEMG Signals to Future Actuator Commands in Biomechatronic Devices

A machine learning model for regression of interrupted Surface Electromyography(sEMG)signals to future control-oriented signals(e.g.,robot’s joint angle and assistive torque)of an active biomechatronic device for high-level myoelectric-based hierarchical control is proposed.A Recurrent Neural Network(RNN)was trained using output data,initially obtained from offline optimization of the biomechatronic(human–robot)device and shifted by the prediction horizon.The input of the RNN consisted of interrupted sEMG signals(to mimic signal disconnections)and previous kinematic signals of the assistive system.The RNN with a 0.1-s prediction horizon could predict the control-oriented joint angle and assistive torque with 92%and 86.5%regression accuracy,respectively,for the test dataset.This proposed approach permits a fast,predictive,and direct estimation of control-oriented signals instead of an iterative process that optimizes assistive torque in the inverse dynamic simulation of a multibody human–robot system.Training with these interrupted input signals significantly improves the regression accuracy in the case of sEMG signal disconnection.This Robust Predictive Control-oriented Machine Learning(Robust-MuscleNET)model can support volitional high-level myoelectric-based control of biomechatronic devices,such as exoskeletons,prostheses,and assistive/resistive robots.Future work should study the application to prosthesis control as well as the repeatability of the high-level controller with electrode shift.The low-level hierarchical controller that manages the human–robot interaction,the assistance/resistance strategy,and the actuator coordination should also be studied.

New ANCF solid-beam element:relationship with Bézier volume and application on leaf spring modeling

An eight-node solid-beam element based on absolute nodal coordinate formulation(ANCF)which uses cubic interpolation at the longitudinal direction and linear at the transverse direction is proposed.The element can accurately discretize the geometry represented by the Bézier volume with the same basis function order.The continuity property of the proposed beam element is verified.Advantages of the proposed element can be found in the application of leaf spring modeling.The parabolically varying thickness of the leaf can be accurately described.Components in leaf spring such as the spring eye and the rubber bushing can be assembled efficiently since the proposed element can be connected at any direction.The ANCF reference node is used to represent the rigid components and the revolute joints in the vehicle suspension system.Static I-shaped cross-section cantilever beam and flexible pendulum model are used to test the accuracy and dynamic performance of the proposed solid-beam element.The leaf spring model shows the flexibility of the element in modeling the complex mechanical system and the balance between the accuracy and the efficiency.

Modeling Method and Application of Rational Finite Element Based on Absolute Nodal Coordinate Formulation

In multibody system dynamics, the absolute nodal coordinate formulation （ANCF） uses power functions as interpolating polynomials to describe the displacement field. It can get accurate results for flexible bodies that undergo large deformation and large rotation. However, the power functions are irrational representation which cannot describe the complex shapes precisely, especially for circular and conic sections. Different from the ANCF representation, the rational absolute nodal coordinate formulation （RANCF） utilizes rational basis functions to describe geometric shapes, which allows the accurate representation of complicated displacement and deformation in dynamics modeling. In this paper, the relationships between the rational surface and volume and the RANCF finite element are provided, and the generalized transformation matrices are established correspondingly. Using these transformation matrices, a new four-node three-dimensional RANCF plate element and a new eight-node three-dimensional RANCF solid element are proposed based on the RANCF. Numerical examples are given to demonstrate the applicability of the proposed elements. It is shown that the proposed elements can depict the geometric characteristics and structure configurations precisely, and lead to better convergence in comparison with the ANCF finite elements for the dynamic analysis of flexible bodies.