Predictive dynamics: an optimization-based novel approach for human motion simulation

Predictive dynamics is a novel approach for simulating human motion. It avoids direct integration of differential-algebraic equations in order to create the resulting simulations for redundant digital human models. Instead, it formulates an optimization problem by defining appropriate performance measures and constraints to recover the real motion of the dynamic system. In the formulation, both kinematics and kinetics parameters serve as unknowns, and equations of motion are treated as equality constraints. Procedures to choose physical performance measures and appropriate constraints based on the available information about the bio-system are presented. The proposed methodology is illustrated and studied by first predicting the swinging motion of a single pendulum with externally applied torque. The pendulum can represent the motion of upper and lower extremities. This simple problem has analytical solutions and is used to gain insights for the predictive dynamics approach. In addition, a complex human walking task is simulated by using the approach, and realistic results are obtained. Such motion prediction capabilities have a wide variety of applications for industries ranging from automotive to military to clinical analysis and design.     

LCD models to analyze and simulate motion artifacts

Abstract— In this paper, two models to evaluate the temporal behavior of liquid‐crystal displays are described: a model assuming a linear display behavior and a model that incorporates non‐linear effects. For the linear temporal model, it can be predicted that the response time starts to contribute to motion blur when it is longer than one‐sixth of the hold time and becomes dominant when it is longer than eight times the hold time. The non‐linear model can be used to visualize the appearance of effects that cannot be determined via linear system theory. Also, some means to reduce display artifacts are described and its impact is illustrated. Although the main focus in this article is on the temporal behavior of liquid‐crystal displays, the spatial properties defined by the pixel structure can be simulated as well. A formula for the spatio‐temporal display behavior is given, which can be evaluated numerically to simulate the perceived image for arbitrary image‐sequence input material.     

Cofields: a physically inspired approach to motion coordination

As computing becomes increasingly pervasive, autonomous computers are going to be embedded in everyday objects in our physical environment. In such scenarios, mobility itself will be pervasive. Mobile users, mobile devices, computer-enabled vehicles, and mobile software components will define a dynamic, networked world in which a large set of autonomous components will interact with each other to orchestrate their activities. We focus on the problem of coordinating autonomous agent's movements in a distributed environment. Orchestrating mobile autonomous agents can take inspiration from the laws of physics. Agent's movements could be driven by locally perceived computational force fields, or cofields, generated by the agents themselves and propagated through an embedded infrastructure.     

Impulse-based substructuring in a floating frame to simulate high frequency dynamics in flexible multibody dynamics

An original approach for flexible multibody dynamics is proposed, which combines the free–free formulation of elastic body deformation with an impulse-based representation of linear vibration. The resulting system of equations being remarkably simple, this impulse-based substructuring method is straightforward to implement. Simple applications of a flexible rotating beam submitted to various excitation inputs have been selected and developed so as to assess the accuracy of the proposed methodology.     

A new approach to code motion and its application to hoisting

An approach to code motion and hoisting, a program optimization technique, is discussed. The safety and profitability of optimization in general and hoisting in particular are analyzed. By restricting the analysis to a spanning tree imposed on the program graph, a linear algorithm is developed that provides sufficient but not necessary conditions for hoisting.     

A new predictive motion control of a planar vehicle under uncertainty via convex optimization

This paper proposes an optimization-based planning framework for target tracking of a planar vehicle in the presence of some moving obstacles and uncertainty resources. The framework is modeled based on Linear-Time Varying Model Predictive Control (LTV-MPC). Moreover, a successive linearization approach is presented to formulate the obstacle avoidance mission as a set of linear constraints. Finally this LTV-MPC problem is translated to a close-form convex optimization problem by applying some mathematical and algebraic simplifications. Simulation results show beneficially of the proposed path planning method to navigate an autonomous vehicle successfully regardless of different uncertainty resources.     

Subspace clustering of high-dimensional data: a predictive approach

In several application domains, high-dimensional observations are collected and then analysed in search for naturally occurring data clusters which might provide further insights about the nature of the problem. In this paper we describe a new approach for partitioning such high-dimensional data. Our assumption is that, within each cluster, the data can be approximated well by a linear subspace estimated by means of a principal component analysis (PCA). The proposed algorithm, Predictive Subspace Clustering (PSC) partitions the data into clusters while simultaneously estimating cluster-wise PCA parameters. The algorithm minimises an objective function that depends upon a new measure of influence for PCA models. A penalised version of the algorithm is also described for carrying our simultaneous subspace clustering and variable selection. The convergence of PSC is discussed in detail, and extensive simulation results and comparisons to competing methods are presented. The comparative performance of PSC has been assessed on six real gene expression data sets for which PSC often provides state-of-art results.     

An engineering approach to human resources performance evaluation: Hybrid MCDM application with interactions

Human performance evaluation is one of the most important fields to analyze for the continuity of an organization. Evaluation files filled by the managers generally end up in dusty folders where no one looks. This decreases the credibility of the evaluators and the process itself. Whereas the management thinks that they are taking the valuable time from the people who can do better things instead of these evaluations. In this paper, we add an engineering point of view to this process by giving a Hybrid Multicriteria Decision Making (MCDM) approach to evaluate employees’ performances working for a same task and explain an efficient way of handling the qualitative and quantitative data simultaneously. The real life situations where performance criteria show interaction will be possible to solve and the different types of interactions will be handled with the proposed hybrid method using Analytical Network Process (ANP) and Choquet Integral (CI) simultaneously. We also give a numerical illustration at the end of the study with the appropriate concluding remarks including the advantages of the proposed method.     

混合设计方法

在系统设计师和管理者中间,有一场旷日持久的争论,就是到底是应该采用由上至下的设计技术还是采用由下至上的技术。经验证明,两种方法混合的效果最佳。     

Interactive videodiscs: a new approach to healthcare education

Linking computer and video technology has produced a new way to educate healthcare providers. These keyboard classrooms offer faster comprehension, better retention and lower costs than most traditional methods.     

Hybrid control: A constrained motion perspective

Control of manipulators during execution of tasks that require the end-effector to come into contact with objects in its work environment represents an important class of control problem. Hybrid control, a concept which defines the architecture of a class of control laws has been proposed as a method with which to solve this control problem. One interpretation of the hybrid control method is within the framework of constrained motion control. Constrained motion occurs during contact by the manipulator end-effector with rigid objects in the workplace, hence the motion of the manipulator is kinematically constrained. Papers have appeared in the robotics control literature addressing hybrid control within the constrained motion context which do not explicitly use the constrained dynamic formulation that correctly describes the dynamic behaviour of the manipulator. This article serves to link results from constrained motion control and the existing literature on the hybrid control method. In this article a formulation of the constrained manipulator dynamics is presented in which the hybrid control design is most naturally carried out. This formulation of the manipulator dynamics, previously proposed in the robotics literature, is such that the generalized force and position coordinates to be controlled are mutually orthogonal. Hence, the hybrid selection matrices, a key element of the hybrid control design philosophy, are implicit in this coordinate representation. A hybrid control design methodology is then formulated based on this development. Two hybrid control laws are proposed. For each hybrid control law, the global asymptotic stability is readily established due to the natural coordinate representation. One of these hybrid control laws, a constrained motion control law already proposed in the literature, is given to illustrate the equivalence of the hybrid control design and certain existing constrained motion control methods. Finally, a concrete example of a three-degree-of-freedom robotic manipulator illustrates the hybrid design methodology proposed.     

Cognitive robotics: a new approach to artificial intelligence

This paper describes a research program about how to achieve artificial intelligence by building robots. It is part of the behavior-oriented AI approach, but differs in some of its hypotheses and methodological approach. This work was presented, in part, at the Third International Symposium on Artificial Life and Robotics, Oita, Japan, January 19–21, 1998     

Satisficing: a new approach to constructive nonlinear control

The main contribution of this paper is a constructive parameterization of the class of almost smooth universal formulas which render a system asymptotically stable with respect to a known control Lyapunov function (CLF), and a constructive parameterization of a class of inverse optimal universal formulas having Kalman-like stability margins. The novelty of the parameterization is that it is given in terms of two function which are constrained to be locally Lipschitz and satisfy convex constraints. The implication of this result is that the CLF/universal formula approach can be combined with a priori performance objectives to design high performance control strategies. Two examples illustrate the approach.     

Optimal control of nonlinear systems: a predictive control approach

A new nonlinear predictive control law for a class of multivariable nonlinear systems is presented in this paper. It is shown that the closed-loop dynamics under this nonlinear predictive controller explicitly depend on design parameters (prediction time and control order). The main features of this result are that an explicitly analytical form of the optimal predictive controller is given, on-line optimisation is not required, stability of the closed-loop system is guaranteed, the whole design procedure is transparent to designers and the resultant controller is easy to implement. By establishing the relationship between the design parameters and time-domain transient, it is shown that the design of an optimal generalised predictive controller to achieve desired time-domain specifications for nonlinear systems can be performed by looking up tables. The design procedure is illustrated by designing an autopilot for a missile.     

The photodissociation of ozone: a quasi-classical approach to a quantum dynamics problem

It is extremely difficult to execute full quantum dynamics calculations on complex systems (more then three degrees of freedom) due to the exponential increase in computer resources required by these methods. Classical mechanics simulations do not suffer from this problem, but are unable to treat quantum mechanical phenomena such as non-adiabatic effects, which often play a vital role in photochemical processes. A method has been implemented for carrying out dynamical calculations using quasi-classical theory. The time dependent Schr?dinger equation is solved using a swarm of trajectories treated under Newtonian laws and Tully's fewest switches trajectory surface hopping is applied to implement the surface switches. The method was applied to ozone, looking at the photodissociation that takes place after excitation into the Chappuis band of the absorption spectrum. While the goal is to treat larger systems, comparison can be made for ozone with numerically exact wavepacket calculations. The method proved successful at calculating quantities such as the rate of population transfer, but there were discrepancies in the details, especially when surface switching occurred from the lower state.     

On the model granularity to simulate protein dynamics: A biological physics view on biomolecular computing

What granularity is needed to carry out computer simulations of biomolecular reactions/motions? This is one of the central issues of the in silico biomolecular computing. In this paper, we addressed this issue by studying model granularity dependence of the native structure dynamics of protein molecules. We conducted molecular dynamics simulations employing three different protein models: the model with full atomic details and two coarse-grained models in which only C_α atoms interacting with each other through simple potentials are considered. In addition to the observed agreement among the three models in terms of isotropic thermal fluctuation, principal component analysis showed that the coarse-grained models can also reproduce the anisotropy (or directionality) of the fluctuation, particularly of collective modes having relevance to molecular function. This indicates that the dependence of the essential dynamics of a protein molecule on the model granularity is weak, although it was also shown that incorporation of the Lennard–Jones-type potential into the harmonic-potential-based coarse-grained model improves the reproducibility to some degree, and that a plastic nature of structural dynamics observed in the full atomic model transforms into an elastic one in the coarse-grained models. The coarse-grained model can be applied to a molecular motor system, which may lead to a new view of biomolecular computing in the context of biological physics.     

Watermarking protocols: an excursus to motivate a new approach

Digital watermarking is one of the techniques that can be used to support digital copyright protection. It enables copyright owners to insert a perceptually invisible watermark within any copy of content that is distributed on the Internet. The main aim is to use watermarks to implement copyright identification and content tracking. However, digital watermarking is not sufficient to ensure an adequate copyright protection, if it is not employed in conjunction with watermarking protocols, which define the schemes of the web transactions by which buyers can purchase protected digital content distributed by content providers in a secure manner. Therefore, watermarking protocols play a central role in the field of digital copyright protection by means of watermarking techniques. They have to ensure both a correct content protection and an easy participation of buyers in the purchase transactions of the contents distributed on the Internet. This paper presents a new watermarking protocol based on an innovative, buyer centric, and mediated design approach. The result is a secure protocol that enables buyers to easily participate in the transactions by which they can buy digital protected content, and this makes the protocol suited for the current web context.     

Smart graphics: a new approach to meeting user needs

A convergence of computer graphics and other technologies such as artificial intelligence is leading to the development of smart graphics, which recognize application requirements, user characteristics, host-machine capabilities, and target usage, and adapt themselves accordingly. These new, dynamic capabilities extend the utility of computer graphics across a broad range of uses and potential applications. They may also engender the creation of new applications. For example, smart-graphics bar charts could change in accordance with fluctuating stock-market conditions. Developers could use smart graphics to customize GUIs and thus make them more useful. In addition, the technology has the potential to revolutionize graphics use in training, simulation, business productivity, and other applications