Dynamic programming for impulse controls

This paper describes the theory of feedback control in the class of inputs which allow delta-functions and their derivatives. It indicates a modification of dynamic programming techniques appropriate for such problems. Introduced are physically realizable bang-bang-type approximations of the “ideal” impulse-type solutions. These may also serve as “fast” feedback controls which solve the terminal control problem in arbitrary small time. Examples of damping high-order oscillations in finite time are presented.     

Stability analysis of a class of digital filters utilizing single saturation nonlinearity

A novel criterion for the global asymptotic stability of a class of digital filters utilizing single saturation nonlinearity is presented. An example showing the effectiveness of the present criterion is given.     

A Reinforcement Learning Algorithm in Cooperative Multi-Robot Domains

Reinforcement learning has been widely applied to solve a diverse set of learning tasks, from board games to robot behaviours. In some of them, results have been very successful, but some tasks present several characteristics that make the application of reinforcement learning harder to define. One of these areas is multi-robot learning, which has two important problems. The first is credit assignment, or how to define the reinforcement signal to each robot belonging to a cooperative team depending on the results achieved by the whole team. The second one is working with large domains, where the amount of data can be large and different in each moment of a learning step. This paper studies both issues in a multi-robot environment, showing that introducing domain knowledge and machine learning algorithms can be combined to achieve successful cooperative behaviours.     

Reducing Symmetries to Generate Easier SAT Instances

Finding countermodels is an effective way of disproving false conjectures. In first-order predicate logic, model finding is an undecidable problem. But if a finite model exists, it can be found by exhaustive search. The finite model generation problem in the first-order logic can also be translated to the satisfiability problem in the propositional logic. But a direct translation may not be very efficient. This paper discusses how to take the symmetries into account so as to make the resulting problem easier. A static method for adding constraints is presented, which can be thought of as an approximation of the least number heuristic (LNH). Also described is a dynamic method, which asks a model searcher like SEM to generate a set of partial models, and then gives each partial model to a propositional prover. The two methods are analyzed, and compared with each other.     

Fragment shaders for agent animation using finite state machines

In a previous paper we generated animated agents and their behavior using a combination of XML and images. The behavior of agents was specified as a finite state machine (FSM) in XML. We used images to determine properties of the world that agents react to. While this approach is very flexible, it can be made much faster by using the power available in modern GPUs. In this paper we implement FSMs as fragment shaders using three kinds of images: world space images, agent space images and FSM table images. We show a simple example and compare performance of CPU and GPU implementations. Then we examine a more complex example involving more maps and two types of agents (predator–prey). Furthermore we explore how to render agents in 3D more efficiently by using a variation on pseudoinstancing.     

A finite volume method for solid mechanics incorporating rotational degrees of freedom

A novel finite volume (FV) based discretization method for determining displacement, strain and stress distributions in loaded two dimensional structures with complex geometries is presented. The method incorporates rotation variables in addition to the displacement degrees of freedom employed in earlier FV based structural analysis procedures and conventional displacement based finite element (FE) formulations. The method is used to predict the displacement fields in a number of test cases for which solutions are already available. The effect of mesh refinement upon the accuracy of the solutions predicted by the method is assessed. The results of this assessment indicate that the new method is more accurate than previous FV procedures incorporating displacement variables only, particularly in cases where bending is the predominant mode of deformation, and therefore the new method represents a significant advance in the development of this type of discretization procedure. Interestingly, the results of the accuracy assessment exercise also indicate that the new FV procedure is also more accurate than the equivalent FE formulation incorporating displacement and rotational degrees of freedom.     

Development of computational software for analysis of curved girders under construction loads

The analysis of horizontally curved, trapezoidal steel girders presents a variety of computational challenges. During the erection and construction stages before a concrete deck is available to form a closed section, these girders are weak in torsion and susceptible to warping. Considering the design of an entire bridge system, current design approaches favor the use of a grid analysis methodology. While the use of a grid analysis procedure offers the advantage of computational efficiency, it is unable to capture girder stresses and brace member forces with sufficient accuracy, particularly during the critical erection and construction stages. In this paper, we present an alternative analysis approach based on the finite element method. The developed software has been designed to be computationally efficient and easy to use for bridge designers.     

Experimental study and simulation of a hydraulic engine mount with fully coupled fluid-structure interaction finite element analysis model

Non-linear properties of a generic hydraulic engine mount (HEM) are identified and characterized by experiment and simulation approaches. The experimental methods for obtaining static and dynamic performances of the HEM are presented. The characteristics of two configurations of the take-apart HEMs (one is with an inertia track and a free decoupler and the other one is only with an inertia track) and their rubber springs are studied. The relations between static stiffness of an HEM and its rubber springs in three orthogonal directions are investigated. The influences of preload, excitation amplitudes and frequencies on the dynamic stiffness of an HEM are verified. The effects of the free decoupler on dynamic characteristics of the HEM are studied by comparisons of the test results of the two HEMs. The dynamic fluid pressure and the temperature in the upper chamber of an HEM are measured under different excitation conditions. A fully coupled fluid-structure interaction (FSI) and finite element analysis (FEA) model for simulation of HEMs is developed in this paper, which can be used to simulate the static and dynamic performances of the HEMs with only stress versus strain relations of the rubber materials, the fluid physical parameters and the HEMs sizes. The simulated results of one HEM with the proposed model are given, and the results match well with the measured data, or in coincidence with the working mechanisms of HEMs.     

Modelling the orthogonal machining process using coated carbide cutting tools

In this paper finite element methods were used to determine the influence of various coated and uncoated tungsten carbide cutting tools on the machining of a nickel-based super alloy Inconel 718. Disposable coated and uncoated carbide inserts were used both experimentally and as FEA models to study how the stress distribution within different coatings and carbide grades compared to each other, under a range of cutting conditions. Simulation of an orthogonal metal cutting process was performed using FORGE2, an elasto-visco plastic FEA code. All FE models were assumed to be plane strain. The results include the stress and temperature distributions through the primary shear zone, the chip/tool contact region and the coating/substrate boundaries. The tool wear and stress results from the FE modelling agree favourably with those obtained from experimental work.