Control of mechanical systems with uncertain parameters by means of small forces

An approach to the construction of a feedback control for non-linear Lagrange mechanical systems with uncertain parameters is developed. A Lagrange mechanical system with uncertain parameters, which is subject to the action of potential forces, control forces and unknown perturbations is considered is considered. It is assumed that the potential forces can be considerably greater than the control forces which, in their turn, are greater than the perturbations. An approach to the construction of a control, is proposed which enables one to bring a system from an arbitrary initial state to a specified final state in a finite time using a bounded control. A procedure, in which the specified nominal trajectory of the motion is tracked, is used. Initially, the trajectory, joining the specified initial and final states of the system, is constructed for a certain dynamical system which is close to the initial system but with completely known parameters. Then, using deviation equations, a control is constructed which brings the initial system onto this nominal trajectory in a finite time and subsequently forces the system to move along this nominal trajectory up to the final state. The control law used in tracking the nominal trajectory is based on a linear feedback, the gains of which depends on the discrepancy between the real trajectory and the nominal trajectory. The gain increase and tend to infinity as the discrepancies tend to zero but the control forces remain bounded and satisfy the conditions imposed on them. The results of numerical modelling of the controlled motions of a plane double pendulum are presented as an illustration.     

Optimizing system performance in the event of feedback failure

The problem of maintaining acceptable performance of a perturbed control system during a disruption of the feedback signal is addressed. The objective is to maximize the time during which performance remains within desirable bounds without feedback, given that the parameters of the controlled system are within a specified neighborhood of their nominal values. The existence of an optimal open-loop controller that achieves this objective is proved. It is also shown that a bang-bang input can approximate the performance of the optimal controller. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

On the efficiency of optimal grid synthesis in optimal control problems with fixed terminal time

In the paper, we consider nonlinear optimal control problems with the Bolza functional and with fixed terminal time. We suggest a construction of optimal grid synthesis. For each initial state of the control system, we obtain an estimate for the difference between the optimal result and the value of the functional on the trajectory generated by the suggested grid positional control. The considered feedback control constructions and the estimates of their efficiency are based on a backward dynamic programming procedure. We also use necessary and sufficient optimality conditions in terms of characteristics of the Bellman equation and the sub-differential of the minimax viscosity solution of this equation in the Cauchy problem specified for the fixed terminal time. The results are illustrated by the numerical solution of a nonlinear optimal control problem.     

On the complexity of solving feasible systems of linear inequalities specified with approximate data

An algorithm that gives an approximate solution of a desired accuracy to a system of linear inequalities specified with approximate data is presented. It uses knowledge that the actual instance is feasible to reduce the data precision necessary to give an approximate solution to the actual instance. In some cases, this additional information allows problem instances that are ill-posed without the knowledge of feasibility to be solved.The algorithm is computationally efficient and requires not much more data accuracy than the minimal amount necessary to give an approximate solution of the desired accuracy. This work aids in the development of a computational complexity theory that uses approximate data and knowledge.     

When are NP-hard location problems easy?

We discuss in this paper several location problems for which it is an NP-hard problem to find an approximate solution. Given certain assumptions on the input distributions, we present polynomial algorithms that deliver a solution asymptotically close to the optimum with probability that is asymptotically one (the exact nature of this asymptotic convergence is described in the paper). In that sense the subproblems defined on the specified family of inputs are in fact easy.This research was supported in part by the National Science Foundation under grant ECS-8204695.     

Forced nonlinear vibrations of a viscoelastic body

An approximate solution of the problem of the forced, geometrically nonlinear vibrations of an arbitrary viscoelastic body is found in the form of an expansion in eigenfunctions of the corresponding linear elastic problem. With the aid of the virtual displacement principle the problem is reduced to a system of nonlinear integro-differential equations whose periodic solution is constructed by the small-parameter method.     

Inverse problem for a quasilinear system of partial differential equations with a nonlocal boundary condition

An initial boundary-value problem for a quasilinear system of partial differential equations with a nonlocal boundary condition involving a delayed argument is considered. The existence of a unique solution to this problem is proved by reducing it to a system of nonlinear integral-functional equations. The inverse problem of finding a solution-dependent coefficient of the system from additional information on a solution component specified at a fixed point of space as a function of time is formulated. The uniqueness of the solution of the inverse problem is proved. The proof is based on the derivation and analysis of an integral-functional equation for the difference between two solutions of the inverse problem.     

Multirate sampling and input-to-state stable receding horizon control for nonlinear systems

In this paper, a robust receding horizon control for multirate sampled-data nonlinear systems with bounded disturbances is presented. The proposed receding horizon control is based on the solution of Bolza-type optimal control problems for the approximate discrete-time model of the nominal system. “Low measurement rate” is assumed. It is shown that the multistep receding horizon controller that stabilizes the nominal approximate discrete-time model also practically input-to-state stabilizes the exact discrete-time system with disturbances.     

Maximum Divert for Planetary Landing Using Convex Optimization

This paper presents a real-time solution method of the maximum divert trajectory optimization problem for planetary landing. In mid-course, the vehicle is to abort and retarget to a landing site as far from the nominal as physically possible. The divert trajectory must satisfy velocity constraints in the range and cross range directions and a total speed constraint. The thrust magnitude is bounded above and below so that once on, the engine cannot be turned off. Because this constraint is not convex, it is relaxed to a convex constraint and lossless convexification is proved. A transformation of variables is introduced in the nonlinear dynamics and an approximation is made so that the problem becomes a second-order cone problem, which can be solved to global optimality in polynomial time whenever a feasible solution exists. A number of examples are solved to illustrate the effectiveness and efficiency of the solution method.     

A new filled function method for an unconstrained nonlinear equation

In this paper, we transform an unconstrained system of nonlinear equations into a special optimization problem. A new filled function is constructed by employing the special properties of the transformed optimization problem. Theoretical and numerical properties of the proposed filled function are investigated and a solution of the algorithm is proposed. Under some conditions, we can find a solution or an approximate solution to the system of nonlinear equations in finite iterations. The implementation of the algorithm on six test problems is reported with satisfactory numerical results.     

Optimal control for quasi-linear systems with small parameters

We consider the controlled systems where the non-linear term is multiplied by a small scalar parameter ε. In the class of these quasi-linear systems, we shall determine the control and optimal trajectory which minimizes the index of performance represented by quadratics functionals. The initial and final conditions are specified and the final time is free. The presence of the small parameter leads to an approximate solution of the formulated problem of optimum. Thus, the zeroth-order solution is obtained for ε=0. The first order solution results by using the sweep method which determines the perturbation of the control and of the state variable on the optimal neighboring trajectory.     

Solving of second order nonlinear PDE problems by using artificial controls with controlled error

In this paper, we find the approximate solution of a second order nonlinear partial differential equation on a simple connected region inR ². We transfer this problem to a new problem of second order nonlinear partial differential equation on a rectangle. Then, we transformed the later one to an equivalent optimization problem. Then we consider the optimization problem as a distributed parameter system with artificial controls. Finally, by using the theory of measure, we obtain the approximate solution of the original problem. In this paper also the global error inL ₁ is controlled.     

Lebesgue piecewise affine approximation of nonlinear systems

This paper addresses a piecewise affine (PWA) approximation problem, i.e., a problem of finding a PWA system model which approximates a given nonlinear system. First, we propose a new class of PWA systems, called the Lebesgue PWA approximation systems, as a model to approximate nonlinear systems. Next, we derive an error bound of the PWA approximation model, and provide a technique for constructing the approximation model with specified accuracy. Finally, the proposed method is applied to a gene regulatory network with nonlinear dynamics, which shows that the method is a useful approximation tool.     

Tracking within a time interval on the basis of data supplied by finite observers

We solve the tracking control problem, in which one should bring a trajectory of a system of linear ordinary differential equations into a neighborhood of a trajectory of another system within a given time interval. After getting into this neighborhood, one should keep the trajectory of the first subsystem in it for a time interval of given duration. For the control synthesis, we use incomplete and imprecise information on the online deviation of one trajectory from the other, which is obtained in real time from linear equations of observation. We consider distinct structures of observers, which substantially affect the solution of control problems for such systems. The equations of dynamics and admissible measurements contain uncertainty for which one knows only some hard pointwise constraints. To solve the main problem, we use an approach that can be reduced to the construction of auxiliary information sets and weakly invariant sets with a subsequent “aiming” of one set at a tube. We suggest an efficient method for an approximate solution on the basis of ellipsoidal calculus techniques. The results of the algorithm operation are illustrated by an example of the solution of a tracking control problem for two fourth-order subsystems.     

Dual-Sampling-Rate Moving-Horizon Control of a Class of Linear Systems with Input Saturation and Plant Uncertainty

Moving-horizon control is a type of sampled-data feedback control in which the control over each sampling interval is determined by the solution of an open-loop optimal control problem. We develop a dual-sampling-rate moving-horizon control scheme for a class of linear, continuous-time plants with strict input saturation constraints in the presence of plant uncertainty and input disturbances. Our control scheme has two components: a slow-sampling moving-horizon controller for a nominal plant and a fast-sampling state-feedback controller whose function is to force the actual plant to emulate the nominal plant. The design of the moving-horizon controller takes into account the nonnegligible computation time required to compute the optimal control trajectory.We prove the local stability of the resulting feedback system and illustrate its performance with simulations. In these simulations, our dual-sampling-rate controller exhibits performance that is considerably superior to its single-sampling-rate moving-horizon controller counterpart.     

A new method for solving nonlinear second order partial differential equations

In this paper, a new method for finding the approximate solution of a second order nonlinear partial differential equation is introduced. In this method the problem is transformed to an equivalent optimization problem. Then, by considering it as a distributed parameter control system, the theory of measure is used for obtaining the approximate solution of the original problem.     

FINITE ELEMENT NONLINEAR GALERKIN COUPLING METHOD FOR THE EXTERIOR STEADY NAVIER-STOKES PROBLEM

1.IntroductionNonlinearGalerkinmethodsaremultilevelschemesforthedissipativeevolutionpartialdifferentialequations.Theycorrespondtothesplittingsoftheunknownu:u=y z)wherethecomponentsareofdifferentorderofmagnitudewithrespecttoaparameterrelatedtothespati...     

Robust hypothesis testing for asymmetric nominal densities under a relative entropy tolerance

In this paper, we address an open problem raised by Levy (2009) regarding the design of a binary minimax test without the symmetry assumption on the nominal conditional probability densities of observations. In the binary minimax test, the nominal likelihood ratio is a monotonically increasing function and the probability densities of the observations are located in neighborhoods characterized by placing a bound on the relative entropy between the actual and nominal densities. The general minimax testing problem at hand is an infinite-dimensional optimization problem, which is quite difficult to solve. In this paper, we prove that the complicated minimax testing problem can be substantially reduced to solve a nonlinear system of two equations having only two unknown variables, which provides an efficient numerical solution.     

An approach for solving nonlinear programming problems

In this paper we use measure theory to solve a wide range of the nonlinear programming problems. First, we transform a nonlinear programming problem to a classical optimal control problem with no restriction on states and controls. The new problem is modified into one consisting of the minimization of a special linear functional over a set of Radon measures; then we obtain an optimal measure corresponding to functional problem which is then approximated by a finite combination of atomic measures and the problem converted approximately to a finite-dimensional linear programming. Then by the solution of the linear programming problem we obtain the approximate optimal control and then, by the solution of the latter problem we obtain an approximate solution for the original problem. Furthermore, we obtain the path from the initial point to the admissible solution.     

Error Estimation for Nonlinear Complementarity Problems via Linear Systems with Interval Data

For the nonlinear complementarity problem, we derive norm bounds for the error of an approximate solution, generalizing the known results for the linear case. Furthermore, we present a linear system with interval data, whose solution set contains the error of an approximate solution. We perform extensive numerical tests and compare the different approaches.