Solution of nonlinear optimal control problems in Hilbert spaces by means of linear programming techniques

Optimal control problems in Hilbert spaces are considered in a measure-theoretical framework. Instead of minimizing a functional defined on a class of admissible trajectory-control pairs, we minimize one defined on a set of measures; this set is defined by the boundary conditions and the differential equation of the problem. The new problem is an infinite-dimensionallinear programming problem; it is shown that it is possible to approximate its solution by that of a finite-dimensional linear program of sufficiently high dimensions, while this solution itself can be approximated by a trajectory-control pair. This pair may not be strictly admissible; if the dimensionality of the finite-dimensional linear program and the accuracy of the computations are high enough, the conditions of admissibility can be said to be satisfied up to any given accuracy. The value given by this pair to the functional measuring the performance criterion can be about equal to theglobal infimum associated with the classical problem, or it may be less than this number. It appears that this method may become a useful technique for the computation of optimal controls, provided the approximations involved are acceptable.     

Extremal points and optimal control theory

Summary An optimal control problem is considered in a setting akin to that of the theory. of generalized curves. Rather than minimizing a functional depending on pairs of trajectories and controls subject to some constraints, a functional defined on a set of Radon measures is considered; the set of measures is determined by the constraints. An approximation scheme is developed, so that the solution of the optimal control problems can be effected by solving a sequence of nonlinear programming problems. Several existence theorems for this kind of generalized control problems are then proved; the most interesting is the one concerning problems in which the set of allowable controls is unbounded. Entrata in Redazione il 5 febbraio 1975.     

Necessary conditions for optimality in relaxed stochastic control problems

In this paper, we are concerned with optimal control problems where the system is driven by a stochastic differential equation of the Ito type. We study the relaxed model for which an optimal solution exists. This is an extension of the initial control problem, where admissible controls are measure valued processes. Using Ekeland's variational principle and some stability properties of the corresponding state equation and adjoint processes, we establish necessary conditions for optimality satisfied by an optimal relaxed control. This is the first version of the stochastic maximum principle that covers relaxed controls.     

A Technique for Stochastic Control Problems with Unbounded Control Set

We describe a change of time technique for stochastic control problems with unbounded control set. We demonstrate the technique on a class of maximization problems that do not have optimal controls. Given such a problem, we introduce an extended problem which has the same value function as the original problem and for which there exist optimal controls that are expressible in simple terms. This device yields a natural sequence of suboptimal controls for the original problem. By this we mean a sequence of controls for which the payoff functions approach the value function.     

Error Estimates for the Approximation of a Class of Optimal Control Systems Governed by Linear PDEs

This paper deals with a class of optimal control problems in which the system is governed by a linear partial differential equation and the control is distributed and with constraints. The problem is posed in the framework of the theory of optimal control of systems. A numerical method is proposed to approximate the optimal control. In this method, the state space as well as the convex set of admissible controls are discretized. An abstract error estimate for the optimal control problem is obtained that depends on both the approximation of the state equation and the space of controls. This theoretical result is illustrated by some numerical examples from the literature.     

Solving some optimal path planning problems using an approach based on measure theory

In this paper we solve a collection of optimal path planning problems using a method based on measure theory. First we consider the problem as an optimization problem and then we convert it to an optimal control problem by defining some artificial control functions. Then we perform a metamorphosis in the space of problem. In fact we define an injection between the set of admissible pairs, containing the control vector function and a collision-free path defined on free space and the space of positive Radon measures. By properties of this kind of measures we obtain a linear programming problem that its solution gives rise to constructing approximate optimal trajectory of the original problem. Some numerical examples are proposed.     

Optimality Conditions for Nonregular Optimal Control Problems and Duality

We define a new class of optimal control problems and show that this class is the largest one of control problems where every admissible process that satisfies the Extended Pontryaguin Maximum Principle is an optimal solution of nonregular optimal control problems. In this class of problems the local and global minimum coincide. A dual problem is also proposed, which may be seen as a generalization of the Mond–Weir-type dual problem, and it is shown that the 2-invexity notion is a necessary and su?cient condition to establish weak, strong, and converse duality results between a nonregular optimal control problem and its dual problem. We also present an example to illustrate our results.     

Optimal control of a class of strongly nonlinear parabolic systems

This paper considers some typical optimal control problems for a class of strongly nonlinear parabolic systems. After some necessary preparation, it is shown that the family of admissible trajectories is a weakly closed and weakly sequentially compact subset of a reflexive Banach space and that the set of attainable states at any given time is a weakly compact subset of a Hilbert space. Using these basic results, proofs of existence of optimal controls are presented. A terminal control problem, a special Bolza problem, and a time optimal control problem are solved, and the necessary conditions of optimality for the corresponding control problems are given.     

The existence of optimal control on the basis of Weierstrass’s theorem

The problem of existence of an optimal control is solved on the basis of Weierstrass’s classical theorem if the set of admissible controls belongs to the class of piecewise continuous functions. In the process of describing admissible controls, the main assumption is that the number of switchings (points of discontinuity) is uniformly bounded and not just finite, as in the main problem of optimal control theory. On the one hand, this assumption does not restrict the spectrum of optimal control applications. On the other hand, it fits the Weierstrass’s theorem owing to the convenience in characterizing the sequential compactness. The formulation of Weierstrass’s theorem, which asserts the existence of continuous function extrema on sequentially compact sets, is customary, and its proof complies with the traditional scheme, whereas the concepts (convergent sequences and some others) are adapted to the peculiarity of optimal problems.     

Least Square Control Problems in Nonreflexive Spaces

We consider a control problem in a Banach space with a bounded observer, but an unbounded controller which takes values in the extrapolated Favard class. A least square regulator problem can be formulated if the observer and the admissible controls take values in Hilbert spaces. We prove that for this type of LQR-problem the value function is given by a Riccati operator, and that a bounded state feedback based on the Riccati operator yields the optimal control. April 30, 1999     

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.     

Random relaxed controls and partially observed stochastic systems

We introduce a class of generalized controls called random relaxed controls, and show that under quite general conditions, a partially observed, controlled diffusion will have an optimal random relaxed control whose cost equals the infimum over the costs of all ordinary controls. We also show that the optimal admissible control can be approximated arbitrarily well by very simple, ordinary controls. The proofs are based on a close analysis of the standard parts of nonstandard controls.     

The Global Optimization of Variational Problems with Discontinuous Solutions

We consider variational problems in which the slope of the admissible curves is not necessarily bounded, so that they admit discontinuous solutions. A problem is first reformulated as one consisting of the minimization of an integral in a space of functions satisfying a set of integral equalities; this is then transfered to a nonstandard framework, in which Loeb measures take the place of the functions and a near-minimizer can always be found. This is mapped back to the standard world by means of the standard part map; its image is a minimizer, so that the optimization is global. The minimizer is shown to be the solution of an infinite dimensional linear program and by well-proven approximation procedures a finite dimensional linear program is found by means of which nearly-optimal curves can be constructed for the original problem. A numerical example is given.     

Existence of Optimal Controls in the Absence of Cesari-Type Conditions forSemilinear Elliptic and Parabolic Systems

Some new results on the existence of optimal controls are established for control systems governed by semilinear elliptic or parabolic equations. No Cesari type conditions are assumed. By proving existence theorems and analyzing the Pontryagin maximum principle for optimal relaxed state-control pairs for the corresponding relaxed problems, existence theorems of classical optimal pairs for the original problem are established. To treat the case of a noncompact control set, relaxed controls defined by finitely additive measures are used.     

On the existence of optimal controls for nonlinear infinite dimensional systems

We consider nonlinear systems with a priori feedback. We establish the existence of admissible pairs and then we show that the Lagrange optimal control problem admits an optimal pair. As application we work out in detail two examples of optimal control problems for nonlinear parabolic partial differential equations.     

The probabilistic structure of controlled diffusion processes

This paper surveys those aspects of controlled diffusion processes wherein the control problem is treated as an optimization problem on a set of probability measures on the path space. This includes: (i) existence results for optimal admissible or Markov controls (both in nondegenerate and degenerate cases), (ii) a probabilistic treatment of the dynamic programming principle, (iii) the corresponding results for control under partial observations, (iv) a probabilistic approach to the ergodic control problem. The paper is expository in nature and aims at giving a unified treatment of several old and new results that evolve around certain central ideas.     

An optimal control problem for a parabolic equation in the class of smooth controls

We consider an optimal control problem for a parabolic equation with a differential constraint on the boundary. We study this problem in the class of smooth controls satisfying certain pointwise constraints. Such problems describe mass transfer processes in column-type apparatuses, taking into account the longitudinal mixing. Control functions in these problems represent flows of raw materials or finished products. For the problem under consideration we obtain a necessary optimality condition, propose a method for improving admissible controls, and carry out the numerical experiment.     

On the Control of a Nonlinear Dynamic System in a Time-Optimal Problem with State Constraints

An optimal control problem for a nonlinear dynamic system is studied. The required control must satisfy given constraints and provide the fulfilment of a number of conditions on the current state of the system. For the construction of admissible controls in this problem, we propose an approach based on the ideas of solution of control problems with a guide. The results of numerical simulation are presented.     

Sufficient optimality conditions for extremal controls based on functional increment formulas

We consider the optimal control problem without terminal constraints. With the help of nonstandard functional increment formulas we introduce definitions of strongly extremal controls. Such controls are optimal in linear and quadratic problems. In the general case, the optimality property is provided with an additional concavity condition of Pontryagin’s function with respect to phase variables.     

An optimal control problem for the multidimensional diffusion equation with a generalized control variable

The present paper is concerned with an optimal control problem for then-dimensional diffusion equation with a sequence of Radon measures as generalized control variables. Suppose that a desired final state is not reachable. We enlarge the set of admissible controls and provide a solution to the corresponding moment problem for the diffusion equation, so that the previously chosen desired final state is actually reachable by the action of a generalized control. Then, we minimize an objective function in this extended space, which can be characterized as consisting of infinite sequences of Radon measures which satisfy some constraints. Then, we approximate the action of the optimal sequence by that of a control, and finally develop numerical methods to estimate these nearly optimal controls. Several numerical examples are presented to illustrate these ideas.