Exact solution of Pontryagin's equations of optimal control—Part 1

In the treatment of constrained optimal control processes, it is customary to employ the Pontryagin maximum principle, which requires the solution of a two-point boundary-value problem. Various economic, mechanical, and biological control processes are of this type, including optimization of hemodialysis. Generally speaking, two-point boundary-value problems are more difficult to treat computationally than initial-value or Cauchy problems. In this paper, a Cauchy system is derived for a class of optimal control processes, and it is then shown that the solution of the Cauchy problem satisfies the Pontryagin equations.This research was supported by the National Science Foundation, Grant No. GF-294, and the National Institutes of Health, Grants Nos. GM-16197-01 and GM-16437-01.     

Necessary conditions for optimality for control problems with time delays appearing in both state and control variables

An integral maximum principle is developed for a class of nonlinear systems containing time delays in state and control variables. Its proof is based on the theory of quasiconvex families of functions, originally developed by Gamkrelidze and extended by Banks. This result is used to obtain a pointwise principle of the Pontryagin type.The authors wish to acknowledge Professor J. M. Blatt for suggesting this problem. Further, they also wish to acknowledge the referee of the paper for bringing to their attention the problems discussed in Section 6.     

Solutions to Constrained Optimal Control Problems with Linear Systems

This paper is devoted to present solutions to constrained finite-horizon optimal control problems with linear systems, and the cost functional of the problem is in a general form. According to the Pontryagin’s maximum principle, the extremal control of such problem is a function of the costate trajectory, but an implicit function. We here develop the canonical backward differential flows method and then give the extremal control explicitly with the costate trajectory by canonical backward differential flows. Moreover, there exists an optimal control if and only if there exists a unique extremal control. We give the proof of the existence of the optimal solution for this optimal control problem with Green functions.     

On constrained impulsive control problems

This paper considers constrained impulsive control problems for which the authors propose a new mathematical concept of control required for the impulsive framework. These controls can arise in engineering, in particular, in problems of space navigation. We derive necessary extremum conditions in the form of the Pontryagin maximum principle and also study conditions under which the constraint regularity clarifications become weaker. In the proof of the main result, Ekeland’s variational principle is used.     

Piecewise Continuous Controls in Dieudonné-Rashevsky Type Problems

This paper considers multidimensional control problems governed by a first-order PDE system. It is known that, if the structure of the problem is linear-convex, then the so-called ε-maximum principle, a set of necessary optimality conditions involving a perturbation parameter ε > 0, holds. Assuming that the optimal controls are piecewise continuous, we are able to drop the perturbation parameter within the conditions, proving the Pontryagin maximum principle with piecewise regular multipliers (measures). The Lebesgue and Hahn decompositions of the multipliers lead to refined maximum conditions. Our proof is based on the Baire classification of the admissible controls.     

The normal sub-Riemannian geodesic flow on E(2) generated by a left-invariant metric and a right-invariant distribution

We consider a sub-Riemannian problem on the three-dimensional solvable Lie group E(2) endowed with a left-invariant metric and a right-invariant distribution. The problem is based on construction of a Hamilton structure for the given metric by the Pontryagin maximum principle.     

The Pontryagin maximum principle. Ab ovo usque ad mala

A proof of the Pontryagin maximum principle for a sufficiently general optimal control problem is presented; the proof is based on the implicit function theorem and the theorem on the solvability of a finite-dimensional system of nonlinear equations. The exposition is self-contained: all necessary preliminary facts are proved. These facts are mainly related to the properties of solutions to differential equations with discontinuous right-hand side and are derived as corollaries to the implicit function theorem, which, in turn, is a direct consequence of Newton’s method for solving nonlinear equations.     

Positional strengthenings of the maximum principle and sufficient optimality conditions

We derive nonlocal necessary optimality conditions, which efficiently strengthen the classical Pontryagin maximum principle and its modification obtained by B. Ka?kosz and S. ?ojasiewicz as well as our previous result of a similar kind named the “feedback minimum principle.” The strengthening of the feedback minimum principle (and, hence, of the Pontryagin principle) is owing to the employment of two types of feedback controls “compatible” with a reference trajectory (i.e., producing this trajectory as a Carath´eodory solution). In each of the versions, the strengthened feedback minimum principle states that the optimality of a reference process implies the optimality of its trajectory in a certain family of variational problems generated by cotrajectories of the original and compatible controls. The basic construction of the feedback minimum principle—a perturbation of a solution to the adjoint system—is employed to prove an exact formula for the increment of the cost functional. We use this formula to obtain sufficient conditions for the strong and global minimum of Pontryagin’s extremals. These conditions are much milder than their known analogs, which require the convexity in the state variable of the functional and of the lower Hamiltonian. Our study is focused on a nonlinear smooth Mayer problem with free terminal states. All assertions are illustrated by examples.     

Necessary conditions for functional differential inclusions

We consider an optimal control problem in which the dynamic equation and cost function depend on the recent past of the trajectory. The regularity assumed in the basic data is Lipschitz continuity with respect to the sup norm. It is shown that, for a given optimal solution, an adjoint arc of bounded variation exists that satisfies an associated Hamiltonian inclusion. From this result, known smooth versions of the Pontryagin maximum principle for hereditary problems can be easily derived. Problems with Euclidean endpoint constraints are also considered.     

Maximum principle for optimal control problems with intermediate constraints

We consider optimal control problems with constraints at intermediate points of the trajectory. A natural technique (propagation of phase and control variables) is applied to reduce these problems to a standard optimal control problem of Pontryagin type with equality and inequality constraints at the trajectory endpoints. In this way we derive necessary optimality conditions that generalize the Pontryagin classical maximum principle. The same technique is applied to so-called variable structure problems and to some hybrid problems. The new optimality conditions are compared with the results of other authors and five examples illustrating their application are presented.     

On the solution of inverse problems of dynamics of linearly controlled systems by the negative discrepancy method

The paper is devoted to the substantiation of the negative discrepancy method for solving inverse problems of the dynamics of deterministic control systems that are nonlinear in state variables and linear in control. The problem statements include the known sampling history of trajectories measured inaccurately, with known error estimates. The investigation is based on the Pontryagin maximum principle. The results of simulation of an inverse problem for a macroeconomic model are presented.     

Non-degenerate necessary optimality conditions for the optimal control problem with equality-type state constraints

In this work, an optimal control problem with state constraints of equality type is considered. Novelty of the problem formulation is justified. Under various regularity assumptions imposed on the optimal trajectory, a non-degenerate Pontryagin Maximum Principle is proven. As a consequence of the maximum principle, the Euler–Lagrange and Legendre conditions for a variational problem with equality and inequality state constraints are obtained. As an application, the equation of the geodesic curve for a complex domain is derived. In control theory, the Maximum Principle suggests the global maximum condition, also known as the Weierstrass–Pontryagin maximum condition, due to which the optimal control function, at each instant of time, turns out to be a solution to a global finite-dimensional optimization problem.     

A note on the classical brachistochrone

It is shown that the classical Bernoulli's brachistochrone problem and the brachistochrone problem in a central force field may be solved by the maximum principle of Pontryagin. According to the optimum control theory these problems are singular.     

Optimal control laws for the model of information diffusion in a social group

The article investigates two models of information diffusion in a social group. The dynamics of the process is described by a one-dimensional controlled Riccati differential equation. Our two models differ from the original model of K. V. Izmodenova and A. P. Mikhailov in the choice of the functional being optimized. Two different choices of the optimand functional are considered. The optimal control problems are solved by the Pontryagin maximum principle. It is shown that the optimal control program is a relay function of time with at most one switching point. Conditions on the problem parameters are proposed that are easy to check and guarantee the existence of an optimal-control switching point. The theoretical analysis leads to a one-dimensional convex minimization problem to find the optimal-control switching point. The article also describes an alternative approach to the construction of the optimal solution, which does not resort to the maximum principle and instead utilizes a special representation of the optimand functional and works with reachability sets that are independent of the functional. For the two models considered in this article optimal feedback controls are derived from the programmed optimal controls.     

Invariant imbedding and optimal control theory

A method for directly converting an optimal control problem to a Cauchy problem is presented. No use is made of the Euler equations, Pontryagin's maximum principle, or dynamic programming in the derivation. The initial-value problem, in addition to being desirable from the computational point of view, possesses stable characteristics. The results are directly applicable in the study of guidance and control and are particularly useful for obtaining numerical solutions to control problems.     

Solution of an optimal control problem with phase constraints by the double variation method

An optimal control problem with an integral quality index specified in a finite time interval is formulated for a model of economic growth that leads to emission of greenhouse gases. The controlled system is linear with respect to control. The problem contains phase constraints that abandon emission of greenhouse gases above some predefined time-dependent limit. As is known, optimal control problems with phase constraints fall beyond the sphere of efficient application of the Pontryagin maximum principle because, for such problems, this principle is formulated in a complicated form difficult for analytic treatment in particular situations. In this study, the analytic structure of the optimal control and phase trajectories is constructed using the double variation method.     

Optimum control laws in a non-autonomous dynamic model of a gas field

A model of gas field development described as a nonlinear optimum control problem with an infinite planning horizon is considered. The Pontryagin maximum principle is used to solve it. The theorem on sufficient optimumity conditions in terms of constructions of the Pontryagin maximum principles is used to substantiate the optimumity of the extremal solution. A procedure for constructing the optimum solution by dynamic programming is described and is of some methodological interest. The obtained optimum solution is used to construct the Bellman function. Reference is made to a work containing an economic interpretation of the problem.     

Existence of Optimal Controls for Semilinear Parabolic Equations without Cesari-Type Conditions

Abstract. Optimal control problems governed by semilinear parabolic partial differential equations are considered. No Cesari-type conditions are assumed. By proving the existence theorem and the Pontryagin maximum principle of optimal ``state-control" pairs for the corresponding relaxed problems, an existence theorem of optimal pairs for the original problem is established.     

Optimality conditions for reflecting boundary control problems

We consider a control problem with reflecting boundary and obtain necessary optimality conditions in the form of the maximum Pontryagin principle. To derive these results we transform the constrained problem in an unconstrained one or we use penalization techniques of Morreau-Yosida type to approach the original problem by a sequence of optimal control problems with Lipschitz dynamics. Then nonsmooth analysis theory is used to study the convergence of the penalization in order to obtain optimality conditions.     

A Solution to Fuller’s Problem Using Constructions of Pontryagin’s Maximum Principle

The classical two-dimensional Fuller problem is considered. The boundary value problem of Pontryagin’s maximum principle is considered. Based on the central symmetry of solutions to the boundary value problem, the Pontryagin maximum principle as a necessary condition of optimality, and the hypothesis of the form of the switching line, a solution to the boundary value problem is constructed and its optimality is substantiated. Invariant group analysis is in this case not used. The results are of considerable methodological interest.