Abstract
In this chapter we discuss turnpike properties and optimality criterions over infinite horizon for a class of convex dynamic optimization problems.
The study of optimal control problems and dynamic games defined on infinite intervals and on sufficiently large intervals has been a rapidly growing area of research [1, 4, 5, 9–16, 21, 23, 24, 26, 27, 29–33, 35, 36, 40, 41, 43, 44, 49–52, 54–56, 61] which has various applications in engineering [2, 19, 38, 70], in models of economic growth [3, 6, 17–20, 25, 28, 34, 37, 42, 48, 53, 59, 60, 62, 70, 79, 85, 94, 95], in infinite discrete models of solid-state physics related to dislocations in one-dimensional crystals [8, 57, 63], and in the theory of thermodynamical equilibrium for materials [22, 39, 45–47].
In this chapter we discuss turnpike properties and optimality criterions over infinite horizon for a class of convex dynamic optimization problems.
1.1 Convex Discrete-Time Problems
Let R n be the n-dimensional Euclidean space with the inner product 〈⋅ , ⋅ 〉 which induces the norm
Let K be a nonempty convex subset of R n. A function f: K → R 1 is called convex (strictly convex respectively) if for all x, y ∈ K such that x ≠ y and all α ∈ (0, 1),
respectively) [58].
Let v: R n × R n → R 1 be a bounded from below function. We consider the minimization problem
where T is a natural number and the points y, z ∈ R n.
The interest in discrete-time optimal problems of type (P0 ) stems from the study of various optimization problems which can be reduced to it, e.g., continuous-time control systems which are represented by ordinary differential equations whose cost integrand contains a discounting factor [37], tracking problems in engineering [2, 38], the study of Frenkel–Kontorova model [8, 63], and the analysis of a long slender bar of a polymeric material under tension in [22, 39, 45–47]. Optimization problems of the type ( P 0 ) were considered in [64–66, 68, 69, 74].
In this section we suppose that the function v: R n × R n → R 1 is strictly convex and differentiable and satisfies the growth condition
We intend to analyze the structure of solutions of the problem ( P 0 ) when the points y, z and the real number T vary and T is sufficiently large. More precisely, we are interested to study a turnpike property of solutions of ( P 0 ) which is independent of the length of the interval T, for all sufficiently large intervals. To have this property means, roughly speaking, that solutions of the optimal control problems are determined mainly by the objective function v, and are essentially independent of T, y, and z. Turnpike properties are well known in mathematical economics (see, for example, [42, 48, 59, 60] and the references mentioned there). Many turnpike results are collected in [70, 81, 84, 93, 95].
In order to meet our goal we consider the auxiliary optimization problem
In view of the strict convexity of v and (1.1), problem (P1 ) has a unique solution \(\bar{x}\). Let
where l 1, l 2 ∈ R n. Since \(\bar{x}\) is a solution of (P1 ) it follows from (1.2) that for every h ∈ R n, we have
Therefore
l 2 = −l 1 and
For every (y, z) ∈ R n × R n define
It is easy to see that the function L: R n × R n → R 1 is differentiable and strictly convex. By (1.1) and (1.4), we have
Since the functions v and L are both strictly convex [58] it follows from (1.4) that
and
We show that the function L: R n × R n → R 1 possesses the following property:
(C) If a sequence {(y i , z i )} i = 1 ∞ ⊂ R n × R n satisfies the equality
then
Assume that a sequence {(y i , z i )} i = 1 ∞ ⊂ R n × R n and that lim i → ∞ L(y i , z i ) = 0. It follows from (1.5) that the sequence {(y i , z i )} i = 1 ∞ is bounded. Let (y, z) be its limit point. Then it is not difficult to see that the equality
is true and in view of (1.7), we have \((y,z) = (\bar{x},\bar{x}).\) This implies that \((\bar{x},\bar{x}) =\lim _{i\rightarrow \infty }(y_{i},z_{i})\).
Therefore property (C) holds.
Consider an auxiliary minimization problem
where T ≥ 1 is an integer and y, z ∈ R n.
By (1.4), for every natural number T and every sequence {x i } i = 0 T ⊂ R n,
It follows from (1.8) that problems ( P 0 ) and (P2 ) are equivalent. More precisely, {x i } i = 0 T ⊂ R n is a solution of problem ( P 0 ) if and only if it is a solution of problem ( P 2 ).
Let T ≥ 1 be an integer and Δ be a positive number. A sequence {x i } i = 0 T ⊂ R n is called (Δ)-optimal if for every sequence {x i ′} i = 0 T ⊂ R n which satisfies x i = x i ′, i = 0, T the inequality
is valid. It is clear that if a sequence {x i } i = 0 T ⊂ R n is (0)-optimal, then it is a solution of problems (P0 ) and (P2 ) with z = x 0 and y = x T .
We prove the following existence result.
Proposition 1.1.
Let T > 1 be an integer and y,z ∈ R n . Then problem (P0 ) possesses a solution.
Proof.
In order to prove the proposition it is sufficient to show that problem ( P 2 ) possesses a solution. Consider a sequence {x i ′} i = 0 T ⊂ R n which satisfies x 0′ = z, x T ′ = y. Put
and
Evidently,
We may assume without loss of generality that
There exists a sequence {x i (k)} i = 0 T ⊂ R n, k = 1, 2, … such that for every integer k ≥ 1, we have
and
By (1.10), (1.11), and (1.12), we may assume that
It follows from (1.13) and (1.5) that there exists a positive number M 3 such that
By (1.14), extracting subsequences, using diagonalization process and re-indexing, if necessary, we may assume without loss of generality that for every integer i ∈ { 0, …, T} there exists
In view of (1.15) and (1.11), we have
Relations (1.15) and (1.12) imply that
Combined with (1.16) and (1.9) this implies that the finite sequence \(\{\widehat{x}_{i}\}_{i=0}^{T}\) is a solution of problem (P2 ). This completes the proof of Proposition 1.1.
Denote by Card(A) the cardinality of a set A.
The following result establishes a turnpike property for approximate solutions of the problem ( P 0 ).
Proposition 1.2.
Let M 1 ,M 2 ,ε > 0. Then there exists an integer k 0 ≥ 1 such that for every natural number T > 1 and every (M 1 )-optimal sequence {x i } i=0 T ⊂ R n which satisfies
the inequality
is valid.
Proof.
Condition (C) implies that there exists a positive number δ such that for every point (y, z) ∈ R n × R n which satisfies
the inequality
is true. Set
and fix a natural number
Assume that an integer T > 1 and that an (M 1)-optimal sequence {x i } i = 0 T ⊂ R n satisfies (1.17). Define
Since the sequence {x i } i = 0 T is (M 1)-optimal it follows from (1.22) that
Combined with (1.7), (1.8), and (1.22) this implies that
Together with (1.17) and (1.20) this implies that
In view of the choice of δ (see (1.18) and (1.19)), (1.21) and the inequality above, we have
Proposition 1.2 is proved.
Proposition 1.2 implies the following turnpike result for exact solutions of the problem ( P 0 ).
Proposition 1.3.
Let M,ε > 0. Then there exists an integer k 0 ≥ 1 such that for every natural number T > 1, every pair of points y,z ∈ R n which satisfies |y|, |z|≤ M and every optimal solution {x i } i=0 T ⊂ R n of problem (P0 ) the inequality
is valid.
Now it is clear that the optimal solution {x i } i = 0 T of problem ( P 0 ) spends most of the time in an ε-neighborhood of the point \(\bar{x}\). In view of Proposition 1.3, the number of all integers i ∈ { 0, …, T − 1} for which that x i does not belong to this ε-neighborhood, does not exceed the constant k 0 which depends only on M, ε and does not depend on T. Following the tradition, the point \(\bar{x}\) is called the turnpike. Moreover we can show that the set
is contained in the union of two intervals [0, k 1] ∪ [T − k 1, T], where k 1 is a constant depending only on M, ε.
We also study the infinite horizon problem associated with problem ( P 0 ).
By (1.1) there is M ∗ > 0 such that
We suppose that the sum over empty set is zero.
Proposition 1.4.
Let M 0 be a positive number. Then there exists a positive number M 1 such that for every natural number T and every finite sequence {x i } i=0 T ⊂ R n which satisfies |x 0 |≤ M 0 the inequality
holds.
Proof.
Set
Assume that T is a natural number and a that a finite sequence {x i } i = 0 T ⊂ R n satisfies
If | x i | > M ∗, i = 1, …, T, then in view of (1.23), we have
and inequality (1.24) is true. Therefore we may assume that there exists an integer q ≥ 1 for which
We may assume without loss of generality that
In view of (1.23) and (1.27), we have
By the equation above, (1.8), (1.6), (1.25), (1.26) and the choice of M 1, we have
This completes the proof of Proposition 1.4.
Choose a positive number \(\tilde{M}\) for which
Proposition 1.5.
Let {x i } i=0 ∞ ⊂ R n . Then either the sequence
is bounded or
Proof.
In view of (1.23), if for all sufficiently large natural numbers i, we have | x i | ≥ M ∗, then equality (1.29) is valid. Therefore we may assume without loss of generality that there exists a strictly increasing sequence of natural numbers {t k } k = 1 ∞ such that
Proposition 1.4 implies that the sequence \(\{\sum _{i=0}^{T-1}(v(x_{i},x_{i+1}) - v(\bar{x},\bar{x}))\}_{T=1}^{\infty }\) is bounded from below.
Assume that this sequence is not bounded from above. In order to complete the proof it is sufficient to show that equality (1.29) is valid.
Let Q > 0 be given. Then there exists an integer T 0 ≥ 1 for which
Fix an integer k ≥ 1 such that
Let an integer
In view of (1.30), (1.32), and (1.33), there exists an integer S such that
By (1.31), (1.34), (1.36), (1.23), (1.35), (1.28), and Proposition 1.4, we have
Thus for any integer T > t k ,
Since Q is any positive number (1.29) is valid. This completes the proof of Proposition 1.5.
A sequence {x i } i = 0 ∞ ⊂ R n is called good [28, 70, 81] if the sequence \(\{\sum _{i=0}^{T-1}(v(x_{i},x_{i+1}) - v(\bar{x},\bar{x}))\}_{T=1}^{\infty }\) is bounded.
Proposition 1.6.
1. A sequence {x i } i=0 ∞ ⊂ R n is good if and only if
2. If a sequence {x i } i=0 ∞ ⊂ R n is good, then it converges to \(\bar{x}\) .
Proof.
Assume that a sequence {x i } i = 0 ∞ ⊂ R n is good. Then there exists a positive number M 0 such that
In view of (1.37) and (1.23), there exists a strictly increasing sequence of natural numbers {t k } k = 1 ∞ such that
Let k ≥ 1 be an integer. It follows from (1.8), (1.37), and (1.38) that
and
Since the inequality above is valid for all natural numbers k we conclude that
Property (C) implies that the sequence {x i } i = 0 ∞ converges to \(\bar{x}\) and Assertion 2 is proved.
Assume that
In view of (1.5), there exists a positive number M 2 such that
By (1.8), (1.39), and (1.40), for all integers T ≥ 1, we have
Combined with Proposition 1.5 this implies that the sequence {x i } i = 0 ∞ is good and completes the proof of Proposition 1.6.
Proposition 1.7.
Let x ∈ R n . Then there exists a sequence {x i } i=0 ∞ ⊂ R n such that x 0 = x and for each sequence {y i } i=0 ∞ ⊂ R n satisfying y 0 = x the inequality
holds.
Proof.
Set
It is clear that M 0 is well defined and nonnegative. There exists a sequence {x i (k)} i = 0 ∞ ⊂ R n, k = 1, 2, … such that
In view of (1.43) and (1.5), there exists a positive number M 1 such that
By (1.44) using diagonalization process, extracting subsequences and re-indexing we may assume without loss of generality that for every nonnegative integer i there exists
It follows from (1.42) and (1.45) that
In view of (1.6), (1.43), and (1.45), for every integer T ≥ 1, we have
Since T is an arbitrary natural number we conclude that
Combined with (1.41) and (1.46) this implies that
Proposition 1.7 is proved.
In our study we use the following optimality criterion introduced in the economic literature [6, 28, 62] and used in the optimal control [19, 70, 81, 84].
A sequence {x i } i = 0 ∞ ⊂ R n is called overtaking optimal if
for any sequence {y i } i = 0 ∞ ⊂ R n satisfying y 0 = x 0.
Proposition 1.8.
Let {x i } i=0 ∞ ⊂ R n . Then the following assertions are equivalent:
1. the sequence {x i } i=0 ∞ is overtaking optimal;
2.
for every sequence {y i } i=0 ∞ ⊂ R n which satisfies y 0 = x 0 .
Proof.
Assume that the sequence {x i } i = 0 ∞ is overtaking optimal. Evidently, it is good. Proposition 1.6 implies that
Let a sequence {y i } i = 0 ∞ ⊂ R n satisfy
We claim that
We may assume that
Property (C) implies that
Since the sequence {x i } i = 0 ∞ is overtaking optimal it follows from (1.47), (1.8), and (1.48) that
Thus assertion 2 holds.
Assume that assertion 2 holds. We claim that the sequence {x i } i = 0 ∞ is overtaking optimal. It is clear that
Proposition 1.6 implies that the sequence {x i } i = 0 ∞ is good and that
Assume that a sequence {y i } i = 0 ∞ ⊂ R n satisfies
We show that
We may assume without loss of generality that the sequence {y i } i = 0 ∞ is good. In view of Proposition 1.6, we have
In view of (1.8), (1.49), (1.50), (1.51), and assertion 2, we have
Thus assertion 1 holds and Proposition 1.8 is proved.
Propositions 1.7 and 1.8 imply the following existence result.
Proposition 1.9.
For every point x ∈ R n there exists an overtaking optimal sequence {x i } i=0 ∞ ⊂ R n such that x 0 = x.
1.2 The Turnpike Phenomenon
In the previous section we proved the turnpike result and the existence of overtaking optimal solutions for rather simple class of discrete-time problems. The problems of this class are unconstrained and their objective functions are convex and differentiable. In this book our goal is to study the structure of approximate solutions over large intervals for a class of discrete-time constrained optimal control problems without convexity (concavity) assumptions. In particular, in Chaps. 2–4 we study the structure of approximate solutions of an autonomous discrete-time control system with a compact metric space of states X. This control system is described by a bounded upper semicontinuous function v: X × X → R 1 which determines an optimality criterion and by a nonempty closed set Ω ⊂ X × X which determines a class of admissible trajectories (programs). We study the problems
and
where T ≥ 1 is an integer and the points y, z ∈ X.
In the classical turnpike theory the objective function v possesses the turnpike property (TP) if there exists a point \(\bar{x} \in X\) (a turnpike) such that the following condition holds:
For each positive number ε there exists an integer L ≥ 1 such that for each integer T ≥ 2L and each solution {x i } i = 0 T ⊂ X of the problem (P1) the inequality \(\rho (x_{i},\bar{x}) \leq \epsilon\) is true for all i = L, …, T − L.
It should be mentioned that the constant L depends neither on T nor on y, z.
The turnpike phenomenon has the following interpretation. If one wishes to reach a point A from a point B by a car in an optimal way, then one should turn to a turnpike, spend most of time on it, and then leave the turnpike to reach the required point.
In the classical turnpike theory [28, 48, 59, 62] the space X is a compact convex subset of a finite-dimensional Euclidean space, the set Ω is convex, and the function v is strictly concave. Under these assumptions the turnpike property can be established and the turnpike \(\bar{x}\) is a unique solution of the maximization problem v(x, x) → max, (x, x) ∈ Ω. In this situation it is shown that for each program {x t } t = 0 ∞ either the sequence \(\{\sum _{t=0}^{T-1}v(x_{t},x_{t+1}) - Tv(\bar{x},\bar{x})\}_{T=1}^{\infty }\) is bounded (in this case the program {x t } t = 0 ∞ is called (v)-good) or it diverges to −∞. Moreover, it is also established that any (v)-good program converges to the turnpike \(\bar{x}\). In the sequel this property is called as the asymptotic turnpike property.
Recently it was shown that the turnpike property is a general phenomenon which holds for large classes of variational and optimal control problems without convexity assumptions. (See, for example, [70] and the references mentioned therein). For these classes of problems a turnpike is not necessarily a singleton but may instead be a nonstationary trajectory (in the discrete time nonautonomous case) or an absolutely continuous function on the interval [0, ∞) (in the continuous time nonautonomous case) or a compact subset of the space X (in the autonomous case). Note that all of these results were obtained for unconstrained problems. In particular, the turnpike results for the problems of the type (P1) were obtained in the case Ω = X × X.
For classes of problems considered in [70], using the Baire category approach, it was shown that the turnpike property holds for a generic (typical) problem. In this book we are interested in individual (non-generic) results describing the structure of approximate solutions. We study the problems (P1)–(P3) with the constraint {(x i , x i+1)} i = 0 T−1 ⊂ Ω where Ω is an arbitrary nonempty closed subset of X × X. Clearly, these constrained problems are more difficult and less understood than their unconstrained prototypes in the previous section and in [64–66, 68]. They are also more realistic from the point of view of mathematical economics. As we have mentioned before in general a turnpike is not necessarily a singleton. Nevertheless problems of the type (P1)–(P3) for which the turnpike is a singleton are of great importance because of the following reasons: there are many models of economic growth for which a turnpike is a singleton; if a turnpike is a singleton, then approximate solutions have very simple structure and this is very important for applications; if a turnpike is a singleton, then it can be easily calculated as a solution of the problem v(x, x) → max, (x, x) ∈ Ω.
The turnpike property is very important for applications. Suppose that our objective function v has the turnpike property and we know a finite number of “approximate” solutions of the problem (P1). Then we know the turnpike \(\bar{x}\), or at least its approximation, and the constant L (see the definition of (TP)) which is an estimate for the time period required to reach the turnpike. This information can be useful if we need to find an “approximate” solution of the problem (P1) with a new time interval [m 1, m 2] and the new values z, y ∈ X at the end points m 1 and m 2. Namely instead of solving this new problem on the “large” interval [m 1, m 2] we can find an “approximate” solution of the problem (P1) on the “small” interval [m 1, m 1 + L] with the values \(z,\bar{x}\) at the end points and an “approximate” solution of the problem (P1) on the “small” interval [m 2 − L, m 2] with the values \(\bar{x},y\) at the end points. Then the concatenation of the first solution, the constant sequence \(x_{i} =\bar{ x}\), i = m 1 + L, …, m 2 − L and the second solution is an “approximate” solution of the problem (P1) on the interval [m 1, m 2] with the values z, y at the end points. Sometimes as an “approximate” solution of the problem (P1) we can choose any admissible sequence \(\{x_{i}\}_{i=m_{1}}^{m_{2}}\) satisfying
1.3 Nonconcave (Nonconvex) Problems
In Chap. 2 we study the structure of approximate solutions of discrete-time optimal control problems introduced in Sect. 1.2.
Let (X, ρ) be a compact metric space, Ω be a nonempty closed subset of X × X, and let v: X × X → R 1 be a bounded upper semicontinuous function.
A sequence {x t } t = 0 ∞ ⊂ X is called an (Ω)-program (or just a program if the set Ω is understood) if (x t , x t+1) ∈ Ω for all nonnegative integers t. A sequence {x t } t = 0 T where T ≥ 1 is an integer is called an (Ω)-program (or just a program if the set Ω is understood) if (x t , x t+1) ∈ Ω for all integers t ∈ [0, T − 1].
We consider the problems
and
where T ≥ 1 is an integer and the points y, z ∈ X.
We suppose that there exist a point \(\bar{x} \in X\) and a positive number \(\bar{c}\) such that the following assumptions hold:
(i) \((\bar{x},\bar{x})\) is an interior point of Ω;
(ii) \(\sum _{t=0}^{T-1}v(x_{t},x_{t+1}) \leq Tv(\bar{x},\bar{x}) +\bar{ c}\) for any natural number T and any program {x t } t = 0 T.
The property (ii) implies that for each program {x t } t = 0 ∞ either the sequence
is bounded or \(\lim _{T\rightarrow \infty }[\sum _{t=0}^{T-1}v(x_{t},x_{t+1}) - Tv(\bar{x},\bar{x})] = -\infty.\)
A program {x t } t = 0 ∞ is called (v)-good if the sequence
is bounded.
We suppose that the following assumption holds.
(iii) (the asymptotic turnpike property) For any (v)-good program {x t } t = 0 ∞, \(\lim _{t\rightarrow \infty }\rho (x_{t},\bar{x}) = 0\).
Note that the properties (i)–(iii) hold for models of economic dynamics considered in the classical turnpike theory.
For each positive number M denote by X M the set of all points x ∈ X for which there exists a program {x t } t = 0 ∞ such that x 0 = x and that for all natural numbers T the following inequality holds:
It is not difficult to see that ∪{ X M : M ∈ (0, ∞)} is the set of all points x ∈ X for which there exists a (v)-good program {x t } t = 0 ∞ satisfying x 0 = x.
Let T ≥ 1 be an integer and Δ ≥ 0. A program {x i } i = 0 T ⊂ X is called (Δ)-optimal if for any program {x i ′} i = 0 T satisfying x 0 = x 0′, the inequality
holds.
The turnpike theory for problems (P1) and (P2) is presented in [84] which summaries our research [71–73, 75–77].
In particular, in Chap. 2 of [84] we prove the following turnpike result for approximate solutions of our second optimization problem stated above.
Theorem 1.10.
Let ε,M be positive numbers. Then there exist a natural number L and a positive number δ such that for each integer T > 2L and each (δ)-optimal program {x t } t=0 T which satisfies x 0 ∈ X M there exist nonnegative integers τ 1 ,τ 2 ≤ L such that \(\rho (x_{t},\bar{x}) \leq \epsilon\) for all t = τ 1 ,…,T −τ 2 and if \(\rho (x_{0},\bar{x}) \leq \delta\) , then τ 1 = 0.
An analogous turnpike result for approximate solutions of our first optimization problem is also proved in Chap. 2 of [84].
A program {x t } t = 0 ∞ is called (v)-overtaking optimal if for each program {y t } t = 0 ∞ satisfying y 0 = x 0 the inequality limsup T → ∞ ∑ t = 0 T−1[v(y t , y t+1) − v(x t , x t+1)] ≤ 0 holds.
In Chap. 2 of [84] we prove the following result which establishes the existence of an overtaking optimal program.
Theorem 1.11.
Assume that x ∈ X and that there exists a (v)-good program {x t } t=0 ∞ such that x 0 = x. Then there exists a (v)-overtaking optimal program {x t ∗ } t=0 ∞ such that x 0 ∗ = x.
In Chap. 2 of [84] for problems which satisfy concavity assumption common in the literature we study the structure of approximate solutions in the regions containing end points and obtain a full description of the structure of approximate solutions. More precisely, we study the structure of approximate solutions of our second optimization problem stated above in the regions [0, L] and [T − L, T] (see the definition of the turnpike property). We show that if {x i } i = 0 T ⊂ X is an approximate solution of our problem, then for all integers t = 0, …, L the state x t is closed enough to z t where {z t } t = 0 ∞ ⊂ X is a unique solution of a certain infinite horizon optimal control problem satisfying z 0 = z. We also show that if {x i } i = 0 T ⊂ X is an approximate solution of our second optimization problem, then for all integers t = 0, …, L the state x T−t is closed enough to Λ t where {Λ t } t = 0 ∞ ⊂ X is a unique solution of a certain infinite horizon optimal control problem which does not depend on z. These results are established when the set X is a convex subset of the Euclidean space R n, the set Ω is convex, and the function v is strictly concave. In this case we obtain the full description of the structure of approximate solutions of our second optimization problem. Note that the structure of approximate solutions in the region [0, L] depends on z while their structure in the region [T − L, T] does not depend on z. Actually it depends only on v and Ω.
In Chap. 2 of the present book we prove the generalizations of the results of Chap. 2 of [84] on the structure of approximate solutions in the regions containing end points. These generalizations are established for problems (P1)–(P3) without concavity assumptions. The results of this chapter were obtained in [83, 86, 90].
In Chap. 3 we consider optimal control systems which are discrete-time analogs of Bolza problems in the calculus of variations. They are described by a pair of objective functions which determines an optimality criterion. We consider two classes of Bolza problems and obtain for each of them the full description of approximate solutions of these problems on large intervals. This description shows that on large intervals the approximate solutions are determined mainly by our optimality criterion and are essentially independent of the choice of time intervals and data. The results of Chap. 4 were obtained in [91, 92].
In Chap. 4 we continue to study the discrete-time analogs of Bolza problems in the calculus of variations. We show that the turnpike phenomenon and the structure of solutions on finite intervals in the regions close to the endpoints are stable under small perturbations of the objective functions and the set Ω. The results of this chapter are new.
1.4 Examples
Example 1.12.
Let (X, ρ) be a compact metric space, Ω be a nonempty closed subset of X × X, \(\bar{x} \in X\), \((\bar{x},\bar{x})\) be an interior point of Ω, π: X → R 1 be a continuous function, α be a real number, and L: X × X → [0, ∞) be a continuous function such that for each (x, y) ∈ X × X the equality L(x, y) = 0 holds if and only if \((x,y) = (\bar{x},\bar{x}).\) Set
for all x, y ∈ X. It is not difficult to see that assumptions (i), (ii), and (iii) hold.
Example 1.13.
Let X be a compact convex subset of the Euclidean space R n with the norm | ⋅ | induced by the scalar product 〈⋅ , ⋅ 〉, let ρ(x, y) = | x − y | , x, y ∈ R n, Ω be a nonempty closed subset of X × X, a point \(\bar{x} \in X\), \((\bar{x},\bar{x})\) be an interior point of Ω, and let v: X × X → R 1 be a strictly concave continuous function such that
We assume that there exists a positive constant \(\bar{r}\) such that
It is a well-known fact of convex analysis [58] that there exists a point l ∈ R n such that
for any point (x, y) ∈ X × X. Set
for all (x, y) ∈ X × X. It is not difficult to see that this example is a particular case of Example 1.12. Therefore assumptions (i), (ii), and (iii) hold.
Example 1.14.
Let X = [0, 1], Ω = { (x, y) ∈ [0, 1] × [0, 1]: y ≤ x 1∕2}, v(x, y) = x 1∕2 − y 2, x, y ∈ X. It is not difficult to see that the set Ω is convex, the function v is strictly concave, the optimization problem v(z, z) → max, z ∈ X, and (z, z) ∈ Ω has a unique solution 16−1∕3 and (16−1∕3, 16−1∕3) is an interior point of Ω. Therefore this example is a particular case of Example 1.13 and assumptions (i), (ii), and (iii) hold.
Example 1.15.
Let X = [0, 1], Ω = { (x, y) ∈ [0, 1] × [0, 1]: y ≤ x 1∕2}, v(x, y) = x 1∕2 − y, x, y ∈ X. It is not difficult to see that the set Ω is convex, the function v is concave but not strictly concave, the optimization problem v(z, z) → max, z ∈ X, and (z, z) ∈ Ω has a unique solution 4−1 and (4−1, 4−1) is an interior point of Ω. Since the function v is concave for all x, y ∈ X,
and
is equal zero if and only if x = 4−1. Now it is not difficult to see that assumptions (i), (ii), and (iii) hold.
Example 1.16.
Consider the sets X, Ω, and the function v defined in Example 1.15 and set u(x, y) = x 1∕2 − x 2 − y + y 2, x, y ∈ X. The function u is strictly convex with respect to the variable y. Nevertheless assumptions (i), (ii), and (iii) hold for the function u because for any integer T and any program {x t } t = 0 T,
1.5 Two-Player Zero-Sum Games
In Chaps. 5–8 we prove turnpike results for classes of dynamic discrete-time two-player zero-sum games. These results describe the structure of approximate solutions, for all sufficiently large intervals. We also study the structure of approximate solutions on large intervals in the regions close to the endpoints of the intervals and examine the existence of a pair of overtaking equilibria strategies over an infinite horizon.
Let \(X \subset R^{m_{1}}\) and \(Y \subset R^{m_{2}}\) be nonempty convex compact sets. Denote by \(\mathcal{M}\) the set of all continuous functions f: X × X × Y × Y → R 1 such that:
for each point (y 1, y 2) ∈ Y × Y the function (x 1, x 2) → f(x 1, x 2, y 1, y 2), (x 1, x 2) ∈ X × X is convex;
for each point (x 1, x 2) ∈ X × X the function (y 1, y 2) → f(x 1, x 2, y 1, y 2), (y 1, y 2) ∈ Y × Y is concave.
The set \(\mathcal{M}\) is equipped with a metric \(\rho: \mathcal{M}\times \mathcal{M}\rightarrow R^{1}\) defined by
It is clearly that \((\mathcal{M},\rho )\) is a complete metric space.
Given \(f \in \mathcal{M}\) and a natural number n we consider a discrete-time two-player zero-sum game over the interval [0, n]. For this game {{x i } i = 0 n: x i ∈ X, i = 0, … n} is the set of strategies for the first player, {{y i } i = 0 n: y i ∈ Y, i = 0, … n} is the set of strategies for the second player, and the objective function for the first player associated with the strategies {x i } i = 0 n, {y i } i = 0 n is given by ∑ i = 0 n−1 f(x i , x i+1, y i , y i+1).
Let \(f \in \mathcal{M}\), n be a natural number and let M ∈ [0, ∞). A pair of sequences \(\{\bar{x}_{i}\}_{i=0}^{n} \subset X,\;\{\bar{y}_{i}\}_{i=0}^{n} \subset Y\) is called (f, M)-good if the following properties hold:
(i) for each sequence {x i } i = 0 n ⊂ X satisfying \(x_{0} =\bar{ x}_{0}\), \(x_{n} =\bar{ x}_{n}\) the inequality
holds;
(ii) for each sequence {y i } i = 0 n ⊂ Y satisfying \(y_{0} =\bar{ y}_{0}\), \(y_{n} =\bar{ y}_{n}\) the inequality
holds.
If a pair of sequences {x i } i = 0 n ⊂ X, {y i } i = 0 n ⊂ Y is (f, 0)-good then it is called (f)-optimal.
Let \(f \in \mathcal{M}\). We say that the function f possesses the turnpike property if there exists a unique pair (x f , y f ) ∈ X × Y for which the following assertion holds:
For each positive number ε there exist an integer n 0 ≥ 2 and a positive number δ such that for each integer n ≥ 2n 0 and each (f, δ)-good pair of sequences {x i } i = 0 n ⊂ X, {y i } i = 0 n ⊂ Y the inequalities | x i − x f | , | y i − y f | ≤ ε holds for all integers i ∈ [n 0, n − n 0].
In [67] we showed that the turnpike property holds for a generic \(f \in \mathcal{M}\). Namely, in [67] we proved the existence of a set \(\mathcal{F}\subset \mathcal{M}\) which is a countable intersection of open everywhere dense sets in \(\mathcal{M}\) such that each \(f \in \mathcal{F}\) has the turnpike property. Thus for most functions \(f \in \mathcal{M}\) the turnpike property holds. Nevertheless it is very important to have conditions on \(f \in \mathcal{M}\) which imply the turnpike property. These conditions are discussed in [78] and in Chap. 5 of [84].
In Chaps. 5–8 we study the structure of solutions of more general and complicated classes of dynamic discrete-time two-player zero-sum games. In Chap. 5 we study a class of unconstrained dynamic discrete-time two-player zero-sum games without using standard convexity–concavity assumptions and prove two turnpike results of [82]. Chapter 6 contains the study of the existence and turnpike properties of approximate solutions for a class of dynamic constrained discrete-time two-player zero-sum games without convexity–concavity assumptions. Its results were obtained in [88]. In Chap. 7 we study turnpike properties of approximate solutions for a class of dynamic discrete-time two-player zero-sum nonautonomous games with convexity–concavity assumptions which were established in [87]. In Chap. 8 we analyze the existence and turnpike properties of approximate solutions for a class of dynamic constrained discrete-time two-player zero-sum games which satisfy convexity–concavity assumptions. We describe the structure of approximate solutions which is independent of the length of the interval, for all sufficiently large intervals and show that approximate solutions are determined mainly by the objective function, and are essentially independent of the choice of intervals and endpoint conditions. We show that the turnpike phenomenon is stable under small perturbations of objective functions and analyze the structure of approximate solutions in regions closed to the endpoints of domains.
References
Alvarez O, Bardi M (2007) Ergodic problems in differential games. In: Advances in dynamic game theory. Birkhäuser, Boston, pp 131–152
Anderson BDO, Moore JB (1971) Linear optimal control. Prentice-Hall, Englewood Cliffs
Arkin VI, Evstigneev IV (1987) Stochastic models of control and economic dynamics. Academic, London
Aseev SM, Kryazhimskiy AV (2004) The Pontryagin maximum principle and transversality conditions for a class of optimal control problems with infinite time horizons. SIAM J Control Optim 43:1094–1119
Aseev SM, Veliov VM (2012) Maximum principle for infinite-horizon optimal control problems with dominating discount. Dyn Continuous Discrete Impuls Syst Ser B 19:43–63
Atsumi H (1965) Neoclassical growth and the efficient program of capital accumulation. Rev Econ Stud 32:127–136
Aubry S, Le Daeron PY (1983) The discrete Frenkel-Kontorova model and its extensions I. Physica D 8:381–422
Bachir M, Blot J (2015) Infinite dimensional infinite-horizon Pontryagin principles for discrete-time problems. Set-Valued Var Anal 23:43–54
Bright I (2012) A reduction of topological infinite-horizon optimization to periodic optimization in a class of compact 2-manifolds. J Math Anal Appl 394:84–101
Brock WA (1970) On existence of weakly maximal programmes in a multi-sector economy. Rev Econ Stud 37:275–280
Carlson DA, Haurie A, Leizarowitz A (1991) Infinite horizon optimal control. Springer, Berlin
Carlson DA, Jabrane A, Haurie A (1987) Existence of overtaking solutions to infinite dimensional control problems on unbounded time intervals. SIAM J Control Optim 25:1517–1541
Cartigny P, Michel P (2003) On a sufficient transversality condition for infinite horizon optimal control problems. Autom J IFAC 39:1007–1010
Coleman BD, Marcus M, Mizel VJ (1992) On the thermodynamics of periodic phases. Arch Ration Mech Anal 117:321–347
Damm T, Grune L, Stieler M, Worthmann K (2014) An exponential turnpike theorem for dissipative discrete time optimal control problems. SIAM J Control Optim 52:1935–1957
De Oliveira VA, Silva GN (2009) Optimality conditions for infinite horizon control problems with state constraints. Nonlinear Anal 71:1788–1795
Evstigneev IV, Flam SD (1998) Rapid growth paths in multivalued dynamical systems generated by homogeneous convex stochastic operators. Set-Valued Anal 6:61–81
Gaitsgory V, Grune L, Thatcher N (2015) Stabilization with discounted optimal control. Syst Control Lett 82:91–98
Gaitsgory V, Rossomakhine S, Thatcher N (2012) Approximate solution of the HJB inequality related to the infinite horizon optimal control problem with discounting. Dyn Continuous Discrete Impuls Syst Ser B 19:65–92
Gale D (1967) On optimal development in a multi-sector economy. Rev Econ Stud 34:1–18
Ghosh MK, Mallikarjuna Rao KS (2005) Differential games with ergodic payoff. SIAM J Control Optim 43:2020–2035
Jasso-Fuentes H, Hernandez-Lerma O (2008) Characterizations of overtaking optimality for controlled diffusion processes. Appl Math Optim 57:349–369
Khan MA, Zaslavski AJ (2010) On two classical turnpike results for the Robinson-Solow-Srinivisan (RSS) model. Adv Math Econ 13:47–97
Khlopin DV (2013) Necessity of vanishing shadow price in infinite horizon control problems. J Dyn Control Syst 19:519–552
Kolokoltsov V, Yang W (2012) The turnpike theorems for Markov games. Dyn Games Appl 2:294–312
Leizarowitz A (1985) Infinite horizon autonomous systems with unbounded cost. Appl Math Opt 13:19–43
Leizarowitz A (1986) Tracking nonperiodic trajectories with the overtaking criterion. Appl Math Opt 14:155–171
Leizarowitz A, Mizel VJ (1989) One dimensional infinite horizon variational problems arising in continuum mechanics. Arch Ration Mech Anal 106:161–194
Leizarowitz A, Zaslavski AJ (2005) On a class of infinite horizon optimal control problems with periodic cost functions. J Nonlinear Convex Anal 6:71–91
Lykina V, Pickenhain S, Wagner M (2008) Different interpretations of the improper integral objective in an infinite horizon control problem. J Math Anal Appl 340:498–510
Makarov VL, Rubinov AM (1977) Mathematical theory of economic dynamics and equilibria. Springer, New York
Malinowska AB, Martins N, Torres DFM (2011) Transversality conditions for infinite horizon variational problems on time scales. Optim Lett 5:41–53
Mammadov M (2014) Turnpike theorem for an infinite horizon optimal control problem with time delay. SIAM J Control Optim 52:420–438
Marcus M, Zaslavski AJ (1999) On a class of second order variational problems with constraints. Israel J Math 111:1–28
Marcus M, Zaslavski AJ (2002) The structure and limiting behavior of locally optimal minimizers. Ann Inst H Poincaré Anal Non Linéaire 19:343–370
McKenzie LW (1976) Turnpike theory. Econometrica 44:841–866
Mordukhovich BS (1990) Minimax design for a class of distributed parameter systems. Autom Remote Control 50:1333–1340
Moser J (1986) Minimal solutions of variational problems on a torus. Ann Inst H Poincaré Anal Non Linéaire 3:229–272
Ocana Anaya E, Cartigny P, Loisel P (2009) Singular infinite horizon calculus of variations. Applications to fisheries management. J Nonlinear Convex Anal 10:157–176
Pickenhain S, Lykina V, Wagner M (2008) On the lower semicontinuity of functionals involving Lebesgue or improper Riemann integrals in infinite horizon optimal control problems. Control Cybern 37:451–468
Prieto-Rumeau T, Hernandez-Lerma O (2005) Bias and overtaking equilibria for zero-sum continuous-time Markov games. Math Methods Oper Res 61:437–454
Reich S, Zaslavski AJ (2014) Genericity in nonlinear analysis. Springer, New York
Rockafellar RT (1970) Convex analysis. Princeton University Press, Princeton
Rubinov AM (1984) Economic dynamics. J Sov Math 26:1975–2012
Samuelson PA (1965) A catenary turnpike theorem involving consumption and the golden rule. Am Econ Rev 55:486–496
Trelat E, Zuazua E (2015) The turnpike property in finite-dimensional nonlinear optimal control. J Differ Equ 218:81–114
von Weizsacker CC (1965) Existence of optimal programs of accumulation for an infinite horizon. Rev Econ Stud 32:85–104
Zaslavski AJ (1987) Ground states in Frenkel-Kontorova model. Math USSR Izv 29:323–354
Zaslavski AJ (1995) Optimal programs on infinite horizon 1. SIAM J Control Optim 33:1643–1660
Zaslavski AJ (1998) Turnpike theorem for convex infinite dimensional discrete-time control systems. J Convex Anal 5:237–248
Zaslavski AJ (1999) Turnpike property for dynamic discrete time zero-sum games. Abst Appl Anal 4:21–48
Zaslavski AJ (2000) Turnpike theorem for nonautonomous infinite dimensional discrete-time control systems. Optimization 48:69–92
Zaslavski AJ (2005) The turnpike property of discrete-time control problems arising in economic dynamics. Discrete Continuous Dyn Syst B 5:861–880
Zaslavski AJ (2006) Turnpike properties in the calculus of variations and optimal control. Springer, New York
Zaslavski AJ (2007) Turnpike results for a discrete-time optimal control systems arising in economic dynamics. Nonlinear Anal 67:2024–2049
Zaslavski AJ (2010) Stability of a turnpike phenomenon for a discrete-time optimal control systems. J Optim Theory Appl 145:597–612
Zaslavski AJ (2010) Optimization on metric and normed spaces. Springer, New York
Zaslavski AJ (2011) Turnpike properties of approximate solutions for discrete-time control systems. Commun Math Anal 11:36–45
Zaslavski AJ (2011) Stability of a turnpike phenomenon for a class of optimal control systems in metric spaces. Numer Algebra Control Optim 1:245–260
Zaslavski AJ (2011) The existence and structure of approximate solutions of dynamic discrete time zero-sum games. J Nonlinear Convex Anal 12:49–68
Zaslavski AJ (2012) A generic turnpike result for a class of discrete-time optimal control systems. Dyn Continuous Discrete Impuls Syst Ser B 19:225–265
Zaslavski AJ (2013) Nonconvex optimal control and variational problems. Springer optimization and its applications. Springer, New York
Zaslavski AJ (2013) Two turnpike results for dynamic discrete time zero-sum games. J Nonlinear Convex Anal 14:687–717
Zaslavski AJ (2014) Turnpike properties of approximate solutions of nonconcave discrete-time optimal control problems. J Convex Anal 21:681–701
Zaslavski AJ (2014) Turnpike phenomenon and infinite horizon optimal control. Springer optimization and its applications. Springer, New York
Zaslavski AJ (2014) Turnpike properties for nonconcave problems. Adv Math Econ 18:101–134
Zaslavski AJ (2014) Structure of solutions of discrete time optimal control problems in the regions close to the endpoints. Set-Valued Var Anal 22:809–842
Zaslavski AJ (2014) Turnpike theory for dynamic zero-sum games. In: Proceedings of the workshop “Variational and optimal control problems on unbounded domains”, Haifa, 2012. Contemporary mathematics, vol. 619, pp 225–247
Zaslavski AJ (2014) Turnpike properties of approximate solutions of dynamic discrete time zero-sum games. J Dyn Games 2014:299–330
Zaslavski AJ (2015) Convergence of solutions of optimal control problems with discounting on large intervals in the regions close to the endpoints. Set-Valued Var Anal 23:191–204
Zaslavski AJ (2015) Structure of approximate solutions of discrete time optimal control Bolza problems on large intervals. Nonlinear Anal 123/124:23–55
Zaslavski AJ (2015) Discrete time optimal control problems on large intervals. Adv Math Econ 19:91–135
Zaslavski AJ (2015) Turnpike theory of continuous-time linear optimal control problems. Springer optimization and its applications. Springer, New York
Zaslavski AJ (2016) Convergence of solutions of concave discrete optimal control problems in the regions close to the endpoints. Commun Appl Nonlinear Anal 23:1–10
Zaslavski AJ (2016) Structure of solutions of optimal control problems on large intervals: a survey of recent results. Pure Appl Funct Anal 1:123–158
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this chapter
Cite this chapter
Zaslavski, A.J. (2017). Introduction. In: Discrete-Time Optimal Control and Games on Large Intervals. Springer Optimization and Its Applications, vol 119. Springer, Cham. https://doi.org/10.1007/978-3-319-52932-5_1
Download citation
DOI: https://doi.org/10.1007/978-3-319-52932-5_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-52931-8
Online ISBN: 978-3-319-52932-5
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)