Optimal production mix

This paper deals with a production plant in which two different products can be produced. The plant consists of three subsystemsS _i . Before or after a phase of separate processing in subsystemsS ₁ andS ₂, the two products have to be processed in subsystemS ₃. Each of these subsystems has a limited capacity.In the first part, we assume empty stocks at the beginning; at a fixed timeT in the future, certain quantitiesX _i of the two products have to be delivered to the customers. Facing linear holding costs, convex production costs, and stringent capacity constraints, the problem is to decide when to produce which product at what rate.It is shown that the optimal solution consists of up to six different regimes and that the time paths of the production rates need not be monotonic. These results, which can be obtained analytically, are also illustrated in several numerical examples.Finally, the case is considered where the terminal demand at timeT is replaced by a continuous and seasonally fluctuating demand rate. It is demonstrated that the optimal production rates show an interesting and nontrivial behavior. In particular, it may happen that, on intervals where the demand for the one product increases, the optimal production rate decreases. This is also demonstrated by computer plots in some numerical examples.The first author gratefully acknowledges support from the Austrian Science Foundation under Grant S3204 and the second author from Stiftung Volkswagenwerk. An earlier version of this paper was presented at the DGOR-NSOR Joint Conference, Eindhoven, Holland, September 23–25, 1987.     

Multiple local solutions to nonlinear control processes

We use the topological tool of Nielsen fixed-point theory to study the controllability of perturbed linear control processes whose control space can be reduced to a finite-dimensional one. These methods produce a lower bound on the number of controls that achieve a given target. We concentrate particularly on the case where the perturbation has sublinear growth, and we exhibit examples of multiple controls of various types that are detected by the Nielsen theory method.This research was partly supported by the Research Grant Teoria del Controllo dei Sistemi Dinamici, Ministero della Pubblica Istruzione, Roma, Italy.     

Stabilization and control of distributed systems with time-dependent spatial domains

This paper considers the problem of the stabilization and control of distributed systems with time-dependent spatial domains. The evolution of the spatial domains with time is described by a finite-dimensional system of ordinary differential equations, while the distributed systems are described by first-order or second-order linear evolution equations defined on appropriate Hilbert spaces. First, results pertaining to the existence and uniqueness of solutions of the system equations are presented. Then, various optimal control and stabilization problems are considered. The paper concludes with some examples which illustrate the application of the main results.This work was supported by the Air Force Office of Scientific Research, Grant No. AFOSR 86-0132, by the National Science Foundation, Grant No. 87-18473, and by the Jet Propulsion Laboratory, Pasadena, California.     

The Kumar-Becker-Lin scheme revisited

The Kumar-Becker-Lin scheme introduces a slowly vanishing cost bias in the parameter estimation part of self-tuning control in order to improve its performance. This paper establishes the a.s. optimality of a variant of this scheme for Markov chains on a countable state space when the action space is compact metric and the parameter space is a compact subset ofR ^m .     

Deterministic equivalent for a continuous-time linear-convex stochastic control problem

We consider a finite-horizon control model with additive input. There are two convex functions which describe the running cost and the terminal cost within the system. The cost of input is proportional to the input and can take both positive and negative values. It is shown that there exists a deterministic control problem whose optimal cost is the same as the one in the stochastic control problem. The optimal policy for the stochastic problem consists of keeping the process as close to the optimal deterministic trajectory as possible.This research is supported by NSERC Grant A4619, MRCO, NSF Grant DMS-86-01510, and AFOSR Grant 87-0278.     

Deterministic impulse control in native forest ecosystems management

This paper illustrates an application of impulse control systems analysis to the management of renewable resources. We obtain an explicit definition for the impulse optimal strategy in the case of a nonlinear natural system which models certain types of forests. The result is in agreement with an econo-ecological conservative action on the forest. The examples developed in this paper associate new theories on impulse or jump control with natural forest management. In addition, we compare our results with the classical ones. We also show that a nonoptimal managing action on the forest is frequently the cause for significant medium-term losses.This work was supported by CONICET, Consejo Nacional de Investigaciones Científicas y Técnicas, and by the Universidad Nacional del Litoral, Argentina.     

里德伯钠原子经绝热快速通道的布居数跃迁相干控制

运用含时多态展开方法和B-样条函数研究了啁啾频率微波场中里德伯钠原子的量子态之间的布居数迁移.计算了里德伯钠原子n=70-77的开普勒频率.计算了在不同的微波场中六个态的布居数从n=70到n=77随时间的迁移,布居数跃迁到最终态n=77达到了98%,这可以通过连续的单光子跃迁来实现.结果表明,通过优化微波脉冲参数可以实现从低态到较高态的布居跃迁的相干控制.     

On the Exceptional Zeros of Rankin–Selberg L-Functions

The main objects of study in this article are two classes of Rankin–Selberg L-functions, namely L(s,f×g) and L(s, sym²(g)× sym²(g)), where f,g are newforms, holomorphic or of Maass type, on the upper half plane, and sym²(g) denotes the symmetric square lift of g to GL(3). We prove that in general, i.e., when these L-functions are not divisible by L-functions of quadratic characters (such divisibility happening rarely), they do not admit any LandauSiegel zeros. Such zeros, which are real and close to s=1, are highly mysterious and are not expected to occur. There are corollaries of our result, one of them being a strong lower bound for special value at s=1, which is of interest both geometrically and analytically. One also gets this way a good bound onthe norm of sym²(g).     

Index Heuristics for Multiclass M/G/1 Systems with Nonpreemptive Service and Convex Holding Costs

We consider the optimal service control of a multiclass M/G/1 queueing system in which customers are served nonpreemptively and the system cost rate is additive across classes and increasing convex in the numbers present in each class. Following Whittle's approach to a class of restless bandit problems, we develop a Langrangian relaxation of the service control problem which serves to motivate the development of a class of index heuristics. The index for a particular customer class is characterised as a fair charge for service of that class. The paper develops these indices and reports an extensive numerical investigation which exhibits strong performance of the index heuristics for both discounted and average costs.     

Application of a Near-Optimal Reinforcement Learning Controller to a Robotics Problem in Manufacturing: A Hybrid Approach

Optimization theory provides a framework for determining the best decisions or actions with respect to some mathematical model of a process. This paper focuses on learning to act in a near-optimal manner through reinforcement learning for problems that either have no model or the model is too complex. One approach to solving this class of problems is via approximate dynamic programming. The application of these methods are established primarily for the case of discrete state and action spaces. In this paper we develop efficient methods of learning which act in complex systems with continuous state and action spaces. Monte-Carlo approaches are employed to estimate function values in an iterative, incremental procedure. Derivative-free line search methods are used to obtain a near-optimal action in the continuous action space for a discrete subset of the state space. This near-optimal control policy is then extended to the entire continuous state space via a fuzzy additive model. To compensate for approximation errors, a modified procedure for perturbing the generated control policy is developed. Convergence results under moderate assumptions and stopping criteria are established.