面向多最优解组合优化问题的决策求解算法

针对具有固定物品总和、多最优解特征的组合优化问题,以固定总和实数子集问题和购买鸡翅问题为例,给出了这类多最优解组合优化问题的形式化表示。在分析枚举等经典算法基础上,提出了基于整数状态表示和实数状态表示的0-1决策递归搜索多最优解动态规划算法。针对该算法在最优解数量较大时,时间复杂度趋向O(mⁿ)的问题,提出了基于相同决策路径合并和基于0-x决策的两种改进算法。实验中两种改进算法的计算时间基本符合与O(nb+nm)的正比关系,表明对于这类多最优解组合优化问题具有良好的求解性能。     

Minimizing weighted mean absolute deviation of flow times in single machine systems

We discuss the problem of scheduling several jobs on a single machine with the objective of minimizing the weighted mean absolute deviation of flow times around the weighted mean flow time. We first show that the optimal schedule is W-shaped. For the unweighted case, we show that all optimal schedules are V-shaped. This characterization enables us to show that the problem is NP-hard. We then provide a pseudopolynomial algorithm for the unweighted problem. Finally, we consider three heuristic algorithms for the unweighted problem and report computational experience with these algorithms. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 297–311, 1998     

Class scheduling algorithms for Navy training schools

The problem of developing good schedules for Navy C-Schools has been modeled as a combinatorial optimization problem. The only complicating feature of the problem is that classes must be grouped together into sequences known as pipelines. An ideal schedule will have all classes in a pipeline scheduled in consecutive weeks. The objective is to eliminate the nonproductive time spent by sailors at C-Schools who are waiting for the next class in a pipeline. In this investigation an implicit enumeration procedure for this problem was developed. The key component of our algorithm is a specialized greedy algorithm which is used to obtain a good initial incumbent. Often this initial incumbent is either an optimal schedule or a near optimal schedule. In an empirical analysis with the only other competing software system, our greedy heuristic found equivalent or better solutions in substantially less computer time. This greedy heuristic was extended and modified for the A-School scheduling problem and was found to be superior to its only competitor. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 533–551, 1998     

A multifaceted heuristic for the orienteering problem

The orienteering problem involves the selection of a path between an origin and a destination which maximizes total score subject to a time restriction. In previous work we presented an effective heuristic for this NP-hard problem that outperformed other heuristics from the literature. In this article we describe and test a significantly improved procedure. The new procedure is based on four concepts—center of gravity, randomness, subgravity, and learning. These concepts combine to yield a procedure which is much faster and which results in more nearly optimal solutions than previous procedures.     

The knapsack problem with disjoint multiple-choice constraints

In this article we consider the binary knapsack problem under disjoint multiple-choice constraints. We propose a two-stage algorithm based on Lagrangian relaxation. The first stage determines in polynomial time an optimal Lagrange multiplier value, which is then used within a branch-and-bound scheme to rank-order the solutions, leading to an optimal solution in a relatively low depth of search. The validity of the algorithm is established, a numerical example is included, and computational experience is described.     

Single machine scheduling with a restricted rate‐modifying activity

In this paper we consider the problem of scheduling a set of jobs on a single machine on which a rate‐modifying activity may be performed. The rate‐modifying activity is an activity that changes the production rate of the machine. So the processing time of a job is a variable, which depends on whether it is scheduled before or after the rate‐modifying activity. We assume that the rate‐modifying activity can take place only at certain predetermined time points, which is a constrained case of a similar problem discussed in the literature. The decisions under consideration are whether and when to schedule the rate‐modifying activity, and how to sequence the jobs in order to minimize some objectives. We study the problems of minimizing makespan and total completion time. We first analyze the computational complexity of both problems for most of the possible versions. The analysis shows that the problems are NP‐hard even for some special cases. Furthermore, for the NP‐hard cases of the makespan problem, we present a pseudo‐polynomial time optimal algorithm and a fully polynomial time approximation scheme. For the total completion time problem, we provide a pseudo‐polynomial time optimal algorithm for the case with agreeable modifying rates. © 2005 Wiley Periodicals, Inc. Naval Research Logistics, 2005     

Capacity expansion and cost efficiency improvement in the warehouse problem

The warehouse problem with deterministic production cost, selling prices, and demand was introduced in the 1950s and there is a renewed interest recently due to its applications in energy storage and arbitrage. In this paper, we consider two extensions of the warehouse problem and develop efficient computational algorithms for finding their optimal solutions. First, we consider a model where the firm can invest in capacity expansion projects for the warehouse while simultaneously making production and sales decisions in each period. We show that this problem can be solved with a computational complexity that is linear in the product of the length of the planning horizon and the number of capacity expansion projects. We then consider a problem in which the firm can invest to improve production cost efficiency while simultaneously making production and sales decisions in each period. The resulting optimization problem is non‐convex with integer decision variables. We show that, under some mild conditions on the cost data, the problem can be solved in linear computational time. © 2016 Wiley Periodicals, Inc. Naval Research Logistics 63: 367–373, 2016     

Scheduling with tool changes to minimize total completion time: A study of heuristics and their performance

The machine scheduling literature does not consider the issue of tool change. The parallel literature on tool management addresses this issue but assumes that the change is due only to part mix. In practice, however, a tool change is caused most frequently by tool wear. That is why we consider here the problem of scheduling a set of jobs on a single CNC machine where the cutting tool is subject to wear; our objective is to minimize the total completion time. We first describe the problem and discuss its peculiarities. After briefly reviewing available theoretical results, we then go on to provide a mixed 0–1 linear programming model for the exact solution of the problem; this is useful in solving problem instances with up to 20 jobs and has been used in our computational study. As our main contribution, we next propose a number of heuristic algorithms based on simple dispatch rules and generic search. We then discuss the results of a computational study where the performance of the various heuristics is tested; we note that the well‐known SPT rule remains good when the tool change time is small but deteriorates as this time increases and further that the proposed algorithms promise significant improvement over the SPT rule. © 2002 Wiley Periodicals, Inc. Naval Research Logistics, 2003     

A competitive genetic algorithm for resource-constrained project scheduling

In this paper we consider the resource-constrained project scheduling problem (RCPSP) with makespan minimization as objective. We propose a new genetic algorithm approach to solve this problem. Subsequently, we compare it to two genetic algorithm concepts from the literature. While our approach makes use of a permutation based genetic encoding that contains problem-specific knowledge, the other two procedures employ a priority value based and a priority rule based representation, respectively. Then we present the results of our thorough computational study for which standard sets of project instances have been used. The outcome reveals that our procedure is the most promising genetic algorithm to solve the RCPSP. Finally, we show that our genetic algorithm yields better results than several heuristic procedures presented in the literature. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 733–750, 1998     

Dynamic pricing and replenishment: Optimality,bounds, and asymptotics

In many applications, managers face the problem of replenishing and selling products during a finite time horizon. We investigate the problem of making dynamic and joint decisions on product replenishment and selling in order to improve profit. We consider a backlog scenario in which penalty cost (resulting from fulfillment delay) and accommodation cost (resulting from shortage at the end of the selling horizon) are incurred. Based on continuous‐time and discrete‐state dynamic programming, we study the optimal joint decisions and characterize their structural properties. We establish an upper bound for the optimal expected profit and develop a fluid policy by resorting to the deterministic version of the problem (ie, the fluid problem). The fluid policy is shown to be asymptotically optimal for the original stochastic problem when the problem size is sufficiently large. The static nature of the fluid policy and its lack of flexibility in matching supply with demand motivate us to develop a “target‐inventory” heuristic, which is shown, numerically, to be a significant improvement over the fluid policy. Scenarios with discrete feasible sets and lost‐sales are also discussed in this article.     

Analysis of the greedy approach in problems of maximum k-coverage

In this paper, we consider a general covering problem in which k subsets are to be selected such that their union covers as large a weight of objects from a universal set of elements as possible. Each subset selected must satisfy some structural constraints. We analyze the quality of a k-stage covering algorithm that relies, at each stage, on greedily selecting a subset that gives maximum improvement in terms of overall coverage. We show that such greedily constructed solutions are guaranteed to be within a factor of 1 − 1/e of the optimal solution. In some cases, selecting a best solution at each stage may itself be difficult; we show that if a β-approximate best solution is chosen at each stage, then the overall solution constructed is guaranteed to be within a factor of 1 − 1/e^β of the optimal. Our results also yield a simple proof that the number of subsets used by the greedy approach to achieve entire coverage of the universal set is within a logarithmic factor of the optimal number of subsets. Examples of problems that fall into the family of general covering problems considered, and for which the algorithmic results apply, are discussed. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 615–627, 1998     

Solution algorithms for the parallel replacement problem under economy of scale

We consider the parallel replacement problem in which machine investment costs exhibit economy of scale which is modeled through associating both fixed and variable costs with machine investment costs. Both finite- and infinite-horizon cases are investigated. Under the three assumptions made in the literature on the problem parameters, we show that the finite-horizon problem with time-varying parameters is equivalent to a shortest path problem and hence can be solved very efficiently, and give a very simple and fast algorithm for the infinite-horizon problem with time-invariant parameters. For the general finite-horizon problem without any assumption on the problem parameters, we formulate it as a zero-one integer program and propose an algorithm for solving it exactly based on Benders' decomposition. Computational results show that this solution algorithm is efficient, i.e., it is capable of solving large scale problems within a reasonable cpu time, and robust, i.e., the number of iterations needed to solve a problem does not increase quickly with the problem size. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 279–295, 1998     

Branch and bound methods for a search problem

The problem of searching for randomly moving targets such as children and submarines is known to be fundamentally difficult, but finding efficient methods for generating optimal or near optimal solutions is nonetheless an important practical problem. This paper investigates the efficiency of Branch and Bound methods, with emphasis on the tradeoff between the accuracy of the bound employed and the time required to compute it. A variety of bounds are investigated, some of which are new. In most cases the best bounds turn out to be imprecise, but very easy to compute. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 243–257, 1998     

Continuous-time markov decision processes with nonzero terminal reward

In this article we consider a continuous-time Markov decision process with a denumerable state space and nonzero terminal rewards. We first establish the necessary and sufficient optimality condition without any restriction on the cost functions. The necessary condition is derived through the Pontryagin maximum principle and the sufficient condition, by the inherent structure of the problem. We introduce a dynamic programming approximation algorithm for the finite-horizon problem. As the time between discrete points decreases, the optimal policy of the discretized problem converges to that of the continuous-time problem in the sense of weak convergence. For the infinite-horizon problem, a successive approximation method is introduced as an alternative to a policy iteration method.     

Makespan minimization in the two‐machine flowshop batch scheduling problem

In this paper we consider a practical scheduling problem commonly arising from batch production in a flexible manufacturing environment. Different part‐types are to be produced in a flexible manufacturing cell organized into a two‐stage production line. The jobs are processed in batches on the first machine, and the completion time of a job is defined as the completion time of the batch containing it. When processing of all jobs in a batch is completed on the first machine, the whole batch of jobs is transferred intact to the second machine. A constant setup time is incurred whenever a batch is formed on any machine. The tradeoff between the setup times and batch processing times gives rise to the batch composition decision. The problem is to find the optimal batch composition and the optimal schedule of the batches so that the makespan is minimized. The problem is shown to be strongly NP‐hard. We identify some special cases by introducing their corresponding solution methods. Heuristic algorithms are also proposed to derive approximate solutions. We conduct computational experiments to study the effectiveness of the proposed heuristics. © 2000 John Wiley & Sons, Inc. Naval Research Logistics 47: 128–144, 2000     

A worst‐case formulation for constrained ranking and selection with input uncertainty

In this research, we consider robust simulation optimization with stochastic constraints. In particular, we focus on the ranking and selection problem in which the computing time is sufficient to evaluate all the designs (solutions) under consideration. Given a fixed simulation budget, we aim at maximizing the probability of correct selection (PCS) for the best feasible design, where the objective and constraint measures are assessed by their worst‐case performances. To simplify the complexity of PCS, we develop an approximated probability measure and derive the asymptotic optimality condition (optimality condition as the simulation budget goes to infinity) of the resulting problem. A sequential selection procedure is then designed within the optimal computing budget allocation framework. The high efficiency of the proposed procedure is tested via a number of numerical examples. In addition, we provide some useful insights into the efficiency of a budget allocation procedure.     

Probabilistic analysis of an asymptotically optimal solution for the completion time variance problem

Scheduling a set of n jobs on a single machine so as to minimize the completion time variance is a well‐known NP‐hard problem. In this paper, we propose a sequence, which can be constructed in O(n log n) time, as a solution for the problem. Our primary concern is to establish the asymptotical optimality of the sequence within the framework of probabilistic analysis. Our main result is that, when the processing times are randomly and independently drawn from the same uniform distribution, the sequence is asymptotically optimal in the sense that its relative error converges to zero in probability as n increases. Other theoretical results are also derived, including: (i) When the processing times follow a symmetric structure, the problem has 2^{⌊(n−1)/2⌋} optimal sequences, which include our proposed sequence and other heuristic sequences suggested in the literature; and (ii) when these 2^{⌊(n−1)/2⌋} sequences are used as approximate solutions for a general problem, our proposed sequence yields the best approximation (in an average sense) while another sequence, which is commonly believed to be a good approximation in the literature, is interestingly the worst. © 1999 John Wiley & Sons, Inc. Naval Research Logistics 46: 373–398, 1999     

Bounding the optimal burn‐in time for a system with two types of failure

Burn‐in is a widely used method to improve the quality of products or systems after they have been produced. In this paper, we consider the problem of determining bounds to the optimal burn‐in time and optimal replacement policy maximizing the steady state availability of a repairable system. It is assumed that two types of system failures may occur: One is Type I failure (minor failure), which can be removed by a minimal repair, and the other is Type II failure (catastrophic failure), which can be removed only by a complete repair. Assuming that the underlying lifetime distribution of the system has a bathtub‐shaped failure rate function, upper and lower bounds for the optimal burn‐in time are provided. Furthermore, some other applications of optimal burn‐in are also considered. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2004     

Optimal lot sizing and inspection policy for an EMQ model with imperfect inspections

An EMQ model with a production process subject to random deterioration is considered. The process can be monitored through inspections, and both the lot size and the inspection schedule are subject to control. The “in-control” periods are assumed to be generally distributed and the inspections are imperfect, i.e., the true state of the process is not necessarily revealed through an inspection. The objective is the joint determination of the lot size and the inspection schedule, minimizing the long-run expected average cost per unit time. Both discrete and continuous cases are examined. A dynamic programming formulation is considered in the case where the inspections can be performed only at discrete times, which is typical for the parts industry. In the continuous case, an optimum inspection schedule is obtained for a given production time and given number of inspections by solving a nonlinear programming problem. A two-dimensional search procedure can be used to find the optimal policy. In the exponential case, the structure of the optimal inspection policy is established using Lagrange's method, and it is shown that the optimal inspection times can be found by solving a nonlinear equation. Numerical studies indicate that the optimal policy performs much better than the optimal policy with periodic inspections considered previously in the literature. The case of perfect inspections is discussed, and an extension of the results obtained previously in the literature is presented. © 1998 John Wiley & Sons, Inc. Naval Research Logistics 45: 165–186, 1998     

Algorithms for the multi‐item multi‐vehicles dynamic lot sizing problem

We consider a two‐stage supply chain, in which multi‐items are shipped from a manufacturing facility or a central warehouse to a downstream retailer that faces deterministic external demand for each of the items over a finite planning horizon. The items are shipped through identical capacitated vehicles, each incurring a fixed cost per trip. In addition, there exist item‐dependent variable shipping costs and inventory holding costs at the retailer for items stored at the end of the period; these costs are constant over time. The sum of all costs must be minimized while satisfying the external demand without backlogging. In this paper we develop a search algorithm to solve the problem optimally. Our search algorithm, although exponential in the worst case, is very efficient empirically due to new properties of the optimal solution that we found, which allow us to restrict the number of solutions examined. Second, we perform a computational study that compares the empirical running time of our search methods to other available exact solution methods to the problem. Finally, we characterize the conditions under which each of the solution methods is likely to be faster than the others and suggest efficient heuristic solutions that we recommend using when the problem is large in all dimensions. © 2005 Wiley Periodicals, Inc. Naval Research Logistics, 2006.