Showing papers on "Heuristic (computer science) published in 1994"

Genetic Algorithms and Random Keys for Sequencing and Optimization

Abstract: In this paper we present a general genetic algorithm to address a wide variety of sequencing and optimization problems including multiple machine scheduling, resource allocation, and the quadratic assignment problem. When addressing such problems, genetic algorithms typically have difficulty maintaining feasibility from parent to offspring. This is overcome with a robust representation technique called random keys. Computational results are shown for multiple machine scheduling, resource allocation, and quadratic assignment problems. INFORMS Journal on Computing, ISSN 1091-9856, was published as ORSA Journal on Computing from 1989 to 1995 under ISSN 0899-1499.

The Reactive Tabu Search

Abstract: We propose an algorithm for combinatorial optimization where an explicit check for the repetition of configurations is added to the basic scheme of Tabu search. In our Tabu scheme the appropriate size of the list is learned in an automated way by reacting to the occurrence of cycles. In addition, if the search appears to be repeating an excessive number of solutions excessively often, then the search is diversified by making a number of random moves proportional to a moving average of the cycle length. The reactive scheme is compared to a “strict” Tabu scheme that forbids the repetition of configurations and to schemes with a fixed or randomly varying list size. From the implementation point of view we show that the Hashing or Digital Tree techniques can be used in order to search for repetitions in a time that is approximately constant. We present the results obtained for a series of computational tests on a benchmark function, on the 0-1 Knapsack Problem, and on the Quadratic Assignment Problem. INFORMS...

The traveling salesman: computational solutions for TSP applications

Abstract: Basic Concepts.- Related Problems and Applications.- Geometric Concepts.- Candidate Sets.- Construction Heuristics.- Improving Solutions.- Heuristics for Large Geometric Problems.- Further Heuristic Approaches.- Lower Bounds.- A Case Study: TSPs in Printed Circuit Board Production.- Practical TSP Solving.

Ant system for Job-shop scheduling

Abstract: The study of natural processes has inspired several heuristic optimization algorithms which have proved to be very effective in combinatorial optimization. In this paper we show how a new heuristic called ant system, in which the search task is distributed over manysimple, loosely interacting agents, can be successfully applied to find good solutions of job-shop scheduling problems.

A Greedy Randomized Adaptive Search Procedure for Maximum Independent Set

Abstract: An efficient randomized heuristic for a maximum independent set is presented. The procedure is tested on randomly generated graphs having from 400 to 3,500 vertices and edge probabilities from 0.2 to 0.9. The heuristic can be implemented trivially in parallel and is tested on an MIMD computer with 1, 2, 4 and 8 processors. Computational results indicate that the heuristic frequently finds the optimal or expected optimal solution in a fraction of the time required by implementations of simulated annealing, tabu search, and an exact partial enumeration method.

A No-wait Flowshop Scheduling Heuristic to Minimize Makespan

Abstract: The scheduling problem in the no-wait or constrained flowshop, with the makespan objective, is considered in this article. A simple heuristic algorithm is proposed on the basis of heuristic preference relations and job insertion. When evaluated over a large number of problems of various sizes, the solutions given by the proposed heuristic are found to be fairly accurate and much superior to those given by the two existing heuristics.

Solution reuse in dynamic constraint satisfaction problems

Abstract: Many AI problems can be modeled as constraint satisfaction problems (CSP), but many of them are actually dynamic: the set of constraints to consider evolves because of the environment, the user or other agents in the framework of a distributed system. In this context, computing a new solution from scratch after each problem change is possible, but has two important drawbacks: inefficiency and instability of the successive solutions. In this paper, we propose a method for reusing any previous solution and producing a new one by local changes on the previous one. First we give the key idea and the corresponding algorithm. Then we establish its properties: termination, correctness and completeness. We show how it can be used to produce a solution, either from an empty assignment, or from any previous assignment and how it can be improved using filtering or learning methods, such as forward-checking or nogood-recording. Experimental results related to efficiency and stability are given, with comparisons with well known algorithms such as backtrack, heuristic repair or dynamic backtracking.

Strategies for Creative Problem Solving

Abstract: 1. Introduction: A Rationale for Problem Solving Strategies. 2. A Heuristic. 3. Problem Definition. 4. Generating Alternatives. 5. Decide the Course of Action. 6. Carry Through. 7. Evaluation. Index.

Methods for optimization of nonlinear problems with discrete variables : a review

Abstract: The methods for discrete-integer-continuous variable nonlinear optimization are reviewed. They are classified into the following six categories: branch and bound, simulated annealing, sequential linearization, penalty functions, Lagrangian relaxation, and other methods. Basic ideas of each method are described and details of some of the algorithms are given. They are transcribed into a step-by-step format for easy implementation into a computer. Under “other methods”, rounding-off, heuristic, cutting-plane, pure discrete, and genetic algorithms are described. For nonlinear problems, none of the methods are guaranteed to produce the global minimizer; however, “good practical” solutions can be obtained.

The Multi-Airport Ground-Holding Problem in Air Traffic Control

Abstract: Motivated by the important problem of congestion costs (they were estimated to be $2 billion in 1991) in air transportation and observing that ground delays are more preferable than airborne delays, we have formulated and studied several integer programming models to assign ground-holding delays optimally in a general network of airports, so that the total (ground plus airborne) delay cost of all flights is minimized. All previous research on this problem has been restricted to the single-airport case, which neglects “down-the-road” effects due to transmission of delays between successive flights performed by the same aircraft. We formulate several models, and then propose a heuristic algorithm which finds a feasible solution to the integer program by rounding the optimal solution of the LP relaxation. Finally, we present extensive computational results with the goal of obtaining qualitative insights on the behavior of the problem under various combinations of the input parameters. We demonstrate that the...

Synthesis of fixed-structure controllers via numerical optimization

Abstract: We propose an iterative algorithm for designing linear time-invariant controllers with some prespecified structure. The iterations require the solution of optimization problems based on linear matrix inequalities, in which either the Lyapunov function proving a certain property or the controller to be designed is alternately regarded as the optimization variable (while the other is fixed). A number of structure constraints on the controller (reduced-order, decentralized, etc.) can be addressed using this technique, which also extends to plants with nonlinearities or uncertainties. The algorithm is heuristic in nature, and is not guaranteed to converge globally. However it provides a locally optimal solution which depends on the initialization of the algorithm, and serves as a useful design tool. >

Exact local solution of the continuous network design problem via stochastic user equilibrium assignment

Abstract: The continuous Network Design Problem (NDP) deals with determining optimal expansions for the capacities of a street network, subject to the constraint that the street traffic volumes must be the outcome of a user-optimal equilibrium assignment. Although the use of deterministic equilibrium methods tends to produce computationally intractable problems, in this paper it is shown that a stochastic user equilibrium based on the logit model leads to a differentiable and large-scale, but tractable, version of the NDP. A procedure for computing the derivatives of the stochastic user equilibrium (SUE) assignment without having to first compute the route choice probabilities is given, and this procedure is coupled with two standard algorithms for solving nonlinear programs, the generalized reduced gradient method and sequential quadratic programming. These algorithms are tested on several example networks, and the results of these tests suggest that the SUE-constrained version of the NDP offers both a promising heuristic for solving DUE-constrained problems as well as a viable procedure in its own right.

Efficient network flow based min-cut balanced partitioning

Abstract: We consider the problem of bipartitioning a circuit into two balanced components that minimizes the number of crossing nets. Previously, the Kernighan and Lin type (K&L) heuristics, the simulated annealing approach, and the spectral method were given to solve the problem. However, network flow techniques were overlooked as a viable approach to min-cut balanced bipartition to due its high complexity. In this paper we propose a balanced bipartition heuristic based on repeated max-flow min-cut techniques, and give an efficient implementation that has the same asymptotic time complexity as that of one max-flow computation. We implemented our heuristic algorithm in a package called FBB. The experimental results demonstrate that FBB outperforms the K&L heuristics and the spectral method in terms of the number of crossing nets, and the efficient implementation makes it possible to partition large, circuit instances with reasonable runtime. For example, the average elapsed time for bipartitioning a circuit S35932 of almost 20K gates is less than 20 minutes.

Multiple sequence alignment using simulated annealing

Abstract: Multiple sequence alignment is a useful technique for studying molecular evolution and analyzing structure-sequence relationships. Dynamic programming of multiple sequence alignment has been widely used to find an optimal alignment. However, dynamic programming does not allow for certain types of gap costs, and it limits the number of sequences that can be aligned due to its high computational complexity. The focus of this paper is to use simulated annealing as the basis for developing an efficient multiple sequence alignment algorithm. An algorithm called Multiple Sequence Alignment using Simulated Annealing (MSASA) has been developed. The computational complexity of MSASA is significantly reduced by replacing the high-temperature phase of the annealing process by a fast heuristic algorithm. This heuristic algorithm facilitates in minimizing the solution set of the low-temperature phase of the annealing process. Compared to the dynamic programming approach, MSASA can (i) use natural gap costs which can generate better solution, (ii) align more sequences and (iii) take less computation time.

Statistical Physics Algorithms That Converge

Abstract: In recent years there has been significant interest in adapting techniques from statistical physics, in particular mean field theory, to provide deterministic heuristic algorithms for obtaining approximate solutions to optimization problems. Although these algorithms have been shown experimentally to be successful there has been little theoretical analysis of them. In this paper we demonstrate connections between mean field theory methods and other approaches, in particular, barrier function and interior point methods. As an explicit example, we summarize our work on the linear assignment problem. In this previous work we defined a number of algorithms, including deterministic annealing, for solving the assignment problem. We proved convergence, gave bounds on the convergence times, and showed relations to other optimization algorithms.

Line-of-sight communication on terrain models

Abstract: Line-of-sight communication on topographic surfaces has relevance for several applications of Geographical Information Systems. In this paper, we study the problem of linking a set of transceiver stations in a visibility-connected communication network, by placing a minimum number of relays on the terrain surface. The problem is studied in the framework of a discrete visibility model, where the mutual visibility of a finite set of sites on the terrain is represented through a graph, called the visibility graph. While in the special case of only two transceivers an optimal solution can be found in polynomial time, by computing a minimum path on the visibility graph, the general problem is equivalent to a Steiner problem on the visibility graph, and, thus, it is untractable in practice. In the latter case, we propose a practical approximate solution based on a Steiner heuristic. For both the special and the general case, we propose both a static and a dynamic algorithm that allow computation of a s...

A promising hybrid GA/heuristic approach for open-shop scheduling problems

Abstract: Many problems in industry are a form of open-shop scheduling problem (OSSP). We describe a hybrid approach to this problem which combines a Genetic Algorithm (GA) with simple heuristic schedule building rules. Excellent performance is found on some benchmark OSS problems, including improvements on previous best-known results. We describe how our approach can be simply amended to deal with the more complex style of open shop scheduling problems which occur in industry, and discuss issues relating to further improvement of performance and integration of the approach into industrial job shop environments.

RELAX-IV : a faster version of the RELAX code for solving minimum cost flow problems

Abstract: The structure of dual ascent methods is particularly well-suited for taking advantage of good initial dual solutions of minimum cost flow problems. For this reason, these methods are extremely efficient for reoptimization and sensitivity analysis. In the absence of prior knowledge of a good initial dual solution, one may attempt to find such a solution by means of a heuristic initialization. RELAX-IV is a minimum cost flow code that combines the RELAX code of [BeT88a], [BeT88b] with an initialization based on a recently proposed auction/sequential shortest path algorithm. This initialization is shown to be extremely helpful in speeding up the solution of difficult problems, involving for example long augmenting paths, for which the relaxation method has been known to be slow. On the other hand, this initialization procedure does not significantly deteriorate the performance of the relaxation method for the types of problems where it has been known to be very fast.

A Tabu Search Approach for Delivering Pet Food and Flour in Switzerland

Abstract: In this paper, we consider a real-life vehicle routeing problem that occurs in a major Swiss company producing pet food and flour. In contrast with usual hypothetical problems, a large variety of restrictions has to be considered. The main constraints are relative to the accessibility and the time windows at customers, the carrying capacities of vehicles, the total duration of routes and the drivers' breaks. To find good solutions to this problem, we propose two heuristic methods: a fast straightforward insertion procedure and a method based on tabu search techniques. Next, the produced solutions are compared with the routes actually covered by the company. Our outcomes indicate that the total distance travelled can be reduced significantly when such methods are used.

Solving the quadratic assignment problem with clues from nature

Abstract: This paper describes a new evolutionary approach to solving quadratic assignment problems. The proposed technique is based loosely on a class of search and optimization algorithms known as evolution strategies (ES). These methods are inspired by the mechanics of biological evolution and have been applied successfully to a variety of difficult problems, particularly in continuous optimization. The combinatorial variant of ES presented here performs very well on the given test problems as compared with the standard 2-Opt heuristic and results with simulated annealing and tabu search. Extensions for practical applications in factory layout are described. >

Modeling and heuristic worst-case performance analysis of the two-level network design problem

Abstract: This paper studies a multi-facility network synthesis problem, called the Two-level Network Design (TLND) problem, that arises in the topological design of hierarchical communication, transportation, and electric power distribution networks. We are given an undirected network containing two types of nodes—primary and secondary—and fixed costs for installing either a primary or a secondary facility on each edge. Primary nodes require higher grade interconnections than secondary nodes, using the more expensive primary facilities. The TLND problem seeks a minimum cost connected design that spans all the nodes, and connects primary nodes via edges containing primary facilities; the design can use either primary or secondary edges to connect the secondary nodes. The TLND problem generalizes the well-known Steiner network problem and the hierarchical network design problem. In this paper, we study the relationship between alternative model formulations for this problem (e.g., directed and undirected models), an...

Exact and heuristic algorithms for the minimization of incompletely specified state machines

Abstract: In this paper we present two exact algorithms for state minimization of FSM's. Our results prove that exact state minimization is feasible for a large class of practical examples, certainly including most hand-designed FSM's. We also present heuristic algorithms, that can handle large, machine-generated, FSM's. The possibly many different reduced machines with the same number of states have different implementation costs. We discuss two steps of the minimization procedure, called state mapping and solution shrinking, that have received little prior attention to the literature, though they play a significant role in delivering an optimally implemented reduced machine. We also introduce an algorithm whose main virtue is the ability to cope with very general cost functions, while providing high performance. >

Some Experiences On Solving Multiconstraint Zero-One Knapsack Problems With Genetic Algorithms

Abstract: In the multiconstraint zero-one knapsack problem one has to decide on how to make efficient use of an entity which consumes multiple resources. The problem is known to be NP-hard, so heuristic solution procedures come into consideration.The purpose of this paper is to analyse different operators within genetic algorithms for solving the multiconstraint zero-one knapsack problem and to compare the results with those of recent investigations of other modern heuristic concepts presented in the literature. Numerical experiences emphasize that one has to incorporate local search improvement operators into simple genetic algorithms to obtain competitive results.