Abstract not found

In this paper, we propose a procedure, based on statistical design of experiments and gradient descent, that finds effective settings for parameters found in heuristics. We develop our procedure using four experiments. We use our procedure and a small subset of problems to find parameter settings for two new vehicle routing heuristics. We then set the parameters of each heuristic and solve 19 capacity-constrained and 15 capacity-constrained and route-length-constrained vehicle routing problems ranging in size from 50 to 483 customers. We conclude that our procedure is an effective method that deserves serious consideration by both researchers and operations research practitioners. Key Words: statistical design of experiments, heuristics, vehicle routing 1.

Neighborhood search algorithms are a wide class of improvement algorithms where at each iteration an improving solution is found by searching the “neighborhood ” of the current solution. This paper discusses neighborhood search algorithms where the size of the neighborhood is “very large” with respect to the size of the input data. For large problem instances, it is impractical to search these neighborhoods explicitly, and one must either search a small portion of the neighborhood or else develop efficient algorithms for searching the neighborhood implicitly. We concentrate on a very large scale neighborhood (VLSN) search technique based on compounding independent moves (CIM) such as 2-opts, swaps, and insertions. We demonstrate that the search for an improving neighbor can be done by finding a negative cost path on an auxiliary graph. In this paper we study CIM algorithms for the vehicle routing problem with capacity and distance constraints. We present results of the computational study which indicates that the CIM algorithms for the capacitated vehicle routing problem are competitive with the current state of the art heuristics.

Many transportation providers such as package delivery companies and postal service organizations face the problem of designing a transportation network in order to service their customers. This network must balance the requirement of on-time delivery under tight time window constraints with the goal of low-cost operations of the fleet. Until recently this task was usually performed by planners without sufficient software aid. This paper describes a decision support system (DSS) which has been designed in order to assist planners of the German postal service, the Deutsche Post AG. It helps the planners in designing improved plans which can be generated either manually or with the help of sophisticated intelligent optimization techniques. Both optimization and system design factors which influenced the design of the DSS are discussed.

The Aerial Fleet Refueling Problem (AFRP) is concerned with the efficient and effective use of a heterogeneous set of tanker (refueling) aircraft, located at diverse geographical locations, in the required operations associated with the deployment of a diverse fleet of military aircraft to a foreign theater of activity. Typically, the “receiving ” aircraft must traverse great distances over large bodies of water and/or over other inhospitable environs where no ground based refueling resources exist. Lacking the ability to complete their flights without refueling, each receiving aircraft must be serviced one or more times during their deployment flights by means of in-flight refueling provided by one of the available tanker aircraft. The receiving aircraft, aggregated into receiver groups (RGs) that fly together, have stipulated departure and destination bases and each RG’s arrival time is bounded by a stated desired earliest and latest time. The excellence of a suggested solution to this very challenging decision making problem is measured relative to a rigorously defined hierarchical multicriteria objective function. This paper describes how the AFRP for the Air Mobility Command (AMC), Scott AFB, IL is efficiently solved using Group Theoretic Tabu Search (GTTS). GTTS uses the symmetric group on n letters (Sn) and applies it to this problem using the Java TM language.