Tobias Buer

Solving a bi-objective winner determination problem in a transportation procurement auction

This paper introduces a bi-objective winner determination problem which arises in the procurement of transportation contracts via combinatorial auctions where bundle bidding is possible. The problem is modelled as a bi-objective extension to the set covering problem. We consider both the minimisation of the total procurement costs and the maximisation of the service-quality level at which the transportation contracts are executed. Taking into account the size of real-world transport auctions, a solution method has to cope with problems of up to some hundred contracts and a few thousand bundle bids. To solve the problem, we propose a bi-objective branch-and-bound algorithm and eight variants of a multiobjective genetic algorithm. Artificial benchmark instances that comply with important economic features of the transport domain are introduced to evaluate the methods. The branch-and-bound approach is able to find the optimal trade-off solutions in reasonable time for very small instances only. The eight variants of the genetic algorithm are compared among each other by means of large instances. The best variant is also evaluated using the small instances with known optimal solutions. The results indicate that the performance largely depends on the initialisation heuristic and suggest also that a well-balanced combination of genetic operators is crucial to obtain good solutions.

A collaborative ant colony metaheuristic for distributed multi-level lot-sizing

The paper presents an ant colony optimization metaheuristic for collaborative planning. Collaborative planning is used to coordinate individual plans of self-interested decision-makers with private information in order to increase the overall benefit of the coalition. The method consists of a new search graph based on encoded solutions. Distributed and private information are integrated via voting mechanisms and via a simple but effective collaborative local search procedure. The approach is applied to a distributed variant of the multi-level lot-sizing problem and evaluated by means of 352 benchmark instances from the literature. The proposed approach clearly outperforms existing approaches on the sets of medium- and large-sized instances. While the best method in the literature so far achieves an average deviation from the best-known non-distributed solutions of 75% for the set of the largest instances, for example, the presented approach reduces the average deviation to 7%.

A Pareto-metaheuristic for a bi-objective winner determination problem in a combinatorial reverse auction

The bi-objective winner determination problem (2WDP-SC) of a combinatorial procurement auction for transport contracts comes up to a multi-criteria set covering problem. We are given a set B of bundle bids. A bundle bid b in B consists of a bidding carrier c_b, a bid price p_b, and a set tau_b of transport contracts which is a subset of the set T of tendered transport contracts. Additionally, the transport quality q_t,c_b is given which is expected to be realized when a transport contract t is executed by a carrier c_b. The task of the auctioneer is to find a set X of winning bids (X is subset of B), such that each transport contract is part of at least one winning bid, the total procurement costs are minimized, and the total transport quality is maximized. This article presents a metaheuristic approach for the 2WDP-SC which integrates the greedy randomized adaptive search procedure, large neighborhood search, and self-adaptive parameter setting in order to find a competitive set of non-dominated solutions. The procedure outperforms existing heuristics. Computational experiments performed on a set of benchmark instances show that, for small instances, the presented procedure is the sole approach that succeeds to find all Pareto-optimal solutions. For each of the large benchmark instances, according to common multi-criteria quality indicators of the literature, it attains new best-known solution sets.

A Model for the Coordination of 20-foot and 40-foot Container Movements in the Hinterland of a Container Terminal

Considered is a carrier that requires decision support to organize an efficient transport of loaded and empty containers in the hinterland of a sea port. Loaded containers are handled as pickup-and-delivery requests, however, requests for empty containers are incomplete because either the pickup location of a container or the delivery location of a container is a priori unknown. The problem is modelled as a generalization of the pickup-and-delivery problem (PDP) with less-than-truckload (LTL) requests. Practically speaking, by using LTL request we are able to consider 20-foot and 40-foot containers simultaneously. This is closer to reality than most previous models discussed in the literature which use full truckload requests, i.e., only containers of homogeneous size are possible. Three types of decisions are involved in the proposed model: a determination of pickup or delivery locations for the incomplete requests, a routing of vehicles, and a routing of empty containers. The presented model is validated by means of a numerical example computed by a MIP solver.

Grasp with Hybrid Path Relinking for Bi-Objective Winner Determination in Combinatorial Transportation Auctions

The procurement of transportation services via large-scale combinatorial auctions involves a couple of complex decisions whose outcome highly influences the performance of the tender process. This paper examines the shipper’s task of selecting a subset of the submitted bids which efficiently trades off total procurement cost against expected carrier performance. To solve this bi-objective winner determination problem, we propose a Pareto-based greedy randomized adaptive search procedure (GRASP). As a post-optimizer we use a path relinking procedure which is hybridized with branch-and-bound. Several variants of this algorithm are evaluated by means of artificial test instances which comply with important real-world characteristics. The two best variants prove superior to a previously published Pareto-based evolutionary algorithm.

Agent-Negotiation of Lot-Sizing Contracts by Simulated Annealing with Part-Way Resets

The distributed multi-level uncapacitated lot-sizing problem is a group decision problem which has to be solved by a set of self-interested and autonomous agents. The agents represent independent companies which have to agree on a joint production plan in a supply chain context. In order to solve the problem we extend a negotiation-based simulated annealing approach introduced by Homberger by a part-way reset procedure. The part-way reset procedure allows a negotiation based search which reaches a deadlock to continue with a different contract proposal and thereby offers a possibility of overcoming disagreements between agents more easily. A benchmark study shows that the approach is competitive on set of 80 medium sized instances from the literature in terms of solution quality, in particular 47 new best-known solutions were computed.

Integrating Side Payments into Collaborative Planning for the Distributed Multi-level Unconstrained Lot Sizing Problem

Collaborative planning mechanisms coordinate the decisions of multiple, autonomous, and self-interested decisions makers under asymmetric information. The approach proposed in this paper extends collaborative planning for the distributed multi-level uncapacitated lot-sizing problem by integrating compensation payments. Compensation or side payments provide an incentive for individual decision makers to accept inferior local solutions that may direct the search to superior global solutions for a coalition of decision makers. The approach uses neighborhood search, voting-based solution acceptance criteria and takes into account varying side payments which are negotiated. Based on 272 benchmark instances the computational study shows that the presented approach is able to achieve substantial progress compared to earlier methods. It therefore is beneficial to incorporate side payments into negotiation processes based on collaborative search.

A Coordination Mechanism for a Collaborative Lot-Sizing Problem with Rivaling Agents

A distributed uncapacitated lot-sizing problem with rivaling agents (DULR) is introduced. The DULR considers concurrent items, i.e., items that can be produced by more than one agent. The agents are self-interested and have private information. They coordinate their local plans in order to find a joint global plan which minimizes total cost. The individual plans are coordinated by a collaborative planning approach based on simulated annealing. In addition to some well-known techniques a production responsibility assignment procedure and Shapley compensation payments computed by means of a modified characteristic function are incorporated. The proposed solution approach outperforms a reference value in 206 out of 272 new DULR instances. The approach is more efficient on small instances than on medium sized instances.

Shipper decision support for the acceptance of bids during the procurement of transport services

Combinatorial reverse auctions can be used by shippers in order to procure transportation services from carriers. After carriers have submitted their bids, a shipper has to decide about the allocation of transport services to carriers, i. e., a shipper has to solve the winner determination problem of the auction. This paper focuses on a bicriteria winner determination problem in which a shipper has to select a subset of the set of bids and simultaneously decide about the desired trade-off between total transportation costs and the quality of the entire transportation services. To solve this bicriteria optimization problem a metaheuristic is developed that computes a set of non-dominated solutions based on the concepts of multi-start, large neighborhood search and a bicriteria branch-and-bound procedure. Compared to previous results in the literature, the proposed algorithm is able to improve the set of non-dominated solutions for 14 out of 30 benchmark instances.

Ein Pareto-Optimierungsverfahren für ein mehrkriterielles Gewinnerermittlungsproblem in einer kombinatorischen Transportausschreibung

This paper extends the NP-hard Winner Determination Problem of Combinatorial Reverse Auctions in a scenario based on the procurement of transportation services. In addition to the common goal of minimizing the total cost of all selected bids, a second objective aims at maximizing the expected carrier performance in the transportation network. Without any a priori preference information a GRASP-based algorithm uses the notion of Pareto dominance to take care of both objectives simultaneously. Variants of this procedure are evaluated by means of 810 generated test instances. Those procedures which consider both objectives during construction stage but only one objective in the local search stage offer the best performance in terms of solution quality, robustness and computing time.