Nicoleta Neagu

Exploiting Interchangeabilities for Case Adaptation

While there are many general methods for case retrieval, case adaptation usually requires problem-specific knowledge and it is still an open problem. In this paper we propose a general method for solving case adaptation problems for the large class of problems which can be formulated as Constraint Satisfaction Problems. This method is based on the concept of interchangeability between values in problem solutions. The method is able to determine how change propagates in a solution set and generate a minimal set of choices which need to be changed to adapt an existing solution to a new problem. The paper presents the proposed method, algorithms and test results for a resource allocation domain.

LS/ATN: Reporting on a Successful Agent-Based Solution for Transport Logistics Optimization

A considerable volume of research exists concerning the domain of automatic planning and scheduling, hut many real-world scheduling problems, and especially that of transportation logistics, remain difficult to solve. In particular, this domain demands schedule-solving for every vehicle in a transportation fleet where pick-up and delivery of customer orders is distributed across multiple geographic locations, while satisfying time-window constraints on pickup and delivery per location. This paper presents a successful commercial-grade solution to this problem called living systems adaptive transportation networks (LS/ATN), which has been proven through real-world deployment to reduce transportation costs through the optimization of route solving for both small and large fleets. LS/ATN is a novel agent-based resource management and decision system designed to address this highly dynamic and complex domain in commercial settings. We show how LS/ATN employs agent cooperation algorithms to derive truck schedules that optimize the use of available resources leading to significant cost savings. The solution is designed to support, rather than replace, the day-to-day activities of human dispatchers

An incremental approach using local-search heuristic for inventory routing problem in industrial gases

Abstract In this paper we solve the inventory routing problem (IRP) occurring in industrial gas distribution where liquefied industrial gases are distributed to customers that have cryogenic tanks to store the gases on-site. We consider a multi-period inventory routing problem with multiple products assuming deterministic demand rates and the proposed model is formulated as a linear mixed-integer program. We propose an incremental approach based on decomposing the set of customers in the original problem into sub-problems. The smallest sub-problem consists of the customer that needs to be delivered most urgently along with a set of its neighbors. We solve each sub-problem with the number of customers growing successively by providing the solution of the previously solved sub-problem as an input. Each sub-problem is then solved with a randomized local-search heuristic method. We also propose an objective function that drives the local-search heuristics toward a long-term optimal solution. The main purpose of this paper is to develop a solution methodology appropriate for large-scale real-life problem instances particularly in industrial gas distribution.

Studying Interchangeability in Constraint Satisfaction Problems

Most work in constraint satisfaction has concentrated on computing a solution to a given problem. In practice, it often happens that an existing solution needs to be modified to satisfy additional criteria or changes in the problem. For example, a schedule or plan might have to be adjusted when a resource is missing. The concept of interchangeability characterizes the possibilities for making local changes to CSP solutions.

Experimental Evaluation of Interchangeability in Soft CSPs

In [8], Freuder defined interchangeability for classical Constraint Satisfaction Problems (CSPs). Recently [2], we extended the definition of interchangeability to Soft CSPs and we introduced two notions of relaxation based on degradation δ and on threshold α ( δ neighborhood interchangeability ( δ NI)and α neighborhood interchangeability ( α NI)). In this paper we extend the study introduced in [11] and we analyze the presence of the relaxed version of interchangeability in random soft CSPs. We give a short description of the implementation we used to compute interchangeabilities and to make the tests. The experiments show that there is high occurrence of α NI and δ NI interchangeability around optimal solution in fuzzy CSPs and weighted CSPs. Thus, these algorithms can be used successfully in solution update applications. Moreover, it is also showed that NI interchangeability can well approximate full interchangeability (FI).

Interchangeability in Soft CSPs

Substitutability and interchangeability in constraint satisfaction problems (CSPs) have been used as a basis for search heuristics, solution adaptation and abstraction techniques. In this paper, we consider how the same concepts can be extended to soft constraint satisfaction problems (SCSPs). We introduce two notions: threshold α and degradation δ for substitutability and interchangeability, (αsubstitutability/interchangeability and δsubstitutability/interchangeability respectively). We show that they satisfy analogous theorems to the ones already known for hard constraints. In αinterchangeability, values are interchangeable in any solution that is better than a threshold α, thus allowing to disregard differences among solutions that are not sufficiently good anyway. In δinterchangeability, values are interchangeable if their exchange could not degrade the solution by more than a factor of δ. We give efficient algorithms to compute (δ/α)interchangeable sets of values for a large class of SCSPs.

Interchangeability with thresholds and degradation factors for Soft CSPs

Substitutability and interchangeability in constraint satisfaction problems (CSPs) have been used as a basis for search heuristics, solution adaptation and abstraction techniques. In this paper, we consider how the same concepts can be extended to soft constraint satisfaction problems (SCSPs). We introduce two notions: thresholdα and degradation factorδ for substitutability and interchangeability, (αsubstitutability/interchangeability and δsubstitutability/interchangeabi-lity respectively). We show that they satisfy analogous theorems to the ones already known for hard constraints. In αinterchangeability, values are interchangeable in any solution that is better than a threshold α, thus allowing to disregard differences among solutions that are not sufficiently good anyway. In δinterchangeability, values are interchangeable if their exchange could not degrade the solution by more than a factor of δ. We give efficient algorithms to compute (δ/α)interchangeable sets of values for a large class of SCSPs, and show an example of their application. Through experimental evaluation based on random generated problem we measure first, how often neighborhood interchangeable values are occurring, second, how well they can approximate fully interchangeable ones, and third, how efficient they are when used as preprocessing techniques for branch and bound search.

A Decomposition Approach to Solve Large-Scale Network Design Problems in Cylinder Gas Distribution

The logistics network has to be optimally designed for an effective supply chain. The focus of this research is to solve network design problem occurring in packaged gases (e.g., cylinder) supply chain. The integrated logistics network design problem for packaged gases is defined as follows: given a set of potential locations for filling plants and hubs, and customers with deterministic demands, determine the configuration of the production-distribution system i.e., optimal facility locations, the filling plant production capacities, the inventory at plants and hubs, and the number of packages to be routed in primary and secondary transportation. The problem is modeled as a deterministic mixed integer program and a decomposition approach is developed which allows a natural split of the production and distribution decisions. The proposed framework is illustrated with numerical examples from real-life packaged gases supply chain. The results show that the decomposition approach is effective in solving a broad range of problem sizes. The results from the decomposition approach are benchmarked by solving optimally the whole packaged gases network design model for smaller test cases. In the end, we perform sensitivity analysis for parameters that are likely to change in the future for better understanding of their impact.

Integrated Supply Chain Network Design for Packaged Gases

Network design of the supply chain is an important and strategic aspect of logistics management. In this paper, we address the network design problem specific to packaged gases (e.g., cylinder) supply chain. We propose an integrated framework that allows for the determination of the optimal facility locations, the filling plant production capacities, the inventory at plants and hubs, and the number of packages to be routed in primary and secondary transportation. We formulate the problem as a mixed integer program and then develop a decomposition approach to solve it. We illustrate the proposed framework with numerical examples from real-life packaged gases supply chain. The results show that the decomposition approach is effective in solving a broad range of problem sizes. We also benchmark the results from the decomposition approach by solving the complete packaged gases network design model for smaller test cases.

Exploiting the CSP Structure by Interchangeability

While there are many AI algorithms designed for finding solutions in a Constraint Satisfaction Problems, finding similar solutions of a CSP requires entirely new and different methods. The method we propose here is based on the interchangeability concept first introduced by Freuder [1].