Algirdas Lančinskas

Solution of multi-objective competitive facility location problems using parallel NSGA-II on large scale computing systems

The multi-objective firm expansion problem on competitive facility location model, and an evolutionary algorithm suitable to solve multi-objective optimization problems are reviewed in the paper. Several strategies to parallelize the algorithm utilizing both the distributed and shared memory parallel programing models are presented. Results of experimental investigation carried out by solving the competitive facility location problem using up to 2048 processing units are presented and discussed.

Approaches to parallelize pareto ranking in NSGA-II algorithm

In this paper several new approaches to parallelize multi-objective optimization algorithm NSGA-II are proposed, theoretically justified and experimentally evaluated. The proposed strategies are based on the optimization and parallelization of the Pareto ranking part of the algorithm NSGA-II. The speed-up of the proposed strategies have been experimentally investigated and compared with each other as well as with other frequently used strategy on up to 64 processors.

A preference-based multi-objective evolutionary algorithm R-NSGA-II with stochastic local search

Incorporation of a decision maker’s preferences into multi-objective evolutionary algorithms has become a relevant trend during the last decade, and several preference-based evolutionary algorithms have been proposed in the literature. Our research is focused on improvement of a well-known preference-based evolutionary algorithm R-NSGA-II by incorporating a local search strategy based on a single agent stochastic approach. The proposed memetic algorithm has been experimentally evaluated by solving a set of well-known multi-objective optimization benchmark problems. It has been experimentally shown that incorporation of the local search strategy has a positive impact to the quality of the algorithm in the sense of the precision and distribution evenness of approximation.

Parallel Multi-objective Memetic Algorithm for Competitive Facility Location

A hybrid genetic algorithm for global multi-objective optimization is parallelized and applied to solve competitive facility location problems. The impact of usage of the local search on the performance of the parallel algorithm has been investigated. An asynchronous version of the parallel genetic algorithm with the local search has been proposed and investigated by solving competitive facility location problem utilizing hybrid distributed and shared memory parallel programming model on high performance computing system.

Local optimization in global multi-objective optimization algorithms

A hybrid Multi-Objective Optimization Algorithm based on the NSGA-II algorithm is presented and evaluated. The local optimization algorithm called SASS has been modified in order to be suitable for multi-objective optimization where the local optimization is intended towards non-dominated points. The modified local optimization algorithm has been incorporated into NSGA-II in order to improve performance.

Investigation of parallel particle swarm optimization algorithm with reduction of the search area

We consider a population based Particle Swarm Optimization (PSO) algorithm and a few modifications to increase quality of optimization. Several strategies are investigated to exchange data between processors in parallel algorithm. Experimental investigation is performed on Multiple Gravity Assist problem. The results are compared with original PSO.

Parallel Optimization Algorithm for Competitive Facility Location

A stochastic search optimization algorithm is developed and applied to solve a bi-objective competitive facility location problem for firm expansion. Parallel versions of the developed algorithm for shared- and distributed-memory parallel computing systems are proposed and experimentally investigated by approximating the Pareto front of the competitive facility location problem of different scope. It is shown that the developed algorithm has advantages against its precursor in the sense of the precision of approximation. It is also shown that the proposed parallel versions of the algorithm have almost linear speed-up when solving competitive facility location problems of different scope reasonable for practical applications.

Effect of Diffusion Limitations on Multianalyte Determination from Biased Biosensor Response

The optimization-based quantitative determination of multianalyte concentrations from biased biosensor responses is investigated under internal and external diffusion-limited conditions. A computational model of a biocatalytic amperometric biosensor utilizing a mono-enzyme-catalyzed (nonspecific) competitive conversion of two substrates was used to generate pseudo-experimental responses to mixtures of compounds. The influence of possible perturbations of the biosensor signal, due to a white noise- and temperature-induced trend, on the precision of the concentration determination has been investigated for different configurations of the biosensor operation. The optimization method was found to be suitable and accurate enough for the quantitative determination of the concentrations of the compounds from a given biosensor transient response. The computational experiments showed a complex dependence of the precision of the concentration estimation on the relative thickness of the outer diffusion layer, as well as on whether the biosensor operates under diffusion- or kinetics-limited conditions. When the biosensor response is affected by the induced exponential trend, the duration of the biosensor action can be optimized for increasing the accuracy of the quantitative analysis.

The Huff Versus the Pareto-Huff Customer Choice Rules in a Discrete Competitive Location Model

An entering firm wants to compete for market share in a given area by opening some new facilities selected among a finite set of potential locations. Customers are spatially separated and other firms are already operating in that area. In this paper, we analyse the effect of two different customers’ behavior over the optimal solutions, the Huff and the Pareto-Huff customer choice rules. In the first, the customer splits its demand among all competing facilities according to its attractions. In the second, the customer splits its demand among the facilities that are Pareto optimal with respect to the attraction (to be maximized) and the distance (to be minimized), proportionally to their attractions. So, a competitive facility location problem on discrete space is considered in which an entering firm wants to locate a fixed number of new facilities for market share maximization when both Huff and Pareto-Huff customer behavior are used. In order to solve these two models, a heuristic procedure is proposed to obtain the best solutions, and it is compared with a classical genetic algorithm for a set of real geographical coordinates and population data of municipalities in Spain.

Single and multi-objective genetic algorithms for the container loading problem

Container Loading Problems (CLPs) deal with determination of the optimal pattern for packing boxes into a given container usually with respect to the maximal utilization of the total container volume. On the other hand, it is also important to maximize the utilization of the maximal container weight for which is paid when buying a shipment service. In this paper we analyze two genetic algorithms specially adopted to solve CLP. One of them is based on the Genetic Algorithm (GA) and is suitable to solve single-objective CLPs, while another one is based on the Non-dominated Sorting Genetic Algorithm (NSGA-II), suitable for solution of CLP by simultaneously considering both of the above mentioned objectives. The algorithms have been experimentally investigated by solving various CLP instances of different complexity. The obtained results showed that simultaneous consideration of both objectives using the proposed multi-objective optimization algorithm gives better results in utilization of container volume when solving complex CLP instances.