الگوریتم بهینه سازی کلونی مورچه بهبود یافته برای مشکلات تسطیح منابع غیر خطی

The notion of using a meta-heuristic approach to solve nonlinear resource-leveling problems has been intensively studied in recent years. Premature convergence and poor exploitation are the main obstacles for the heuristic algorithms. Analyzing the characteristics of the project topology network, this paper introduces a directional ant colony optimization (DACO) algorithm for solving nonlinear resource-leveling problems. The DACO algorithm introduced can efficiently improve the convergence rate and the quality of solution for real-project scheduling.

Given the finite nature of resource availability, a project scheduling may have to be modified so that the project can be successfully performed. Resource-leveling optimization as a project scheduling management technique has been widely studied in the construction and manufacturing industries for solving nonlinear resource allocation problems by searching for the best start time for each activity. Many analytical and heuristic models have been developed for solving nonlinear resource-leveling optimization problems. In computational biology, meta-heuristics such as genetic algorithms, particle swarm optimization approaches and taboo searches have been used for solving such nonlinear resource-leveling problems. For example, Leu [1] suggested a genetic algorithm optimization for tackling the resource-leveling issue in construction; Roca [2] proposed a multi-objective genetic algorithm-based solver for optimizing the extended resource-leveling problem. However, the premature convergence and poor exploitation, which are the main drawbacks of meta-heuristics, have attracted increasing attention from researchers and engineers. This paper employs a directional ant colony optimization (DACO) approach to solve nonlinear resource-leveling problems. The activity-on-node-based DACO technique is effective and efficient in dealing with premature convergence or poor exploitation, and it has an advantage of not translating the real data into code, as compared with genetic algorithms. Simple random exploitation is carried out in the ACO approach, while the DACO algorithm is designed to search for a promising path in the area considered. It is a directional search that combines a globally optimized trail, the local best path and random exploitation for resource-leveling optimization. The original idea of ACO was proposed by Dorigo [3] and [4] for searching for the optimal path in a graph.

A directional ant colony optimization (DACO) algorithm is introduced in this paper for searching for the best starting time in non-critical path activities and establishing a nonlinear resource leveling. In the DACO algorithm, a swarm of ants search for the best path node by node along one direction, which can improve the performance of the traditional ACO approach. The DACO algorithm can not only perform global exploration efficiently and rapidly, but also improve the solution quality greatly. Further studies are needed to enhance the ability of the DACO approach to solve more complex resource-constrained project scheduling problems.