Application of Evolutionary Algorithms in Project Management

Abstract

The paper deals with "resource leveling optimization problems", a class of problems that are often met in modern project management. The problems of this kind refer to the optimal handling of available resources in a candidate project and have emerged, as the result of the even increasing needs of project managers in facing project complexity, controlling related budgeting and finances and managing the construction production line. For the effective resource leveling optimization in problem analysis, evolutionary intelligent methodologies are proposed. Traditional approaches, such as exhaustive or greedy search methodologies, often fail to provide near-optimum solutions in a short amount of time, whereas the proposed intelligent approaches manage to quickly reach high quality near-optimal solutions. In this paper, a new genetic algorithm is proposed for the investigation of the start time of the non-critical activities of a project, in order to optimally allocate its resources. Experiments with small and medium size benchmark problems taken from publicly available project data resources, produce highly accurate resource profiles. The proposed methodology proves capable of coping with larger size project management problems, where conventional techniques like complete enumeration is impossible, obtaining near-optimal solutions.

Application of Genetic Algorithms for Decision-Making in Project Management: A Literature Review

“…All identified and analysed limitations of genetic algorithms can be divided into two categories: technic-specific limitations and ones defined by project tasksall identified limitations described in this section relate to relevant problems of the knowledge area. [12], [14], [16], [18]- [25], [27], [29], [31]- [33], [35], [36], [38]- [45] When solving Project, Resource, Team, Activities or Tasks: Planning and Scheduling problems with genetic algorithm, most of the limitations, which are possible to identify, are taskbased genetic algorithm limitations.…”

“…27 GA model-based limitations include: (1) the algorithm not being tested on real data [29], [31]; (2) underestimated implementation potential for use (real problem solving is not discussed, but the described experiment is based on realistic data; the results show that 100 % positive evaluation is not achieved) [29], [31]; (3) a lack of historical database, which describes the previous algorithm activity [28]; (4) suitability of the algorithm for one specific project and a lack of customisation as a universal solution (there are different assumptions and dependencies in each project, which may not be required in other cases/projects) [28]; (5) a lack of indication of necessary time and resources to solve problems [27]; (6) a lack of evaluation of the implementation potential for use [27].…”

“…• The maximum number of iterations/generations [10], [21], [27], [30]; • Mutation rate and crossover rate limitations [10], [21], [22], [27]; • The number of population limitations [21], [27], [30];…”

“…In this case, the GA chromosome encoding strategies are project-specific input data, for example, priorities [10]; delay time [10], [28] and release dates [10]; tasks [21], [22], [29], [30]; project-specific rules [15], [27], [28]; employee DB, employees' knowledge and competences [15], [29]; human resources [22], [27], [28], [29]; scheduled activities and activity lists [5], [31], [32].…”

“…The most applied types of chromosome encoding approaches in the genetic algorithm are graphs, nets or tree nodes [22], [27], [30], [31] or integer number string [5], [28], [32]. Authors also use chromosome encoding approaches [10], Island-based encoding method [15] and binary code [29].…”