Model of justification of project portfolio characteristics based on binary optimization tools

Modern conditions of development of the construction industry are characterized by tightening requirements for the duration and timing of the construction of residential and commercial projects. The formation of a portfolio of construction projects implemented within certain time intervals is one of the most important tasks solved within specialized (construction) organizations to ensure the effectiveness of the implementation of both the relevant technological processes and the processes of developing the infrastructure of large settlements, which has a significant impact on the economic indicators of individual territorial regions. The purpose of the research is to develop tools for solving the problem of justifying the characteristics of a portfolio of construction projects, implemented by a specialized organization over a certain time period. The subject of the research is the characteristics of the above-mentioned project portfolio in terms of indicators of the feasibility of implementing (inclusion in the portfolio) individual projects, the duration and cost of implementing the project portfolio. During the research process the methods of system analysis and statistical processing of information were used, as well as integer linear optimization considering the requirements of binarity for the values of unknown variables. A review and analysis of scientific developments in the field of substantiating the characteristics of a portfolio of construction projects implemented by specialized organizations was carried out, during which shortcomings in existing methods were identified: the lack of a formalized description of tools and the presence of low practical significance due to the lack of consideration of features characteristic of the construction industry. During the research process an optimization model has been developed to determine the characteristics of a portfolio of construction projects implemented by a specialized organization over a certain time period; the model has a linear structure, contains binary constraints for the values of unknown variables, and therefore can be implemented with the use of the branch-and-bound method available in modern optimization modeling software environments. The developed optimization model was implemented using a practical example; based on the results obtained, a conclusion has been made about the high practical significance of the developed tool.