INTELLIGENT MODELS FOR MANAGING COMMUNITY INFRASTRUCTURE DEVELOPMENT PROJECTS UNDER MULTILEVEL RISKS

Abstract

Introduction. The article substantiates the relevance of managing community infrastructure development projects in the context of multilevel risks. Three types of infrastructure projects (security, medical, and energy) are considered. Based on the analysis of modern approaches in science and practice, it is established that existing management models are mostly narrowly focused, do not take into account the dynamic project environment, which causes multi-level risks, and do not provide adaptability to the project environment, the need for rapid changes in projects. Given the important role of timely response and efficient use of limited resources, the management of community development infrastructure projects requires a multisectoral approach. It integrates intelligent methods, risk forecasting, and prioritization of project implementation scenarios. In a dynamic project environment, it is especially important to develop value-based solutions based on the use of computational intelligence and multi-criteria optimization to ensure sustainable community development. 
Purpose. To substantiate the approach and develop an intelligent multisectoral model of project management for the development of security, medical and energy infrastructures of communities, taking into account multilevel risks, which allows to increase the efficiency of management decisions, ensure adaptive resource planning, integrate modern artificial intelligence technologies into the process of strategic development of territories, minimize the consequences of undesirable events and improve the sustainability of infrastructure systems in a dynamic project environment with limited. 
Methods. To achieve the stated research objective, a comprehensive approach was applied, combining modern methods of analysis, modeling, and optimization, along with the use of artificial intelligence tools and geoinformation technologies. This approach enables the consideration of the multicomponent structure of community infrastructure and ensures cross-sectoral integration within a single project management model focused on security, healthcare, and energy development in local communities. One of the key methodological foundations is the use of systems analysis, which allows for identifying interconnections among various components of infrastructure development projects, assessing their sensitivity to external and internal risks, and determining the priority of management interventions. This approach provides a holistic view of governance processes within a community and enables the identification of critical influences by focusing attention on the most significant elements. To construct models of interaction between infrastructure sectors, cognitive modeling was applied. This method makes it possible to visualize cause-and-effect relationships, evaluate system dynamics, and forecast potential development scenarios. The research also proposes the use of computational intelligence methods, including neural network modeling, fuzzy logic, and situational decision-making algorithms. These tools are employed to develop models that adapt projects to changes in the project environment, detect anomalous infrastructure development scenarios, and generate recommendations for future management actions under conditions of uncertainty. 
Results. A conceptual model of intelligent multisectoral management of infrastructure projects (focused on security, healthcare, and energy) at the community level has been developed. The proposed model consists of five interrelated levels: 1) the information environment; 2) the project management office (PMO); 3) the portfolio of infrastructure projects (security, medical, and energy) at the community level; 4) the performance/output layer (infrastructure project deliverables); 5) value creation. The model is multisectoral, as it integrates the security, healthcare, and energy components of community infrastructure within a unified strategic space. It incorporates a multisectoral risk management subsystem, which is based on constructing risk matrices that take into account both the likelihood of threats and the severity of their consequences. Additionally, it includes a decision-making subsystem, which applies a function for determining the integral project value. 
The value of infrastructure projects is calculated based on five key parameters: risk level, implementation duration, resource availability, accessibility of medical services, and energy autonomy. The model also considers the specific priorities of five stakeholder groups - community residents, the state, local administrations, specialists, and project managers - ensuring a well-grounded prioritization of projects. 
Conclusions and specific proposals of the author. Based on the proposed conceptual model, a computer-based implementation was developed, enabling scenario modeling using Python tools. The model was tested on a hypothetical community, which allowed for the evaluation of the integral value of each type of project under three implementation scenarios. In the baseline scenario, the integral value of the healthcare development project was 0.585, the energy project — 0.570, and the security project — 0.609. Under the optimistic scenario, the highest value was observed for the security project (0.828), followed by the healthcare (0.798) and energy projects (0.778). The pessimistic scenario showed a decrease in project value indicators to 0.419 for security, 0.409 for healthcare, and 0.393 for energy. These results confirm the relevance of implementing dynamic models for managing community infrastructure development projects that are sensitive to changes in the project environment. The proposed model enables the consideration of dynamic shifts, supports the justification of priority project selection, and facilitates the adaptation of the infrastructure development portfolio in accordance with resource and risk constraints. 
Further application of the model is advisable for evaluating infrastructure projects in real communities, taking into account the spatial, demographic, and security characteristics of the project environment.