Optimization of construction safety resource allocation based on evolutionary game and genetic algorithm

In the construction industry, ensuring the safety performance of a project relies heavily on the effective allocation of safety resources. As the importance of mental health in the construction industry increases, evolutionary game theory can be used to analyze the interaction mechanism of various factors affecting safety performance during the construction phase. The objective of this paper is to construct an analytical model that combines evolutionary game theory with genetic algorithms from the perspective of Leader-Member Exchange Ambivalence. The model aims to quantify and compare the various factors that influence achieving the expected safety state and identify the specific necessary constraints. Initially, we analyzed the relationships among construction site employees, divided them into superiors and subordinates, and established a game model and payoff matrix based on the research background. Next, we introduced genetic algorithms into the model via the replicator dynamic equation for optimization. We adjusted the coefficients of safety risk level, psychological expected return, moral identity, and other factors to simulate various construction site scenarios. Simulation and optimization results indicate that genetic algorithms provide more accurate reference values for safety resource allocation compared to preset or manually assigned values.