Application of Bayesian network and genetic algorithm enhanced Monte Carlo method in urban transportation infrastructure investment

Traditional urban transportation infrastructure investment evaluation methods cannot fully consider future uncertainties and complex environmental factors, which affect resource allocation and decision-making accuracy. In order to optimize investment decision-making, this research aims to integrate Bayesian networks, genetic algorithms, and Monte Carlo methods to propose a more flexible and accurate investment evaluation method. By combining conditional probability modeling, optimization function, and randomness simulation, we strive to realize the scientific and rational decision-making of urban transportation infrastructure investment. In this paper, a Bayesian network of transportation infrastructure investment risk is constructed. Variable nodes and logical relationships are accurately defined, and a conditional probability table is formed to model uncertainty. The investment portfolio is optimized by genetic algorithm, and the fitness function and iterative selection are set to screen out the optimal solution. Monte Carlo simulation technology is used to verify the robustness of the optimization solution and conduct multiple random experiments. Combining these three, an optimization model is formed that can efficiently evaluate investment decisions and improve the accuracy and adaptability of results. The experimental results show that at a 95% confidence level, the VaR (value at risk), and CVaR (conditional value at risk) values of the vehicle-kilometer service fee method are 32.4512 million yuan and 89.1234 million yuan, respectively, which are significantly better than the 410.2392 million yuan and 691.5735 million yuan of the agreement ticket price law. At the same time, the standard deviation rate of the IRR (internal rate of return) of the vehicle-kilometer service fee method is 0.1548, and the probability that the IRR is not lower than the social average return is 95.49%. The potential return is higher; the risk management and stability are better; it is more suitable for decision makers who pursue stable returns. The study shows that the enhanced Monte Carlo method in this paper not only provides effective theoretical support and technical means for the investment evaluation of urban transportation infrastructure.