A computational framework for resource allocation and reliability assessment in road networks under stochastic link disruptions

Stochastic link disruptions can substantially reduce the reliability and service performance of road networks by interrupting critical connections, increasing travel costs, and weakening accessibility between origins and destinations. To support the computational analysis of protection resource allocation under uncertainty, this paper develops a parametric framework that integrates resource allocation, stochastic link failure, and multidimensional network-performance evaluation. Link scores derived from network topology are incorporated into a Softmax-type allocation rule featuring an adjustable preference parameter. The allocated resources are then coupled with a probabilistic survival model to generate disrupted network states via Monte Carlo simulation. Network performance is assessed using three metrics: average accessibility, average travel time, and the proportion of the largest connected component, along with a disruption-threshold resilience curve (DTRC) for comparing strategies under progressively severe disruptions. A numerical experiment based on a simplified real-world road network demonstrates that accessibility and travel efficiency are more sensitive than structural connectivity to both allocation preference and disruption intensity. The proposed framework offers an interpretable basis for comparing resource-allocation strategies in stochastic road-network reliability assessment.