Pavement deterioration models, being essential elements of Pavement Management System, are in widespread use. The accuracy of predictions yielded by models determines the dependability of the analysis procedures. For a model to satisfy the boundary conditions, logical constraints, or accuracy requirements; a complex form of the equation may be inevitable. Complex equation forms of the models present their own set of difficulties in the determination of solutions. Traditionally adopted methods to solve such equations often suffer from disadvantages of encountering divergence, need for an exhaustive search of initial values, or being stuck in a local minimum. Simulated Annealing (SA), a stochastic method, able to overcome these disadvantages in solving equations is introduced in this paper. The solutions for a pavement deterioration equation display a high degree of accuracy of predictions suggesting the suitability of Simulated Annealing algorithm as a tool for the development of solution of pavement deterioration models. This paper also presents some of the tips that may be adopted for easier determination of solutions.
展开▼
