The stage of operation of a highway is one of the most important stages of its life cycle. The decisions made at this stage have a direct impact on the durability of the road, the safety and transport costs of its users. Optimization of the decisions made is an urgent task that requires a comprehensive consideration of both the technical and economic components. Within the framework of this work, using the algorithms of artificial neural networks (ANN), a mathematical model of the ANN was developed to determine the integral level of pavement safety, based on a complex of structural and operational factors, including the modulus of elasticity of pavement layers, longitudinal evenness, rutting, and the presence of fatigue failures. On the basis of the integral level of pavement safety, its operational condition is predicted, a scale is assigned for the selection of control actions to ensure its required durability.
Keywords: artificial neural networks, integral safety level, falling weight deflectometer, road maintenance, road life cycle management
The possibility of using artificial neural networks to assess the current characteristics of pavements and their potential application in the development of road maintenance strategies is considered. The results of the models showed the convergence between the estimated values of the pavement condition and the actual values at all stages of training. The results show that public road authorities can use the developed models to determine the optimal approach to road maintenance and determine the most effective measures to restore their capacity and operational condition.
Keywords: artificial neural networks, backpropagation algorithm, falling weight deflectometer, pavement maintenance, pavement management system