Modeling the Quantitative Assessment of the Condition of Bridge Components Made of Reinforced Concrete Using ANN

Bridges in Ukraine are one of the most important components infrastructure, requiring attention from government agencies and constant funding. The object study was methodology for quantifying condition bridge components. Artificial Neural Network-based (ANN) tool developed to quantify technical literature analysis showed that cases datasets were obtained during inspection bridges solve problems assessing current condition. lack such a database prompted creation dataset on basis Classification Tables Operating Conditions Bridge Components (CT). Based CTs, five formed assess components: span, deck, pier caps beam, piers abutments, approaches. next step this creating, training, validating testing ANN models. network with ADAM loss function softmax activation best results. optimal values MAPE R2 achieved at 100th epoch 64 neurons hidden layer equal 0.1% 0.99998, respectively. practical application models carried out common type Ukraine, namely, road beam small length, made precast concrete. novelty consists development based use model, proposal modify This will allow minimizing uncertainties associated subjective judgments experts, as well increasing accuracy assessment.