A new method of neural network system for monitoring of asphalt mixtures compaction

Prokopev Andrey Petrovich, Nabizhanov Zhasurbek Ilkhomovich, Emelyanov Rurik Timofeevich, Ivanchura Vladimir Ivanovich

The article considers the task of non-destructive compaction control technology automation of road materials by pavers in real time. A new method is proposed based on the prediction of the compaction coefficient of road materials by an artificial neural network, the input data of which is determined by analyzing the vibration and force state of the sealing working bodies of the compacting equipment. The inputs of the artificial neural network are the type of asphalt mixture (AM), the velocity of the paver, the force in the tamper pusher, the frequency of the tamper, the layer thickness. The theoretical prerequisites of the method of continuous compaction control by pavers are determined.

Keywords:non-destructive technologies, compaction control, artificial neural networks, cyber-physical systems, road construction.