Computing Truck Attributes with Artificial Neural Networks

Abstract
The paper describes the application of neural networks to the problem of determining truck attributes (such as velocity, axle spacings, and axle loads) purely from strain‐response readings taken from the structure over which the truck is traveling. The approach is designed to remove both the need for tape switches on the deck of the bridge to obtain such data and associated problems so as to provide a convenient and viable means of collecting bridge loading statistics. The application and performance of a radial‐Gaussian‐based networking system with its own training algorithm to the truck‐attribute determination problem is detailed. The chosen approach is a two‐layered modular network structure. The artificial neural network in the first layer classifies the trucks, and the second‐layer artificial neural networks—three for each type of truck—compute estimates of the truck's velocity, axle spacings, and axle loads. This solution provides a fast, accurate, and convenient means of determining truck attributes.

This publication has 3 references indexed in Scilit: