Determination of brake force using artificial neural network

Abstract

Brakeforcesfordifferentconditionshavebeenusedtocreateanartificialneuralnetwork(ANN) model. Amathematical formulation is obtained by considering effective factors of brake force (slip, tread depth, road surface, turn angle, tread profile, tire pressure and Toe of the wheel). There is a good agreement between experimental data and ANN model. The best approach with minimum errors is performed using LM algorithm with 4 neurons on hidden layers. Statistical coefficient of multiple determinations (R² value) is obtained as 0,999699 for training data and 0,999178 for testing data.