Detection of steering events based on vehicle logging data using hidden Markov models

Journal article, 2016

In vehicle design it is desirable to model the loads by describing load environment, customer usage and vehicle dynamics. In this study a method will be proposed for detection of steering events such as curves and manoeuvring using on-board logging signals available on trucks. The method is based on hidden Markov models (HMMs), which are probabilistic models that can be used to recognise patterns in time series data. In an HMM, 'hidden' refers to a Markov chain where the states are not observable. However, observations depending on the hidden Markov chain can be observed. The idea here is to consider the current driving event as the hidden state, while the on-board logging signals generate the observed sequence. Examples of curve detection are presented for both simulated and measured data on a truck. The classification results indicate that the method can recognise left and right turns with small misclassification errors.