Improved Driving Behaviors Prediction Based on Fuzzy Logic-Hidden Markov Model (FL-HMM)

Abstract

Research and development of human driving behaviors play an important role in the development of assistance systems. In this contribution, a driving behaviors prediction model is based on a newly developed approach combining different Hidden Markov Models (HMM) cooperatively combined by Fuzzy Logic (FL). Due to variations of individual human drivers decision behavior the task to classify related behaviors based on individually trained models is difficult. The FL approach will be used for additional distinction of driving scenes into very safe, safe, and dangerous driving scenarios. For each scenario corresponding HMMs will be trained. Three different driving behaviors including left/right lane change and lane keeping are modelled as hidden states for the HMM. Based on observations, the algorithm calculates the most possible driving behaviors through the observation sequences. Furthermore, the observed sequences are also used for training of HMM during modeling process. To improve the prediction performance of the model, a prefilter is proposed to quantize the collected signals into observed sequences with specific features. To optimize the model performance NSGA-II was used to define the optimal thresholds of FL and the optimal prefilters of HMMs. Using experimental data from real human driving behaviors (taken from driving simulator) it can be concluded that selecting optimal thresholds will increase the performance of driving behaviors prediction. The effectiveness of the suggested fuzzy-based HMM has been successfully proved based on experiments.