Research on Driver’s Lane Change Intention Recognition Method Based on Principal Component Analysis and GMM-HMM 2022-01-7021
Aiming at the problems of long lane change intention recognition, complicated lane change model, and huge amount of processing data in the current research, this paper uses principal component analysis to improve the driver’s lane change intention recognition model using traditional pattern recognition. Firstly collect 7 parameters including driver operation and vehicle running characteristics. After data standardization and PCA (principal component analysis), the top three principal components that can reflect the information content of the original data are nearly 90%. Then, a lane-change intent recognition model based on GMM-HMM was established, three lane change intents cannot be directly observed as the hidden state of the model; and three principal component quantities obtained through linear changes are used as observational measurements. In the specific implementation process, the seven parameters are first transformed into three principal components through linear changes, and are captured using the sliding time window method, and then the forward probability of three lane change intentions is calculated by using the forward-backward algorithm. Finally, the experiment set was used to verify the accuracy of the model recognition. The results show that the method can be used to identify the intention of changing lanes, and the recognition rate can reach nearly 100% 1.5s after the starting point of changing lanes.
Citation: Shen, C., Li, S., Shi, B., Yu, J. et al., "Research on Driver’s Lane Change Intention Recognition Method Based on Principal Component Analysis and GMM-HMM," SAE Technical Paper 2022-01-7021, 2022, https://doi.org/10.4271/2022-01-7021. Download Citation