Research on the Driving Capability Evaluation Method Basing on the Parallel Driving System for Automated Vehicles

In order to solve the time cost and safety issues brought by testing the driving capability of drivers in real scenarios, this paper develops a hardware in the loop virtual simulation platform based on the parallel driving system between humans, vehicles, and roads, and realizes the evaluation of the driving capability of virtual drivers. Based on BP neural network, we establish a virtual driver driving capability evaluation model. Combining virtual environment and hardware in the loop system for verification, the experimental results show that the virtual vehicle controlled by the virtual driver model and the real vehicle have synchronization, which verifies the effectiveness of the parallel driving system among humans, vehicles, and roads. The virtual driver driving capability evaluation model has high identification accuracy and can accurately output the results of virtual driver driving capability evaluation.