Login / Signup

Vehicle Driving Risk Prediction Model by Reverse Artificial Intelligence Neural Network.

Huizhe DingRaja Ariffin Raja GhazillaRamesh Singh Kuldip SinghLina Wei
Published in: Computational intelligence and neuroscience (2022)
The popularity of private cars has brought great convenience to citizens' travel. However, the number of private cars in society is increasing yearly, and the traffic pressure on the road is also increasing. The number of traffic accidents is increasing yearly, and the vast majority are caused by small private cars. Therefore, it is necessary to improve the traffic safety awareness of drivers and help car manufacturers to design traffic risk prediction systems. The Backpropagation neural network (BPNN) algorithm is used as the technical basis, combined with the MATLAB operation program, to simulate the driving process of the car. Dynamic predictive models are built to predict and analyze vehicle safety risks. Multiple experiments found that: (1) in various simulations, the simulation driving process of MATLAB is more in line with the actual car driving process; (2) the error between BPNN and the actual driving prediction is within 0.4, which can meet the actual needs. Predictive models are optimized to deploy and predict in various traffic situations. The model can effectively prompt risk accidents, reduce the probability of traffic accidents, provide a certain degree of protection for the lives of drivers and passengers, and significantly improve the safety of traffic roads.
Keyphrases
  • air pollution
  • neural network
  • artificial intelligence
  • machine learning
  • healthcare
  • health insurance
  • deep learning
  • molecular dynamics
  • quality improvement