Jia-qiang E

Driving-Record-Based Distributed Path-Planning for Autonomous Vehicle

The influence from individual preference and circumstance has been considered in the distributed path-planning model which is generated by Maximum Likelihood Estimation (MLE) and Bayesian Theory, and the model parameter can be determined with actual driving records of an optimal path-planning model. The new model develops the previous one from a single path into a distribution based on driving records so that the actual optimal path during driving can be adjusted according to the real-time position, which avoids oscillation along the single path and improves practicality and amenity.

Intelligent Forecasting on Composite Regeneration Endpoint of DPF in the Vehicle

A mathematical model based on fuzzy least squares support vector machines has been developed to forecast the composite regeneration endpoint of the diesel particulate filter in Vehicle. The results show that the relative error of forecasting model of composite regeneration endpoint is less than 3.5%, and the oxygen concentration of exhaust gas has the biggest effect on the endpoint of composite regeneration, followed by the mass flow rate of exhaust gas, the microwave power, the temperature of exhaust gas and the amount of cerium-based additive.