A local trajectory fault tolerance method for an intelligent driving vehicle

Abstract

The invention discloses a local trajectory fault tolerance method for an intelligent driving vehicle, and the method comprises the steps: step 1, initializing a following mode, and constructing a vehicle coordinate system of the intelligent driving vehicle; step 2, based on the vehicle coordinate system, performing validity judgment on GPS data according to the obtained GPS data to obtain a GPS data validity judgment result; step 3, determining a lane line recognition state according to the GPS data and obtained lane line data, and determining a real fault-tolerant deviation; and step 4, updating the following mode of the intelligent driving vehicle according to the GPS data validity judgment result, the lane line recognition state and the real fault-tolerant deviation, and determining theplanned trajectory of the intelligent driving vehicle according to the updated following mode. According to the method, fault tolerance can be carried out based on each data state, and correspondingdata can be dynamically updated in real time. The method is easy to apply practically, the robustness of data processing can be improved, and the complexity of the system can be effectively simplified.

Description

technical field [0001] The invention relates to the technical field of intelligent driving, in particular to a local track fault-tolerant method for intelligent driving vehicles. Background technique [0002] Intelligent driving vehicles refer to intelligent vehicles that can automatically plan driving routes and control vehicles to reach predetermined targets by sensing the road environment through on-board sensor systems. Path trajectory planning in intelligent driving technology is one of the key technologies in the field of intelligent driving. In the process of intelligent driving in a semi-structured environment, the vehicle is prone to complex emergencies such as signal instability caused by GPS / IMU interference, unclear or discontinuous lane lines, or misidentified lane lines. Local trajectory planning is dynamic planning in non-specific scenarios, which requires accurate lane following and quick response to abnormal emergencies. [0003] At present, the existing l...

AI Technical Summary

Problems solved by technology

[0003] At present, the existing local path planning methods for intelligent driving vehicles mainly include path planning methods based on heuristic search, intelligent bionics, behavior planning, and reward learning. These methods have achieved good results in global path planning, but have not Consider the difficult adaptability of the planning system due to data instability, such as lane discontinuity and lane error recognition under semi-structured roads, or GPS / IMU positioning distortion

Images