Automatic driving vehicle trajectory planning control implementation method

Abstract

The invention discloses a trajectory planning control implementation method for an automatic driving vehicle, and provides a path planning method for the automatic driving vehicle under a deterministic road framework, an uncertain road framework and a hybrid road framework aiming at the interpretability degree of a complex road scene. Determining various constraint conditions under various frame modes, adopting a reinforcement learning trajectory planning algorithm, performing trajectory planning in a credible security region, performing error optimization of an actually planned trajectory and a virtual security trajectory, and realizing design of a dynamic trajectory interval so as to ensure that the security interval of the reachable trajectory realizes dynamic mobility; and finally, realizing tracking control of path planning by adopting a fusion reinforcement learning and predictive control algorithm. According to the method, path planning control of the automatic driving vehicle in complex road and weather environments such as coastal cities is facilitated, and the risk of path planning caused by the road environment is better reduced.

Description

technical field [0001] The invention belongs to the field of intelligent vehicles, in particular to a method for realizing trajectory planning control of an automatic driving vehicle. Background technique [0002] At present, the technology of self-driving vehicles has been greatly improved, especially in the low-speed application of the road conditions of the cross-sea bridge in the port area or the structured road scene in the closed park, and has achieved a small-scale industrialization trend. The common technologies used are environmental perception, navigation and positioning, The high-precision map, decision-making, path planning, and control links realize the intelligent driving of the vehicle through data transmission between comprehensive technologies. However, the method of realizing the intelligent vehicle by relying on multiple links has the lack of data volume and the delay of calculation time between each link. , high requirements for hardware equipment and man...

AI Technical Summary

Problems solved by technology

[0002] At present, the technology of self-driving vehicles has been greatly improved, especially in the low-speed application of the cross-sea bridge in the port area or the structured road scene of the closed park, which has realized the trend of small-scale industrialization. The common technologies used are environmental perception, navigation and positioning, High-precision maps, decision-making, path planning, and control links realize intelligent driving of vehicles through data transmission among comprehensive technologies. However, the method of realizing intelligent vehicles by relying on multiple links has data loss and calculation time delay between each link. , high requirements on hardware equipment and many other difficulties, especially for the path planning link, which requires accurate data volume and less calculation time to plan a better path. The problem of large amounts of data transmission is still unresolved. At the same time, high-precision maps require a large amount of data, calculation time, and poor real-time updates. Especially for coastal cities, the road environment is easily affected by fog, rain, snow and other weather, resulting in very poor road conditions. Complex, the road environment is often in a complex switching between deterministic road environment and uncertain road environment. For autonomous vehicles, it is difficult to plan a safe and comfortable path. Therefore, the traditional The path planning method of the method is not universal, and cannot ensure that the planned path is in real time. For the road environment of coastal cities, the road environment is more complex and changeable, and the certainty of the road environment cannot be controlled, and the road environment usually conforms to Due to the characteristics of traffic regulations, high mobility of people, and certain regularity in the destination of vehicles, self-driving vehicles usually realize a certain driving function, and then their application value can be realized.

Images