Partly-observable automatic pilot decision-making method and system based on restraint online planning

Abstract

The invention relates to a partly-observable automatic pilot decision-making method and system based on constraint online planning. The decision-making method is mainly used for automatic pilot, so that a vehicle is enabled to have various driving modes. The system comprises a driving environment state unit, a search unit, a simulation unit and a cost constraint unit. By the decision-making method, a driving scheme can be generated according to current driving environment and can be adjusted in real time according to real-time road conditions and vehicle conditions to improve flexibility of the scheme. A history-based Monte Carlo search tree is built by the method, and simulating of a solution is performed on the basis of real conditions, so that reliability is improved. In addition, the method meets certain optimal selection conditions to ensure acquired strategies to be random strategies, so that the defect of deterministic strategies is made up. The decision-making method completelymeets driving needs of current common users, and various modes are provided for selection especially, so that user experience is improved greatly.

Description

technical field [0001] The invention relates to the technical field of automatic driving, in particular to a partially observable automatic driving decision-making method based on constraint online planning. Background technique [0002] Currently, autonomous driving involves three questions: the first is where am I? The second is where am I going? How is the third one going? Real autonomous driving needs to perfectly solve these three problems. The first problem refers to the positioning problem. In reality, driving road conditions are usually more complicated, so we need centimeter-level positioning. The second problem is the path planning problem, which is the problem to be solved in this patent. The third problem is the vehicle execution structure, that is, the control-by-wire system. The main operations include brake-by-wire, steering, and accelerator, that is, to control the vehicle according to the scheme obtained by the planning module. [0003] The traditional ...

AI Technical Summary

Problems solved by technology

[0009] Based on this, it is necessary to provide a partially observable autonomous driving decision-making method based on constrained online programming based on the above technical problems. Based on partially observable Monte Carlo programming, the strategy can be implemented by considering both the primary objective and the secondary objective constraints. The planning method solves the problem that the driving mode is single and the user can only passively accept the driving plan, and improves the user experience. This method has the advantages of flexible driving plan, high reliability, and comprehensive consideration, and has a wide range of application scenarios in the field of automatic driving

Images