Bridge detection unmanned aerial vehicle autonomous navigation and stability control method based on reinforcement learning

Abstract

The invention discloses a bridge detection unmanned aerial vehicle autonomous navigation and stability control method based on reinforcement learning. The method comprises the following steps: 1, carrying out dynamic modeling on a four-rotor unmanned aerial vehicle for detecting a bridge based on a Newton-Euler method; 2, considering surrounding environment information and detection constraint conditions of a to-be-detected bridge, and proposing to realize autonomous obstacle avoidance flight of the unmanned aerial vehicle based on neural network reinforcement learning; 3, adopting binocular vision and IMU combined autonomous navigation under the assistance of a bridge building information model; and 4, carrying out attitude control by adopting an active disturbance rejection algorithm in inner loop control of the system, carrying out position control by adopting PID control in outer loop control, and achieving stable control of the bridge detection unmanned aerial vehicle under strong wind interference. The problems that in the unmanned aerial vehicle bridge detection technology, under-bridge positioning signals are lost, and flight control cannot be stabilized under strong wind interference are solved.

Description

technical field [0001] The invention belongs to the technical field of bridge detection, and in particular relates to an autonomous navigation and stability control method of a bridge detection UAV based on reinforcement learning. Background technique [0002] my country's transportation infrastructure construction is developing rapidly, and bridge construction is growing rapidly at an average annual rate of 20,000. By the end of 2018, my country has built more than 90,000 long-span bridges and more than 5,000 extra-long-span bridges, of which the main span exceeds 400. There are 105 super long-span bridges. The safety of existing bridges has become a core issue related to the national economy and national security. [0003] At present, most bridge inspection operations in my country still use three traditional technical methods: [0004] (1) Bridge inspection vehicle: it is the main method of bridge inspection at present. This type of method uses bridge inspection vehicles...

AI Technical Summary

Problems solved by technology

[0018] The purpose of the present invention is to provide an autonomous navigation and stable control method for bridge detection UAVs based on reinforcement learning to solve the problems of lack of positioning signals under the bridge and unsteady flight control under strong wind interference for UAV bridge detection technology

Images