A deep learning-based low-speed automatic driving car control method and system

Abstract

A low-speed self-driving car control method based on deep learning, one: start the system; two: wait for the user's route information instruction, and then enter the next step after obtaining it; three: obtain map data and vehicle position and attitude data respectively through the network; four: according to User route information, map information and vehicle location information plan vehicle driving route plan; five: the vehicle collects road surface image data, recognizes the drivable road surface, selects a suitable driving route from the planned vehicle driving route plan, and adjusts the vehicle posture to avoid open obstacles; six: issue control commands, and the vehicle will drive according to the driving route determined in five. The vehicle uses global path planning to determine the real-time position of the car on the network map; uses the communication system to remotely send the starting point and destination position of the car from the PC or mobile client and returns the real-time status information of the car; through global path planning, local obstacle avoidance and communication system The combination of realizes low-cost low-speed automatic operation in a semi-open environment.

Description

technical field [0001] The invention relates to the technical field of intelligent control, in particular to a low-speed automatic driving car control method and system based on deep learning. Background technique [0002] In recent years, autonomous driving technology has developed extremely rapidly. Most of the existing automatic driving systems use equipment such as lidar and millimeter-wave radar to obtain road information. Among them, radar has a long detection distance, can accurately obtain three-dimensional information of objects, and has high stability and robustness, but lidar has a narrow detection range and is greatly affected by the environment; millimeter-wave radar cannot perceive pedestrians, and cannot detect pedestrians. Accurate modeling of surrounding obstacles, and most importantly, the application cost of radar is high, which is not suitable for mass production and use. And with the development of economy and technology, various fields are moving towa...

AI Technical Summary

Problems solved by technology

With the mature development of the logistics industry, the traditional method of self-collection of goods has brought a lot of inconvenience, which can no longer meet people's needs

Therefore, there is an increasing demand for autonomous driving in semi-enclosed environments, such as campuses, parks and other areas. However, there are no commercialized autonomous driving cars on the market, let alone functions such as remote control and information interaction.