Study On Ultrasonic Detection Technology Of Autonomous Driving

With the rapid increase of vehicles,frequent traffic accidents and road congestion,people's demand for automatic driving technology has become a hot research area.Waseda University's COMS Auto Driving Project aims to extend the auto-driving technology used in compact electric vehicle COMS(Toyota Motor Company's compact electric vehicle)to conventional vehicles.At present,COMS has been able to avoid obstacles in short distance and low speed.However,in the obstacle location system,the image-ultrasonic integrated positioning system has the problems of slow response speed and low accuracy of obstacle location,which leads to low accuracy of obstacle avoidance path and long time of obstacle avoidance process,so it is difficult to meet the actual needs of obstacle avoidance.The Waseda University COMS Autonomous Drving Program is dedicated to the promotion of self-driving technology for small electric vehicles to COMS.The automatic obstacle avoidance system is an important part of the automatic driving system,and it is also the key technology to realize the driving of the self-driving vehicle in the complex road environment.Therefore,this paper takes the COMS automatic obstacle avoidance system as the research goal and strives to improve the performance of the obstacle avoidance system.In this paper,the key technologies of multi-ultrasonic joint positioning system and path planning and obstacle avoidance are studied for COMS mini-electric vehicle of Waseda University,and the hardware test and verification system is built.In this paper,the performances of different kinds of sensors are analyzed.The ultrasonic sensors with high accuracy,fast speed and low price are selected to design a multi-ultrasonic joint positioning system.The working time of the sensors is designed to avoid the interference between different sensors.By dividing the detection range into regions and using different location calculation methods in different regions,the accuracy of location calculation is improved.By using Doppler effect,the calculation formula under moving state is revised,and the influence of velocity on positioning accuracy is reduced.After completing the design of obstacle positioning system,this paper also improves the path planning which is used in conjunction with the positioning system.Because of the improvement of the accuracy of obstacle location system,obstacle avoidance points can be accurately detected in obstacle avoidance paths,and obstacle avoidance paths are constrained to make the path more accurate and the path planning time shorter.After completing the improvement work of path planning,a simulation platform is designed to test and verify the performance and accuracy of path planning to ensure that the path is feasible.After completing all the design work,this paper builds a hardware test and verification system.HC-SR04 ultrasonic sensor and ADuino MCU are selected to form the hardware part,which is connected to the body central control system to provide positioning data.The different localization algorithms used in different regions are written into MCU in the form of codes to complete the compilation of the code part.In order to test the accuracy of the positioning system,a modular obstacle positioning test is designed.At the same time,the vehicle test has been carried out to verify the feasibility of the automatic obstacle avoidance system and the effect of automatic obstacle avoidance in the actual environment.The test results show that in the shortest obstacle avoidance distance,the farthest obstacle avoidance distance and dark light conditions,the obstacle avoidance effect is better than that of the original image-ultrasonic integrated positioning system on the vehicle.The error of obstacle avoidance is less than 5 cm and the time of obstacle avoidance is less than 5 s,which verifies the superiority of the multi-ultrasonic combined positioning system.