Obstacle Detection Using Ultrasonic Sensors For Unmanned Ground Vehicle In Cross-Country Environment

To drive in unstructured cross-country environment for Autonomous Land Vehicle(ALV),the surrounding environment information is needed to be acquired,processed with various sensors,and the obstacles would be detected and localized.The environment cognition capability of ALV provides the base for the navigation,palnning,and control.Ultrasonic sensors,compared with the main-stream sensors such as LiDAR,microwave radar,intrinsically possess the faults including small detection range,low precision of obstacle localization.Therefore,some opions once were concluded that ultrasonic sensor was unfitful for the obstacle detection of ALV.On the observation of the merits of ultrasonic sensor such as strong adaptability,low cost,one kind of ultrasonic sensor with large detection range and high precision is discussed comprehensively.The innovative contributions including the design of ultrasonic sensor with large range,the obstacle detection algorithm,and the system implementation,are introduced in detail as follows:1.A new scheme of ultrasonic sensor for object detection is proposed with reflective piezoelectric ceramics sonar transducer,and the ultrasonic sensor has been developed with the characters of large range,as well as high directionality,small size.Based on the analysis on the radial mode and the frequency characteristics of composite piezoelectric ceramics cylindrical pipe,the coupling between the resonant frequency and the radial area of the piezoelectric ceramic piece is successfully removed.The usability of the sonar transducer was approved through the vehicle experiment.2.A new optimized scheme on the transmitting circuit and the receiving circuit of ultrasonic sensor,is proposed.The invert type transmitting circuit of sonar transducer is developed through the comprehensive design including the match circuit,the tuned transformer,and the high frequency pulse transformer.The receiving circuit of the vehicle-mounted ultrasonic sensor is developed,which is made up of some key circuits including time-varying gain circuit with low noise,tuned amplifying circuit with high sensitivity.Both the theoretical analysis and the simulated experiment results show that the proposed scheme effectively improve the transmitting power and the receiving sensitivity.So that the developed ultrasonic sensor based obstacle detection system possesses the characteristics of large detection range and high localization precision.Real vehicle experiments approved that the ultrasonic obstacle detection system satisfy the requirement of ALV.3.A new obstacle detection algorithm based on the proposed ultrasonic sensor model is presented,in which all return signals are processed in one cycle,other than only the first return signal is processed in traditional method.Therefore the fake data arising from multi-path phenomena,cross-talk etc.,is automatically filtering out.The field test show that the algorithm is fit for the cross-country environment,with merits such as robustness and well performance.4.A dynamic data filtering method based on ultrasonic array is presented.The sonar return data is firstly processed through static filtering according to the geometric relationship of the ultrasonic array,then,dynamic filtering is executed using the orientation and the trajectory information of the vehicle.The dynamic filtering method is compared with traditional ultrasonic obstacle detection method,static filtering method in a typical field environment,experiment result prove the validity of the method to solve the fake data problem,and the accuracy of obstacle detection is improved.The innovative contributions above would provide key technology foundation for the application of ultrasonic sensors in unmanned ground vehicle.