Research On Ultrasonic Ranging And Positioning Method Based On Digital Signal Processing

As a typical non-contact measurement method,ultrasonic ranging has the advantages of strong environmental adaptability and low price,and it is widely used in obstacle avoidance systems deployed on mobile robots.However,this method might have to make a trade-off between the measurement accuracy and the effective working distance.Besides,due to the existence of the beam angle,it is impossible to locate obstacles precisely and cannot meet the detection requirements of mobile robots in complex environments.Towards the obstacle avoidance of mobile robots,the methods of the distance measurement based on single ultrasonic transducer and the angle measurement based on the multi ultrasonic transducer are explored from the perspective of digital signal processing of ultrasonic echo signals in this thesis.Concerning the contradiction between effective distance and measurement accuracy in ultrasonic distance measurement,an ultrasonic ranging scheme with cross-correlation as the core is explored.Firstly,a lifting wavelet transform is exploited for secondary denoising on echo signals to improve the signal-to-noise ratio,and the selection methods of wavelet base,decomposition level,and threshold function are discussed in detail.Moreover,aiming at the difficulty of obtaining the reference signal,the genetic algorithm and the Levenberg-Marquardt algorithm are utilized to estimate the parameters of the ultrasonic exponential model by fitting the echo signal at different distances,and the model reconstruction signal is used as a reference signal.Finally,a new two-step correlation method is proposed so as to solve the problem of a large amount of calculation in the cross-correlation method,which is roughly cross-correlation in the wavelet domain and accurate cross-correlation in the time domain.Furthermore,the Hilbert transform is used to obtain the envelope of cross-correlation function to search the peak value exactly.Aiming at the fuzzy of obstacle location caused by a large beam angle of the ultrasonic transducer,an ultrasonic positioning scheme with the intersection as the core is proposed,which is based on multi-element linear ultrasonic transducer array.Firstly,obstacles are divided into the point and the plane types according to the reflection of ultrasonic signals,and the corresponding positioning models are introduced.Also,the linear least square and the Taylor least square are utilized to fuse the multi transducer data.Finally,a multi-obstacle localization algorithm based on the hypothesis-test method is proposed to reduce the amount of data associated with multiple obstacle echo signals during a single detection period.Based on the above research,a ranging and positioning experimental system using a fiveelement linear ultrasonic transducer array is developed,which utilizes FPGA of Xilinx Company as the main control chip.The Verilog HDL is exploited to design global control,clock management,echo signal processing,temperature acquisition,serial communication and other functional modules to implement an algorithm of ranging by five-channel ultrasonic.Moreover,the control display interface of the host computer is designed to implement a multi-obstacle positioning algorithm and data display function via Matlab GUI.The experiment of obstacle ranging and positioning is performed to verify the feasibility of the algorithm in embedded systems.