Ranging Method And Accuracy Analysis Based On Vehicle Laser Radar

With the continuous development of scientific and technological information,auto-driving has been favored by people and has leaped into one of the hottest research fields.In automatic driving system,the environment perception is like the visual and auditory system of the traveling vehicle,which is an important prerequisite and basis for vehicles to achieve advanced intelligent behavior.A vehicle laser radar is the core device to monitor the surrounding environment of autonomous vehicle,while accurate measurement of target distance is an important information for vehicle to perceive and judge the driving environment.A effective method of measurement and accuracy decide the ranging performance of the whole laser radar system.In the process of ranging,there are usually many kinds of noises in the received echo signal of laser radar,which lead to the low signal-to-noise ratio(SNR)of the echo signal,thus seriously affects the ranging performance of the system.Most of the past researches elevate the ranging accuracy by improving the hardware circuit of the system.However,the effect is limited.In recent years,the research of wavelet transform in signal denoising has developed rapidly,especially for non-stationary noisy signal,which has a very obvious effect.This thesis firstly determines the specific method of measurement.On the basis of this,form the aspect of SNR,it removes the noise components in the echo signal by using the wavelet transform algorithm,and extracts the effective echo,so as to achieve the ultimate goal of improving the ranging performance.The main research contents are as follows:Firstly,the specific method of measurement of the vehicle laser radar is determined.On the basis of this,the noise characteristics affecting the performance of the laser radar ranging system are studied,and the theoretical model of the laser radar transmitting signal and receiving the noisy echo signal is established.An adaptive wavelet denoising method is proposed,which focuses on the characteristic of system of echo noise and combines it with traditional wavelet denoising theory.The traditional method of choosing a fixed decomposition layer is improved while determining the decomposition layer number,and the final decomposition layer number is determined by the correlation degree of approximate coefficients of adjacent layers in the process of wavelet decomposition.The simulation results show that the proposed adaptive denoising method further improves not only the SNR,but also improves the echo signal waveform recovery performance compared with the denoising method proposed in former thesises.Through the simulation analysis of ranging accuracy,the results also show that the accuracy after denoising has been greatly improved.According to the characteristics that noises and useful signals coexist in the laser radar echo,this thesis puts forward a method of pulse accumulation to process the echo signal under the circumstances that the laser radar echo signal is weak.The simulation results show that it conforms to the theoretical principle.In addition,considering that the cumulative times can not be too large and the noise can be removed better,this thesis adopts a new scheme that combines pulse accumulation with adaptive threshold denoising mentioned above.Finally,the scheme is verified by simulation,and the curves of SNR and root mean square error(RMSE)of output signal versus input SNR are given.The results show that the scheme has a good denoising effect and better improvement effect on ranging accuracy.