| In the past decades, lidar has been widely utilized in many fields. The rapid development of lidar, including the diversification of system, the increasement of functions and the improvement of performance, meets the needs of the various application background. But nowdays, the detection sensitivity of lidar meets the bottleneck of development due to the limitation of traditional detector. Gm- APD detector is a new kind of detector, and it has a single photon response sensitivity. It brings new hope to weak signal detection. It is quickly applied in the field of lidar, and a new lidar system named Photon Counting Lidar(PCL) appears. PCL has high range accuracy, high integration, low laser energy requirements, small size, low power consumption and so on, and has acquired great attention in recent years. Previous researches on PCL concentrates mainly on system research, the researches on the signal processing method of PCL are less. This is because Gm-APD is different from the traditional detector, for example, APD, CCD, etc. The traditional processing method is difficult to be used for reference. Now, there is no perfect model for digital signal processing with probability characteristic. The lack of theoretical research on this subject results in the shortcoming of the function and performance of PCL.This paper uses the poisson probability distribution to describe the response model of Gm-APD, and establishes the signal processing model of PCL according to the characteristic of digital signal and probability statistics. Using this model to investigate three kinds of typical PCL system, some breakthrough progress has been made in restraining the false-alarm probability, obtaining the intensity information, improving SNR, and improving the range accuracy.The false alarms of PCL are researched. According to the Poission probability model of Gm-APD response, the response situation of signal and noise are investigated. In outputs of Gm-APD response, signal is still relevant, but noise is not relevant. According to this difference, the relevance judgment methods for restraining false alarms is proposed. The experimental results show false alarms caused by the sun background light is restrained efficiently. This work provides a guarantee of the application of PCL in the daylight.For the problem of PCL unable to obtain intensity information, Composite Modulation Pulse Sequence(CMPS) is applied. CMPS is modulated not only on the pulse position, but also on the pulse intensity. According to the statistics of the response probability of different intensity echo pulses, an intensity calculating method is established. And then further analyzes the influence factors of the intensity calculating method, including the levels of intensity modulation, pulse sequence length and the echo signal intensity. Finally, the experiment results verify the intensity calculating method, and the intensity accuracy is better than 1%.For the SNR research of chirped amplitude modulation PCL, the signal processing model of chirped amplitude modulation PCL is establised on the basis of the poission probability of Gm-APD response. The influence factors of SNR are investigated. By analyzing the relationship of between echo signal intensity and SNR, an optimum echo signal intensity is obtained. Using an iris diaphram to adjust the echo signal intensity to the optimum, SNR of chirped amplitude modulation PCL is enhanced. Besides, the distribution probability of false alarms caused by noise is deduced, and an interesting phenomenon is discovered. When the signal is strong, the false-alarm probability is low; when the signal is weak, the false-alarm probability is high. According to characteristics of false-alarm probability distribution, use sliding window to obtain the signal envelope and use threshold to reduce false alarms in the trough, and thus SNR of chirped amplitude modulation PCL is improved. When the echo signal intensity is 0.5 photoelectron per nanoseconds, the experimental results show that SNR is increased by 3.3 d B.For the range accuracy research of chirped amplitude modulation PCL, the influence factors of range accuracy are analyzed according to the signal processing model of chirped amplitude modulation PCL with the characteristic of probability and statistics. The relationship of the range accuracy being inversely proportional to the signal modulation bandwidth is discovered. Due to the way of mixing, the signal modulation bandwidth is restrained by the sampling frequency of Gm-APD, which influences the range accuracy. Besides, intermediate frequency spectrum interval also restrains the range accuracy. Aiming at these two factors, on the one hand, a novel mixing method is proposed, which named pre-mixing method. Using the local signal to modulate the sampling window width of Gm-APD to complete mixing with the echo signal, this mehtod can break the limitation of the signal modulation bandwidth and make chirped amplitude modulation PCL work on broader modulation bandwidth to improve the range accuracy. On the other hand, using I/Q quadrature phase detecting method to extract the phase information which isn’t used before. Through the ingenious design, the intermediate frequency information and the phase information can be effectively combined together, resulting in a better range accuracy. The simulation results show that when the echo signal intensity is-10.5 ns and the noise intensity is-10.01 ns, the range accuracy is improved from = 1.47 mrawdR ± to 14.3 cmfinedR = ±143cm...
|
PDF Full Text Request