时间相关单光子计数型激光雷达距离判别法

Distance Estimation Method Based on Photon Counting LIDAR

为了解决光子计数型激光雷达测量系统,收集到的信号存在不确定性的问题,提出一种贝叶斯推论下的马尔可夫信号法,可实现回波峰值位置、峰值幅度等参量的提取,达到信号重建的目的.依据贝叶斯理论,利用后验预设分布模型近似描述参量的先验分布,有效提高对参量存在空间的探测速度.在求解过程中,将这些参量构成一个全局近似解,利用贝叶斯推论对所有全局近似解进行评估,挑选出最合适的全局近似解作为问题的全局最优解.实验表明,该方法具有较高的准确性和良好的鲁棒性.

In order to solve the certain randomness in the signal received of photon counting. Light detection and ranging a method of Markov techniques based on the Bayesian inference was proposed to extract the peak position, peak amplitude and the noise of the present sampling data, so that the accurate distance was easily got. In Bayesian inference, the posterior distribution can be adopted to describe the prior distribution as approximately, which could effectively improve the exploration of the parameter spaces. During processing, the updated parameter and other parameters constituted a global approximate solution, and the Bayesian inference was used for the optimal solution in the global approximation solutions. Experiments demonstrated show that the parameters can be estimated to a high degree of accuracy and the method is robust.