Simulation Of Lunar Regolith Stochastic Media Model And Processing Of LPR Data Onboard The Rover Of Chang’E-3

The moon is the nearest planet to the earth.In Chinese deep space exploration engineering research,lunar exploration is one of the important contents.The success of the Chang’E-3 lunar program played an important role to promote people to understand the structure and features of the moon.The lunar substructure mainly consists of lunar soil and lunar rocks from shallow to deep.The contents of Cr,Sc and Ti in the lunar rock are enriched.It is an important goal for the future lunar exploration to study the moon sea basalt.The composition and structure of lunar soils can provide important clues for the evolution of the lunar geology.As the lunar environment is dry,the relative permittivity is low and the electromagnetic wave decays slowly.So the high frequency electromagnetic radar can be used as an effective mean to detect the near-surface structure of the moon.For the first time,Chang’E-3 carried the lunar radar system to obtain the lunar soil and rock detection results.The lunar penetrating radar(LPR)used the high-frequency electromagnetic wave reflection signal to identify the underground structure.In the detection process,the lunar radar signal often contains random noise and multiple reflections.Noise can seriously affect the recognition of the target body in the subsurface media.Therefore,the parameter modeling of the lunar soil structure and the method of processing the radar signal are very important to the reliability of the LPR interpretation.This paper introduces the terrain features of the landing point,and analyzes the possible impact on the lunar radar profile.Then,the simple chemical compositions of the lunar soil and rock layers are introduced.Based on the sampling data of Apollo exploration plan,the physical properties of the lunar layers are described.Finally,the chapter introduces the working principle and technical parameters of Chang’E-3 lunar radar,and provides the theoretical basis for the establishment of the near-surface model.Numerical modeling gives us insight in how such background features affect the performance of LPR.Besides,numerical modeling can also be used for testing,comparing and optimizing the performance of LPR.In the third chapter,according to the principle of equivalent stochastic medium,the numerical simulation of the lunar radar is carried out on the simple layers model,the surface relief model,the stochastic medium model,the rover influence model and the comprehensive geological model.Based on the numerical simulation,the received radar signal is also a kind of non-linear,non-stationary signals.The two-dimensional empirical(BEMD)is used to improve the target signal of the LPR data.This paper introduces the principle of empirical mode decomposition(EMD)algorithm and iterative screening process,and briefly introduces the principle of Hilbert-Huang transform.This method can decompose the signal into a number of different intrinsic modal functions(IMF).Through the Hilbert transform of the decomposed component,the instantaneous parameters are obtained to prove the frequency resolution ability of the EMD.BEMD used the mean envelope surface to extract the intrinsic modal components.Relative to EMD,the problem of BEMD is more complicated.One of the difficulties of BEMD is fitting the extreme envelope surface.The fitting of the envelope is directly related to the result of each decomposition.According to the finite point interpolation in space for surface fitting,it is a problem in the calculation of geometric disciplines.The increase in the amount of data in the two-dimensional image processing will result in the increase of the calculation cost.Therefore,it is also necessary to improve the computational efficiency.There are many methods to fit the envelope surface of BEMD algorithm in 2D graphics processing.In this paper,a linear interpolation method based on triangular basis function is adopted.The superiority of BEMD algorithm in two-dimensional image processing is proved by decomposing the two-dimensional simulation.In this paper,BEMD algorithm is applied to process data of LPR.The radar data is obtained with the center frequency of 500 MHz.Firstly,the original data of the lunar radar are pretreated.And the pretreated data is removed by the excess track,the splicing of the data block,the removal and adjustment of the delay time,and the removal of the mean value.Then,based on the BEMD theory,the pre-processed data are decomposed into four components.The noise components are removed.The remaining components are reconstructed,so as to achieve the purpose of improving the signal to noise ratio.For the shallow surface of the lunar radar detection data processing,BEMD provides a new processing method.Combined with the lunar geology and previous research results on the 500 MHz lunar radar profile,four layers are divided.They are reworked zone,ejecta,paleoregolith and basalt,respectively.It is estimated that the bedrock is 2.5 billion years ago.Before the Chang’e III crater formed 27 to 80 million years ago,the paleoregolith layer had accumulated at the base of the Eratosthenian basalts.Continuous ejecta from the crater was then deposited over this paleoregolith layer.After that,the area is reworked under the action of the ejecta from small local craters and micrometeorite bombardments.In this paper,the application of lunar exploration radar detection technology is introduced from four aspects: lunar exploration method,stochastic medium modeling,antenna and terrain modeling and signal processing and interpretation technology.It provides an effective theoretical basis for the development of LPR.