CN110930064A - Method for extracting space-time probability of Mars dust storm and evaluating landing safety - Google Patents

Abstract

The invention discloses a method for extracting space-time probability of Mars dust storm and evaluating landing safety, which comprises the following steps: step 1, identifying a dust storm object based on RGB (red, green and blue) color remote sensing images; step 2, analyzing the time probability of the Mars dust storm; step 3, analyzing the space probability of the Mars dust storm; and 4, evaluating the safety of the Mars landing area. The invention has the advantages that: the Mars dust storm object can be accurately identified and the area extraction is carried out; the problem that the regularity of repeated occurrence of dust storms in a plurality of Mars years is neglected in the prior art is solved; the problem that the characteristics of the dust storm space distribution rule are ignored in the prior art is solved; the appropriate landing time and safe area are selected for the Mars mission.

Description

A method for spatial and temporal probability extraction and landing safety evaluation of Martian dust storms

technical field

The invention relates to the technical field of planetary remote sensing and planetary meteorology, in particular to a method for extracting the space-time probability of a Martian dust storm and evaluating the landing safety based on remote sensing images.

Background technique

Mars is the most Earth-like body in the solar system. If there is extraterrestrial life in the solar system, Mars is the most likely. The remote sensing and in-situ detection of Mars has a profound impact on the search for water sources and traces of life. The development of Mars exploration missions is of great strategic significance to my country's scientific and technological, economic and social development.

From the 1960s to the present, the United States and the former Soviet Union and other countries have successively launched a number of Mars landing probes. Limited to the level of engineering technology and science, the early landing probes did not consider meteorological, terrain and other related factors, so the landing success rate was very low. The first lander "Mars 2" was swallowed by a global dust storm at the beginning of its landing, and "Mars 3" also suffered from a dust storm that caused the communication system to be destroyed. In 2004, NASA's "Spirit" and "Opportunity" landed in the southern hemisphere summer, encountered a larger than expected dust storm, and the landing sites were 10.1km and 24.6km away from the center of the preselected landing ellipse, respectively. In June 2018, a global dust storm on Mars caused Opportunity to lose contact with the ground. Therefore, the probability of dust storms in the pre-selected landing zone on Mars is related to the success of the landing mission and affects the landing accuracy, as well as the subsequent normal operation of the probe. The detection and research of the climatic laws and topographical features of Mars have continued to deepen, but dust storms are still a hot and difficult point of study. Both NASA and the China National Aeronautics and Space Administration are expected to launch new Mars exploration missions in 2020. my country's lander is expected to achieve a soft landing on the surface of Mars and a Mars rover tour in 2021. These Mars missions all need to carry out research on the probability analysis and safety evaluation of dust storms in the landing area in advance.

The implementation process and results of previous Mars surface landing missions have the following three problems:

(1) The previous calculation of dust storm time probability only considers the percentage of the daily average dust storm area on Mars, and does not consider the recurrence probability of dust storms between Martian years. Suppose there are two Martian days, the former has dust storms every year, and its coverage is small; while the latter has only one dust storm in many years, but its coverage is larger. The calculated dust storm probability of the latter is greater than that of the former, which is obviously unreasonable.

(2) The predecessors only considered the time probability of the occurrence of dust storms, ignoring the spatial probability and distribution characteristics of dust storms in the pre-selection landing area. The safe landing position and area in the pre-selection landing area need to be considered.

(3) The temporal and spatial distribution law of dust storms in the comprehensive pre-selected landing area was not considered to evaluate and select the appropriate time and safe landing area for the Mars landing mission, so as to provide safety guarantee for the Mars rover's subsequent inspection of the Mars surface.

SUMMARY OF THE INVENTION

Aiming at the defects of the prior art, the present invention provides a method for extracting the space-time probability of a Martian dust storm and evaluating the safety of landing, which solves the defects existing in the prior art.

In order to realize the above purpose of the invention, the technical scheme adopted by the present invention is as follows:

A method for extraction of space-time probability of Martian dust storm and evaluation of landing safety, comprising the following steps:

Step 1, pre-selected landing zone dust storm identification;

The data used in the present invention are the remote sensing images of the Martian surface captured by the Mars orbiting probe, which are RGB color remote sensing images composed of red (Red, R), green (Green, G) and blue (Blue, B) bands, wherein the R band The wavelength range is 580-620nm, the G-band wavelength range is 505-525nm, and the B-band wavelength range is 400-450nm. Different features on the Martian surface absorb and reflect sunlight differently. In RGB color remote sensing images, sand and dust are usually yellow, exposed rocks are black, and clouds (water vapor) are white. Martian dust storms and clouds (water vapor) often accompany and are similar in color and affect each other, so it is necessary to distinguish the two through the B and R bands. Clouds (water vapor) have a strong ability to absorb red light, and have weak reflection in the R band, and the color tone is dark. The reflectivity of sand and dust in a dust storm increases with the wavelength, and the red light absorption ability is weaker than the blue light absorption ability, so the color tone is brighter in the R band and darker than the cloud (water vapor) in the B band. In addition, dust storms generally have special texture features such as feather-like and pebble-like. By comparing multiple RGB color remote sensing images of consecutive Martian days, special forms of Martian dust storms can be found. Combining the above two points, dust storms can be identified and extracted. The specific steps to identify dust storm objects based on RGB color remote sensing images are as follows:

The input RGB color remote sensing image set is I(i,j), the spatial extent covers the preselected landing area, and the time interval is one Martian day. A Martian year MY is the time it takes for Mars to orbit the sun once, assuming a total of n years of RGB color remote sensing images (i=1,2,...,n). A Martian day Ls represents the change in the angle between Mars and the sun is 1°, which corresponds to 0° from the northern hemisphere vernal equinox, and the summer solstice, autumn equinox, and winter solstice are 90°, 180° and 270° respectively. The longitude of the sun represents the seasonal change of Mars, j =1,2,...,360 (unit:°). The sub-images corresponding to the three bands of red, green and blue are I(i,j) _R , I(i,j) _G and I(i,j) _B . Take the RGB color remote sensing image I(i,j) of the ith Martian year and the jth Martian day as an example to identify the dust storm object:

(1) Difference area extraction. On the RGB color remote sensing images, the ground objects are consistent in the image in a short period of time, and the areas with obvious changes can be regarded as possible dust storm areas. Identify the differences and changes between the pre-selected landing areas in the three consecutive Martian remote sensing images I(i,j-1), I(i,j), I(i,j+1) in the ith Martian year, The changed regions are extracted and vectorized into polygonal objects. The identified polygon object set is represented by D ₀ (i, j, Id ₀ ), where Id ₀ represents the polygon object sequence number (Id ₀ =1, 2, . . . , m _0ij ).

(2) Dust storm and cloud (water vapor) distinction. Compare the R-band sub-image I(i,j)R and B-band sub-image I(i,j) of the RGB color remote sensing image with the brightness and darkness of _B. If the ks-th polygonal object D(i,j,ks) is in I( i,j) is bright in the _R image, and dark in the I(i,j) _B image, the polygon object is judged to be a dust storm; otherwise, it is a cloud (water vapor). This generates a dust storm polygon object set D(i,j,Id), where D is a subset of D ₀ , and Id represents the dust storm polygon object number (Id=1,2,...,m _ij ), and calculates each dust storm polygon The area of the object generates the dust storm polygon area set A(i,j,Id).

Step 2: Probability analysis of Martian dust storm time;

The formula for calculating the daily average dust storm probability of the Mars landing zone consists of two parts. First, calculate the area percentage P ₁ (j) of dust storm objects identified in the landing zone on the jth Martian day in n Martian years, then calculate the probability P _E1 (j) of dust storms recurring in n Martian years, and finally combine the above two steps The result of , obtains the time probability P _T (j) of the occurrence of the solar dust storm in the preselected landing zone.

假设n个火星年中同一个火星日j识别出的尘暴多边形总个数为M，则

那么n个火星年中第j个火星日所有识别出的尘暴对象加权面积百分比之和P₁(j)为：(1) Calculate the percentage P ₁ (j) of the daily average dust storm coverage area on Mars. Assuming that the area of the pre-selected landing area on Mars is A _T , the set of dust storm polygons identified in the ith Martian year and the jth Martian day is D(i, j, Id), and there are m _ij dust storm polygon objects (Id=1, 2,...,m _ij ), where the area of the kth dust storm is A(i,j,k). Divide the area of the kth dust storm by the area of the entire landing zone to get the area percentage of the dust storm polygon

Assuming that the total number of dust storm polygons identified on the same Martian day j in n Martian years is M, then

Then the sum of the weighted area percentages of all identified dust storm objects on the jth Martian day in n Martian years P ₁ (j) is:

in,

A _f (i,j) is the sum of the dust storm area percentages of the ith Martian year and the jth Martian day.

(2) Calculation of dust storm recurrence probability P _E1 (j). Whether a dust storm occurs in the jth Martian day of the ith Martian year can be identified by Is(i,j). Is(i,j)=1 if a dust storm occurs, otherwise Is(i,j)=0. Then the probability of repeated occurrence of the jth Martian dust storm in n Martian years is:

(3) Calculation of the time probability P _T (j) of the occurrence of the dust storm. The average dust storm probability of the Martian landing zone on the jth Martian day can be obtained by multiplying the weighted average of the percentage of Martian dust storm coverage and the dust storm recurrence probability:

P _T (j)=P ₁ (j)×P _E1 (j) (4)

Step 3: Spatial probability analysis of Martian dust storms;

Considering the uneven spatial distribution of Martian dust storms in the pre-selected landing zone, the occurrence probability of dust storms in different regions is different. Therefore, the polygons of the pre-selected landing area are divided into uniformly distributed square grids, whose side length is L, there are p total, and the grid data set is Grid(g), g=1,2,...,p. The annual average probability of dust storms in different grids is calculated, and the result of the annual average probability of dust storms in all grids is the spatial distribution characteristics of dust storms in the entire landing area.

假设网格g中n个火星年中识别出的尘暴多边形总个数为M_g，则

那么网格g的年加权尘暴面积百分比为

Take the gth grid as an example to calculate the annual average occurrence probability of dust storms. Assuming that the set of dust storm polygons identified in the g grid in the ith Martian year is D _g (i,Id _g ), there are m _ig dust storm polygon objects (Id _g =1,2,...,m _ig ), The area of the k- _th dust storm is A(i, _kg ). Percentage of dust storm area within the g grid for the ith Martian year

Assuming that the total number of dust storm polygons identified in n Martian years in grid g is M _g , then

Then the annual weighted dust storm area percentage for grid g is

Is(i,g)=1 if a dust storm occurs in the ith Martian year in grid g, otherwise Is(i,g)=0. The probability P _E2 (g) of dust storm recurrence in grid g in n Martian years is:

The annual weighted dust storm area percentage P ₁ (g) of grid g is multiplied by the dust storm recurrence probability of grid g in n Martian years, and the annual average dust storm probability P _s (g) of grid g can be obtained. The specific formula is: :

P _s (g)=P ₁ (g)×P _E2 (g) (6)

Step 4, the safety evaluation of the Mars landing zone;

According to the time probability and spatial distribution of dust storm occurrence in the landing area calculated in the previous steps, the Mars landing area mission can be evaluated and selected from time and space. In terms of time, dust storms with strong intensity and long duration will interfere with the mission of the Mars landing zone. Therefore, a continuous Martian day with a low probability of dust storms should be selected as the preferred time period for the landing zone. Assuming that the daily average probability threshold of Martian dust storm is P _a , Martian days exceeding this threshold are not suitable for the safe landing of the Mars lander. The set of consecutive Martian days T _S less than this threshold is selected as the safe landing zone time. In space, if a grid has a high probability of annual average dust storms, the dust storms in the grid will produce certain images of the lander or the rover after landing, so it is not suitable as a safe landing area grid for Mars missions . Assuming that the grid's annual average probability threshold of Martian dust storms is P _b , the landing zone grid exceeding this threshold is not suitable for the safe landing of Mars landers. According to the threshold, the grid is divided into safe grids (P _S (g)<P _b ) and unsafe grids (P _S (g)>P _b ), and the adjacent safe grids are merged to generate a safe area. A collection of polygons S _S . Then, according to the topographic data of the pre-selected landing area on Mars, the flat area with few rocks and impact craters in the safe area polygon set is selected as the safe area polygon set Ss' that integrates topography and dust storm factors.

Compared with the prior art, the advantages of the present invention are:

According to the difference in reflectivity of dust storms and clouds (water vapor) in the red and blue bands, the Martian dust storm objects can be accurately identified and the area can be extracted; the dust storm time probability of the pre-selected landing zone can be calculated based on the probability of repeated dust storms and the percentage of coverage area. , to solve the problem that the existing technology ignores the regularity of the repeated occurrence of dust storms in multiple Martian years; divide the pre-selected landing area into regular grids, calculate the annual average probability of dust storms in the grid, and study the spatial distribution characteristics and laws of dust storms to solve the existing technology. The problem of the characteristics of the spatial distribution of dust storms is ignored; finally, the safety evaluation of the pre-selected landing area is carried out according to the spatial and temporal distribution of dust storms, and the appropriate landing time and safe area are selected for the Mars mission.

Description of drawings

Figure 1 is a topographic map of the surface of Mars;

Figure 2 is an RGB color remote sensing image map;

Figure 3 is a statistical diagram of the daily average dust storm probability in Martian days;

Figure 4 is a graph of the annual average probability of dust storms in the study area.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below according to the accompanying drawings and examples.

1 Study area and remote sensing images of Mars

The example of the present invention focuses on the pre-selected landing area (black polygon) of Chryse of my country's 2020 Mars exploration mission. The research area is a 1600km circle in the center of the Chris area, and the latitude and longitude range is about (0°-60°S, -60° E-0°), as shown in Figure 1.

The data source of the example of the present invention is the RGB color remote sensing image covering the surface of Mars, from the remote sensing image captured by the Mars Observer camera on the Mars Global Surveyor, with a total of 4 Martian years of data.

2 Safety Analysis of Mars Landing Mission

(1) Dust storm identification in the study area

As shown in Figure 2, (a), (b) and (c) are the corresponding blue-band and red-band images, respectively, based on the RGB color remote sensing images of different Martian days in the study area to identify dust storms. Figure 2(b) and (c) are the blue and red band images corresponding to the RGB color remote sensing image with Mars time MY=27, Ls=203.6°. Among them, white arrows point to identified dust storms, while black arrows point to clouds (water vapor). A total of 1172 dust storm objects were identified in the RGB color remote sensing images of 4 Martian years within the 1600km circle of the Kris region.

(2) Probability analysis of dust storm time in the study area

According to the formulas (1)-(4) of the daily average probability of dust storms, the average daily probability of dust storms in the study area is calculated. The results are shown in Figure 3. In Figure 3, the abscissa is the Martian day Ls, and the ordinate is the average daily probability of dust storms. The daily average probability of dust storm in the study area is the highest at 0.21, which is located at Ls=228°. Among them, the probability of continuous dust storms between Ls=177°-239° and Ls=288°-5° is relatively high, with an average value of 0.095 and 0.041. Therefore, it is not suitable for the landing time of the Mars mission. The time is between Ls=239°-288° to Ls=5°-177°, which can be used as a suitable landing time for the Mars mission.

(3) Spatial probability analysis of dust storms in the study area

The study area is divided into 0.5°×0.5° grids, and the annual average probability of dust storms in each grid is calculated by formulas (6)-(7). The spatial distribution results of the annual average dust storm probability in the study area are shown in Figure 4. The spatial probability of dust storms in the study area ranged from 0% to 10.8%. Combined with the contour data generated by the topographic data of the study area, a flat area without impact craters and a low probability of dust storm space was selected as the safe landing area for the Mars mission. A total of three suitable landing areas were extracted, as shown in the black dotted box in Figure 4. The safe landing areas 1 and 2 are in the west of the Chris area, and the landing area 3 is in the east of the Chris area, with areas of 65,856km ² , 84,744km ² and 70,242km ² , respectively. , the mean values of the dust storm spatial probability are 0.45%, 0.26% and 0.03%, respectively.

Those of ordinary skill in the art will appreciate that the embodiments described herein are intended to help readers understand the implementation method of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations without departing from the essence of the present invention according to the technical teaching disclosed in the present invention, and these modifications and combinations still fall within the protection scope of the present invention.

Claims (1)

1. a method for extraction of Martian dust storm spatiotemporal probability and landing safety evaluation, is characterized in that, comprises the following steps: Step 1: Identify dust storm objects based on RGB color remote sensing images. The specific steps are as follows: The input RGB color remote sensing image set is I(i,j), the spatial range covers the pre-selected landing area, and the time interval is one Martian day; a Martian year MY is the time it takes for Mars to orbit the sun once, assuming a total of n years of RGB color Remote sensing image (i=1,2,…,n); a Martian day Ls represents the change of the angle between Mars and the sun is 1°, which corresponds to 0° from the northern hemisphere vernal equinox, and the summer solstice, autumn equinox, and winter solstice are 90° and 180° respectively. ° and 270°, the seasonal changes of Mars are represented by the solar longitude, j=1,2,...,360 (unit:°); the sub-images corresponding to the three bands of red, green and blue are I(i,j) _R , I (i,j) _G and I(i,j) _B ; Take the RGB color remote sensing image I(i,j) of the i-th Martian year and the j-th Martian day as an example to identify the dust storm object: (1) Extraction of difference areas; on RGB color remote sensing images, the ground objects appear consistent in the image in a short period of time, and the areas with obvious changes can be regarded as possible dust storm areas; identify three consecutive Mars remote sensing images in the ith Martian year I ( i,j-1),I(i,j),I(i,j+1) The difference and change between the pre-selected landing area features, the change area is extracted and vectorized into a polygon object; the identified polygon The object set is represented by D ₀ (i, j, Id ₀ ), where Id ₀ represents the polygon object serial number (Id ₀ =1, 2, . . . , m _0ij ); (2) Distinguish dust storms and clouds (water vapor); compare the R-band sub-image I(i,j) of the RGB color remote sensing image and the B-band sub-image I(i,j) _B tones are bright and dark, if the ks th polygonal object If D(i,j,ks) is bright in the I(i,j) _R image, but dark in the I(i,j) _B image, the polygon object is judged to be a dust storm; otherwise, it is a cloud (water vapor) ; Generate a dust storm polygon object set D(i,j,Id), where D is a subset of D ₀ , Id represents the dust storm polygon object serial number (Id=1,2,...,m _ij ), calculate each dust storm The area of the polygon object generates the dust storm polygon area set A(i,j,Id); Step 2: Probability analysis of Martian dust storm time; The formula for calculating the daily average dust storm probability of the Martian landing zone consists of two parts; first, calculate the percentage P ₁ (j) of the dust storm object area identified by the landing zone on the jth Martian day in n Martian years, and then calculate the dust storm repetition in n Martian years The probability of occurrence P _E1 (j), and finally the results of the above two steps are combined to obtain the time probability P _T (j) of the occurrence of the solar dust storm in the pre-selected landing area; (1) Calculate the percentage P ₁ (j) of the average dust storm coverage area on a Martian day; assuming that the area of the pre-selected landing area on Mars is A _T , the set of dust storm polygons identified in the i-th Martian year and the j-th Martian day is D(i, j,Id), there are m _ij dust storm polygon objects (Id=1,2,...,m _ij ), where the area of the kth dust storm is A(i,j,k); Divide the area by the entire landing zone area to get the area percentage of the dust storm polygon

Assuming that the total number of dust storm polygons identified on the same Martian day j in n Martian years is M, then

Then the sum of the weighted area percentages of all identified dust storm objects on the jth Martian day in n Martian years P ₁ (j) is:

in,

A _f (i,j) is the sum of the dust storm area percentages of the i-th Martian year and the j-th Martian day; (2) Calculate the probability of repeated occurrence of dust storms P _E1 (j); whether dust storms occur in the jth Martian day of the ith Martian year can be identified by Is(i,j); if dust storms occur, Is(i,j )=1, otherwise Is(i,j)=0; then the probability of the jth Martian dust storm recurring in n Martian years is:

(3) Calculate the time probability P _T (j) of the occurrence of dust storms; the average dust storm probability of the j-th Martian day in the Martian landing zone can be obtained by multiplying the weighted average of the percentage of Martian dust storm coverage and the recurrence probability of dust storms: P _T (j)=P ₁ (j)×P _E1 (j) (4) Step 3: Spatial probability analysis of Martian dust storms; Divide the polygons of the pre-selected landing area into uniformly distributed square grids, whose side length is L, there are p total, and the grid data set is Grid(g), g=1,2,...,p; calculate the dust storms in different grids The annual average probability of occurrence, and the result of the annual average probability of dust storms in all grids is the spatial distribution characteristics of dust storms in the entire landing area; Take the g-th grid as an example to calculate the annual average occurrence probability of dust storms. Assuming that the set of dust-storm polygons identified in the g-grid in the i-th Martian year is D _g (i,Id _g ), there are m _ig dust-storm polygons in total Object (Id _g = 1,2,...,m _ig ), where the area of the k _g dust storm is A(i, _kg ); the percentage of dust storm area within the g grid for the ith Martian year

Assuming that the total number of dust storm polygons identified in n Martian years in grid g is M _g , then

Then the annual weighted dust storm area percentage for grid g is

If a dust storm occurs in the ith Martian year in grid g, then Is(i,g)=1, otherwise Is(i,g)=0; the probability of repeated dust storms in grid g in n Martian years P _E2 ( g) is:

The annual weighted dust storm area percentage P ₁ (g) of grid g is multiplied by the dust storm recurrence probability of grid g in n Martian years, and the annual average dust storm probability P _s (g) of grid g can be obtained. The specific formula is: : P _s (g)=P ₁ (g)×P _E2 (g) (6) Step 4, the safety evaluation of the Mars landing zone; According to the time probability and spatial distribution of dust storms in the landing area calculated in the previous steps, the Mars landing area mission can be evaluated and selected from time and space; Therefore, a continuous Martian day with a small probability of occurrence of dust storm is selected as the preferred time period for the landing zone; assuming that the daily average probability threshold of Martian dust storm is P _a , Martian days exceeding this threshold are not suitable for the safe landing of the Mars lander; The set of consecutive Martian days T _S less than this threshold is used as the safe landing zone time; spatially, if a grid has a high probability of occurrence of annual average dust storms, the dust storms in the grid will cause the lander or the rover after landing. There are certain images, so it is not suitable for the safe landing area grid of Mars missions; Assuming that the annual average probability threshold of Martian dust storm of the grid is P _b , the landing area grid exceeding this threshold is not suitable for the safe landing of Mars landers; according to This threshold divides the grid into a safe grid (P _S (g)<P _b ) and an unsafe grid (P _S (g)>P _b ), and merges the adjacent safe grids to generate a polygon of the safe area Set S _S ; then according to the topographic data of the pre-selected landing zone on Mars, select the flat, rocky and crater-less area in the safe area polygon set as the safe area polygon set Ss' that integrates topography and dust storm factors.

Images