Ecological remote sensing detection system for soil polluted area

Abstract

The invention relates to the technical field of soil pollution identification, in particular to an ecological remote sensing detection system for a soil pollution area. Determining a ground monitoring area according to the road coverage area and the tail gas influence distance; dividing according to the topographic data and the environmental data of the ground monitoring area to obtain ground sub-areas; according to the model data and the driving data of the vehicle in the ground subarea, tail gas emission data are obtained; pollution distribution data are obtained according to the tail gas emission data, the topographic data and the environmental data; dividing the ground sub-regions according to the pollution distribution data to obtain pollution sub-regions and pollution sub-data; sampling and testing the soil in the polluted subarea to obtain soil pollution data; analyzing the plants in the polluted sub-regions to obtain ecological abnormal data; obtaining a soil pollution coefficient according to the pollution sub-data of the pollution sub-region, the soil pollution data and the ecological anomaly data; according to the soil pollution coefficient, the soil pollution on the two sides of the road can be accurately identified and early warned.

Description

Technical Field

[0001] The present invention relates to the technical field of soil pollution identification, in particular to an ecological remote sensing detection system for soil pollution areas. Background Art

[0002] With the advancement of industrialization and urbanization, the road network has been expanded on a large scale, and the number of vehicles and mileage have also increased. At the same time, the exhaust gas generated by automobiles has led to increasing soil pollution near roads. It is necessary to identify and warn of soil pollution near roads.

[0003] The identification method of soil pollution usually relies on on-site sampling and chemical analysis, but this method has some obvious shortcomings. For example, the sampling and analysis speed is slow, and it takes time to collect data and conduct laboratory tests, etc., which leads to the inability to effectively control the polluted areas in a timely manner. In contrast, ecological remote sensing technology can monitor pollution in a large area in a short time through satellite images, drones and other remote sensing tools.

[0004] Although remote sensing technology has many advantages in environmental monitoring, its data acquisition methods and processes still have certain limitations. Due to the accuracy of ground information, the influence of environmental factors, resolution limitations and other factors, it is difficult to capture the changes of small pollution sources. Therefore, in practical applications, remote sensing technology still needs to be combined with other pollution monitoring methods to improve the overall identification accuracy and ensure the effectiveness of soil protection work.

[0005] Therefore, an ecological remote sensing detection system for soil contaminated areas is proposed. Summary of the invention

[0006] The purpose of the present invention is to provide an ecological remote sensing detection system for soil pollution areas; determine the ground monitoring area according to the road coverage area and the exhaust gas impact distance; divide the ground sub-area according to the terrain data and environmental data of the ground monitoring area; obtain the exhaust gas emission data according to the model data and driving data of the vehicle in the ground sub-area; obtain the pollution distribution data according to the exhaust gas emission data, terrain data and environmental data; divide the ground sub-area according to the pollution distribution data to obtain the pollution sub-area and pollution sub-data; sample and test the soil in the pollution sub-area to obtain soil pollution data; analyze the plants in the pollution sub-area to obtain ecological abnormality data; obtain the soil pollution coefficient according to the pollution sub-data, soil pollution data and ecological abnormality data of the pollution sub-area; the present invention can accurately identify and warn the soil pollution on both sides of the road according to the soil pollution coefficient.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An ecological remote sensing detection system for soil contaminated areas, comprising:

[0009] The ground area division module identifies the road coverage area and determines the ground monitoring area according to the road coverage area and the exhaust impact distance; obtains the terrain data and environmental data of the ground monitoring area; and divides the ground sub-areas according to the terrain data and environmental data of the ground monitoring area;

[0010] The exhaust pollution identification module obtains exhaust emission data based on the vehicle model data and driving data in the ground sub-area; obtains pollution distribution data based on the exhaust emission data, terrain data and environmental data; divides the ground sub-area according to the pollution distribution data to obtain pollution sub-areas and pollution sub-data;

[0011] A soil pollution test module is used to sample and test the soil in the polluted sub-area to obtain soil pollution data; the soil pollution data includes a first soil pollution index, a second soil pollution index, a third soil pollution index and a fourth soil pollution index; the first soil pollution index is a soil heavy metal content index; the second soil pollution index is a nitrogen oxide content index; the third soil pollution index is a hydrocarbon content index; and the fourth soil pollution index is a particulate matter index;

[0012] The ecological anomaly identification module analyzes the plants in the polluted sub-area through remote sensing images and obtains ecological anomaly data based on plant image feature recognition;

[0013] The soil pollution early warning module obtains the soil pollution coefficient based on the pollution sub-data, soil pollution data and ecological anomaly data of the pollution sub-area; and issues early warning based on the soil pollution coefficient.

[0014] The method for determining the ground monitoring area is:

[0015] Conduct exhaust gas impact distance test according to the road coverage area, and obtain the maximum impact distance as the exhaust gas impact distance;

[0016] Determine the identification direction according to the road coverage area, and determine the left impact area and the right impact area according to the road coverage area, the identification direction and the exhaust impact distance;

[0017] The ground monitoring area is determined based on the road coverage area, the left impact area, and the right impact area.

[0018] The environmental data includes temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data; obtained through IoT sensors;

[0019] The terrain data includes first terrain data, second terrain data and third terrain data; the first terrain data is terrain data of the road coverage area; the second terrain data is terrain data of the left impact area; and the third terrain data is terrain data of the right impact area.

[0020] The process of dividing the ground sub-areas is as follows:

[0021] The first terrain data is divided according to the clustering algorithm and the environmental data to obtain first segmented data; the second terrain data is divided according to the clustering algorithm and the environmental data to obtain second segmented data; the third terrain data is divided according to the clustering algorithm and the environmental data to obtain third segmented data;

[0022] The ground monitoring area is divided according to the first segmented data, the second segmented data and the third segmented data to obtain ground sub-areas.

[0023] The vehicle model data includes engine parameters, fuel type and actual vehicle weight; the vehicle behavior data includes speed data and steering data of the vehicle during the period of being in the ground sub-area.

[0024] The exhaust emission data is obtained through an exhaust prediction model;

[0025] The training process of the exhaust gas prediction model is as follows:

[0026] Select a vehicle to drive on different roads to obtain first test data;

[0027] The first test data includes test road terrain data, test road environment data, test vehicle model data, test vehicle behavior data and collected test exhaust emission data;

[0028] The model is trained and optimized according to the first test data to obtain an exhaust gas prediction model.

[0029] The polluted sub-area and polluted sub-data are obtained through a pollution identification model;

[0030] The pollution identification model includes an exhaust data identification layer, a first pollution identification layer, a second pollution identification layer, a pollution distribution identification layer and a pollution area division layer;

[0031] The exhaust gas data recognition layer recognizes and divides the exhaust gas emission data according to the wind speed data and wind direction data in the environmental data to obtain the left exhaust gas emission data and the right exhaust gas emission data;

[0032] The first pollution identification layer identifies the environmental data, terrain data, and left tail gas emission data of the ground sub-area to obtain left pollution distribution data;

[0033] The second pollution identification layer identifies the environmental data, terrain data, and right tail gas emission data of the ground sub-area to obtain right pollution distribution data;

[0034] The pollution distribution identification layer obtains ground pollution distribution data according to the left pollution distribution data and the right pollution distribution data;

[0035] The pollution area division layer identifies the ground pollution distribution data through a clustering algorithm, divides the ground sub-areas according to the identification results, and obtains pollution sub-areas and pollution sub-data.

[0036] The soil pollution coefficient is obtained through a soil pollution identification model;

[0037] The soil pollution identification model includes a pollution data identification layer, a soil pollution identification layer and a soil pollution calibration layer;

[0038] The pollution data identification layer identifies the pollution sub-data according to the pollution threshold to obtain the pollution factor; clusters the pollution sub-areas according to the pollution factor to divide the pollution sub-areas into multiple pollution sub-area sets;

[0039] The soil pollution identification layer identifies the soil pollution coefficient of the polluted sub-area according to the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological anomaly data and the pollution factor;

[0040] The soil pollution calibration layer performs outlier identification and calibration according to the soil pollution coefficient of the polluted sub-region in the same polluted sub-region set; and outputs the soil pollution coefficient after outlier identification and calibration.

[0041] In the pollution data identification layer, the process of dividing the pollution sub-area set according to the clustering algorithm is as follows:

[0042] Get the pollution factor of the polluted sub-area, and set the data weight for each data dimension in the pollution factor;

[0043] According to the clustering algorithm, the data weights and pollution factors are identified and divided to obtain multiple data clusters; according to the data cluster division, a set of polluted sub-areas is obtained.

[0044] The method for giving early warning according to the soil pollution coefficient is: setting a soil pollution threshold, and identifying and giving early warning to the soil pollution coefficient according to the soil pollution threshold.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. The present invention obtains terrain data of the road coverage area, the left impact area and the right impact area, and then identifies them according to the clustering algorithm and the environmental data to obtain the first segmented data, the second segmented data and the third segmented data; the ground sub-area is obtained by division according to the first segmented data, the second segmented data and the third segmented data; it can effectively ensure that the environmental data and terrain data of the left area, the coverage area and the right area of ​​the divided ground sub-area are similar, respectively, to provide support for soil pollution identification.

[0047] 2. The present invention selects a vehicle to travel on different roads to obtain first test data; trains and optimizes the model based on the first test data to obtain an exhaust gas prediction model; and identifies first terrain data, road environment data, vehicle model data and behavior data of a road sub-area based on the exhaust gas prediction model to accurately obtain the vehicle's exhaust emission data.

[0048] 3. The present invention identifies and divides the exhaust emission data according to the wind speed data and wind direction data in the environmental data to obtain the left exhaust emission data and the right exhaust emission data; identifies the left pollution distribution data according to the environmental data, the terrain data and the left exhaust emission data; identifies the right pollution distribution data according to the environmental data, the terrain data and the right exhaust emission data; accurately identifies the road pollution data according to the left pollution distribution data and the right pollution distribution data; and then divides the pollution sub-areas according to the clustering algorithm to effectively improve the soil pollution identification accuracy of the pollution sub-areas.

[0049] 4. The present invention identifies the polluted sub-data according to the pollution threshold to obtain the pollution factor; divides the polluted sub-area into multiple polluted sub-area sets according to the pollution factor; identifies the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological abnormal data and the pollution factor of the polluted sub-area in the polluted sub-area set to obtain the soil pollution coefficient of the polluted sub-area; then identifies and calibrates the abnormal value of the soil pollution coefficient; and can accurately obtain the soil pollution coefficient reflecting the degree of soil pollution in the polluted sub-area. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a structural schematic diagram of an ecological remote sensing detection system for soil contaminated areas of the present invention;

[0051] Figure 2 It is a structural schematic diagram of the pollution identification model of the present invention;

[0052] Figure 3 It is a structural schematic diagram of the soil pollution identification model of the present invention;

[0053] Figure 4 It is a schematic diagram of segmented data of the present invention;

[0054] Figure 5 It is a schematic diagram of ground area division of the present invention.

[0055] In the figure: 11, first terrain data; 12, second terrain data; 13, third terrain data; 21, first segmented data; 22, second segmented data; 23, third segmented data. DETAILED DESCRIPTION

[0056] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] Embodiment 1

[0058] The present invention proposes an ecological remote sensing detection system for soil contaminated areas, the structure of which is as follows: Figure 1 As shown, it includes: a ground area division module, an exhaust pollution identification module, a soil pollution testing module, an ecological anomaly identification module and a soil pollution early warning module.

[0059] The ground area division module identifies the road coverage area and determines the ground monitoring area according to the road coverage area and the exhaust impact distance; obtains the terrain data and environmental data of the ground monitoring area; and divides the ground sub-areas according to the terrain data and environmental data of the ground monitoring area;

[0060] The method for determining the ground monitoring area is:

[0061] Conduct exhaust gas impact distance test according to the road coverage area, and obtain the maximum impact distance as the exhaust gas impact distance;

[0062] The identification direction is determined according to the road coverage area. The left impact area and the right impact area are determined according to the road coverage area, the identification direction and the exhaust impact distance. The ground monitoring area is determined according to the road coverage area, the left impact area and the right impact area.

[0063] The environmental data includes temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data; obtained through the Internet of Things sensor; the terrain data includes first terrain data, second terrain data and third terrain data; the first terrain data is the terrain data of the road coverage area; the second terrain data is the terrain data of the left impact area; the third terrain data is the terrain data of the right impact area.

[0064] The present invention collects environmental data and terrain data of the ground area where the road is located; through temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data, etc., the environmental data of the road can be accurately measured; through the terrain data of the road coverage area, the terrain data of the left impact area and the right impact area, the terrain data of the road area can be accurately measured, providing a data basis for the emission and distribution identification of automobile exhaust.

[0065] The process of dividing the ground sub-areas is as follows:

[0066] The first terrain data is divided according to the clustering algorithm and the environmental data to obtain the first segmented data; the second terrain data is divided according to the clustering algorithm and the environmental data to obtain the second segmented data; the third terrain data is divided according to the clustering algorithm and the environmental data to obtain the third segmented data; the ground monitoring area is divided according to the first segmented data, the second segmented data and the third segmented data to obtain the ground sub-area.

[0067] The present invention obtains terrain data of the road coverage area, the left impact area and the right impact area, and then identifies them according to the clustering algorithm and the environmental data to obtain the first segmented data, the second segmented data and the third segmented data; the ground sub-area is obtained by division according to the first segmented data, the second segmented data and the third segmented data; it ensures that the environmental data and terrain data of the left area, the coverage area and the right area of ​​the divided ground sub-area are similar, respectively, to provide support for soil pollution identification.

[0068] The exhaust pollution identification module obtains exhaust emission data based on the vehicle model data and driving data in the ground sub-area; obtains pollution distribution data based on the exhaust emission data, terrain data and environmental data; divides the ground sub-area according to the pollution distribution data to obtain pollution sub-areas and pollution sub-data;

[0069] The vehicle model data includes engine parameters, fuel type and actual vehicle weight; the vehicle behavior data includes speed data and steering data of the vehicle during the period of being in the ground sub-area.

[0070] The exhaust emission data is obtained through an exhaust prediction model; the exhaust prediction model is constructed according to a deep neural network;

[0071] The training process of the exhaust gas prediction model is as follows: selecting a vehicle to drive on different roads to obtain first test data;

[0072] The first test data includes test road terrain data, test road environment data, test vehicle model data, test vehicle behavior data and collected test exhaust emission data; the model is trained and optimized according to the first test data to obtain an exhaust prediction model.

[0073] The present invention selects a vehicle to travel on different roads to obtain first test data; trains and optimizes a model based on the first test data to obtain an exhaust gas prediction model; and identifies first terrain data, road environment data, vehicle model data and behavior data of a road sub-area based on the exhaust gas prediction model to accurately obtain the vehicle's exhaust emission data.

[0074] The polluted sub-area and polluted sub-data are obtained through a pollution identification model;

[0075] The pollution identification model is constructed based on a deep neural network, and its structure is as follows: Figure 2 As shown, it includes an exhaust data identification layer, a first pollution identification layer, a second pollution identification layer, a pollution distribution identification layer and a pollution area division layer;

[0076] The exhaust gas data recognition layer recognizes and divides the exhaust gas emission data according to the wind speed data and wind direction data in the environmental data to obtain the left exhaust gas emission data and the right exhaust gas emission data;

[0077] The first pollution identification layer identifies the environmental data, terrain data, and left tail gas emission data of the ground sub-area to obtain left pollution distribution data;

[0078] The second pollution identification layer identifies the environmental data, terrain data, and right tail gas emission data of the ground sub-area to obtain right pollution distribution data;

[0079] The pollution distribution identification layer obtains ground pollution distribution data according to the left pollution distribution data and the right pollution distribution data.

[0080] The pollution area division layer identifies the ground pollution distribution data through a clustering algorithm, divides the ground sub-areas according to the identification results, and obtains pollution sub-areas and pollution sub-data.

[0081] The present invention identifies and divides the exhaust emission data according to the wind speed data and wind direction data in the environmental data, and obtains the left exhaust emission data and the right exhaust emission data; obtains the left pollution distribution data according to the environmental data, terrain data, and the left exhaust emission data; obtains the right pollution distribution data according to the environmental data, terrain data, and the right exhaust emission data; and accurately identifies the road pollution data according to the left pollution distribution data and the right pollution distribution data. Then, the pollution sub-areas are divided according to the clustering algorithm, and the soil pollution identification of the pollution sub-areas is effectively improved.

[0082] A soil pollution test module is used to sample and test the soil in the polluted sub-area to obtain soil pollution data; the soil pollution data includes a first soil pollution index, a second soil pollution index, a third soil pollution index and a fourth soil pollution index; the first soil pollution index is a soil heavy metal content index; the second soil pollution index is a nitrogen oxide content index; the third soil pollution index is a hydrocarbon content index; and the fourth soil pollution index is a particulate matter index;

[0083] The ecological anomaly identification module analyzes the plants in the polluted sub-area through remote sensing images and obtains ecological anomaly data based on plant image feature recognition;

[0084] The soil pollution early warning module obtains the soil pollution coefficient based on the pollution sub-data, soil pollution data and ecological anomaly data of the pollution sub-area; and issues early warning based on the soil pollution coefficient.

[0085] The soil pollution coefficient is obtained through a soil pollution identification model;

[0086] The soil pollution identification model is constructed based on a convolutional neural network, and its structure is as follows: Figure 3 As shown, it includes a pollution data identification layer, a soil pollution identification layer and a soil pollution calibration layer;

[0087] The pollution data identification layer identifies the pollution sub-data according to the pollution threshold to obtain the pollution factor; clusters the pollution sub-areas according to the pollution factor to divide the pollution sub-areas into multiple pollution sub-area sets;

[0088] The soil pollution identification layer identifies the soil pollution coefficient of the polluted sub-area according to the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological anomaly data and the pollution factor;

[0089] The soil pollution calibration layer performs outlier identification and calibration according to the soil pollution coefficient of the polluted sub-region in the same polluted sub-region set; and outputs the soil pollution coefficient after outlier identification and calibration.

[0090] In the pollution data identification layer, the process of dividing the pollution sub-area set according to the clustering algorithm is as follows:

[0091] Get the pollution factor of the polluted sub-area, and set the data weight for each data dimension in the pollution factor;

[0092] According to the clustering algorithm, the data weights and pollution factors are identified and divided to obtain multiple data clusters; according to the data cluster division, a set of polluted sub-areas is obtained.

[0093] The present invention identifies polluted sub-data according to a pollution threshold value to obtain a pollution factor; divides a polluted sub-region into a plurality of polluted sub-region sets according to the pollution factor; identifies a first soil pollution index, a second soil pollution index, a third soil pollution index, a fourth soil pollution index, ecological abnormal data and a pollution factor of the polluted sub-region in the polluted sub-region set to obtain a soil pollution coefficient of the polluted sub-region; then identifies and calibrates abnormal values ​​of the soil pollution coefficient; and accurately obtains a soil pollution coefficient that reflects the degree of soil pollution in the polluted sub-region.

[0094] The method for giving early warning according to the soil pollution coefficient is: setting a soil pollution threshold, and identifying and giving early warning to the soil pollution coefficient according to the soil pollution threshold.

[0095] The present invention determines a ground monitoring area according to a road coverage area and an exhaust gas impact distance; divides the ground monitoring area according to terrain data and environmental data to obtain ground sub-areas; obtains exhaust gas emission data according to vehicle model data and driving data in the ground sub-areas; obtains pollution distribution data according to the exhaust gas emission data, terrain data and environmental data; divides the ground sub-areas according to the pollution distribution data to obtain polluted sub-areas and pollution sub-data; samples and tests the soil in the polluted sub-areas to obtain soil pollution data; analyzes plants in the polluted sub-areas to obtain ecological anomaly data; obtains a soil pollution coefficient according to the pollution sub-data, soil pollution data and ecological anomaly data of the polluted sub-areas; and accurately identifies and warns soil pollution on both sides of the road according to the soil pollution coefficient.

[0096] Embodiment 2

[0097] In order to verify the actual effect of the present invention on soil pollution identification, an actual road area was selected for testing. The testing process according to the actual road area is as follows:

[0098] The ground area division module identifies the road coverage area and determines the ground monitoring area according to the road coverage area and the exhaust impact distance; obtains the terrain data and environmental data of the ground monitoring area; and divides the ground sub-areas according to the terrain data and environmental data of the ground monitoring area;

[0099] The method for determining the ground monitoring area is:

[0100] Conduct exhaust gas impact distance test according to the road coverage area, and obtain the maximum impact distance as the exhaust gas impact distance;

[0101] The identification direction is determined according to the road coverage area. The left impact area and the right impact area are determined according to the road coverage area, the identification direction and the exhaust impact distance. The ground monitoring area is determined according to the road coverage area, the left impact area and the right impact area.

[0102] The environmental data includes temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data; obtained through the Internet of Things sensor; the terrain data includes first terrain data, second terrain data and third terrain data; the first terrain data is the terrain data of the road coverage area; the second terrain data is the terrain data of the left impact area; the third terrain data is the terrain data of the right impact area.

[0103] The present invention collects environmental data and terrain data of the ground area where the road is located; through temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data, etc., the environmental data of the road can be accurately measured; through the terrain data of the road coverage area, the terrain data of the left impact area and the right impact area, the terrain data of the road area can be accurately measured, providing a data basis for the emission and distribution identification of automobile exhaust.

[0104] The process of dividing the ground sub-areas is as follows:

[0105] The first terrain data is divided according to the clustering algorithm and the environmental data to obtain the first segmented data; the second terrain data is divided according to the clustering algorithm and the environmental data to obtain the second segmented data; the third terrain data is divided according to the clustering algorithm and the environmental data to obtain the third segmented data; the ground monitoring area is divided according to the first segmented data, the second segmented data and the third segmented data to obtain the ground sub-area.

[0106] like Figure 4 As shown, the figure includes the first terrain data 11, and from left to right as the identification direction, the second terrain data 12 on the left side of the first terrain data and the third terrain data 13 on the right side of the first terrain data are obtained; the first terrain data 11 includes the first segmentation data 21; the second terrain data 12 includes the second segmentation data 22; the third terrain data 13 includes the third segmentation data 23. The ground monitoring area is divided according to the first segmentation data, the second segmentation data and the third segmentation data to obtain the ground sub-area; as shown in FIG. Figure 5 shown.

[0107] The present invention obtains terrain data of a road coverage area, a left impact area, and a right impact area, and then respectively identifies them according to a clustering algorithm and environmental data to obtain first segmented data, second segmented data, and third segmented data; divides the ground sub-areas according to the first segmented data, the second segmented data, and the third segmented data; ensures that the environmental data and terrain data of the left area, the coverage area, and the right area of ​​the divided ground sub-areas are similar, respectively, to provide support for soil pollution identification.

[0108] The exhaust pollution identification module obtains exhaust emission data based on the vehicle model data and driving data in the ground sub-area; obtains pollution distribution data based on the exhaust emission data, terrain data and environmental data; divides the ground sub-area according to the pollution distribution data to obtain pollution sub-areas and pollution sub-data;

[0109] The vehicle model data includes engine parameters, fuel type and actual vehicle weight; the vehicle behavior data includes speed data and steering data of the vehicle during the period of being in the ground sub-area.

[0110] The exhaust emission data is obtained through an exhaust prediction model;

[0111] The training process of the exhaust gas prediction model is as follows:

[0112] Select a vehicle to drive on different roads to obtain first test data;

[0113] The first test data includes test road terrain data, test road environment data, test vehicle model data, test vehicle behavior data and collected test exhaust emission data;

[0114] The model is trained and optimized according to the first test data to obtain an exhaust gas prediction model.

[0115] The components of the exhaust gas are shown in Table 1.

[0116] Table 1 Exhaust gas composition

[0117]

[0118] The composition of automobile exhaust mainly depends on the fuel type, engine operating conditions and emission control technology.

[0119] Common exhaust components include: carbon monoxide, nitrogen oxides, volatile organic compounds, particulate matter, hydrocarbons, etc.; harmful substances emitted by automobiles may settle in the soil, thereby polluting soil quality and the ecological environment.

[0120] The particulate matter emitted by cars may contain heavy metal elements such as lead, cadmium, copper, and zinc. These heavy metals can be deposited into the soil with rainwater, and long-term accumulation will lead to heavy metal pollution in the soil.

[0121] Nitrogen oxides are converted into nitrates in the soil, increasing soil acidity.

[0122] Hydrocarbons will attach to soil particles, and long-term deposition will change the organic composition of the soil, leading to soil organic pollution. At the same time, particulate matter emitted in exhaust gas not only affects air quality, but may also settle into the soil through precipitation, increasing the particle load of the soil, hindering the aeration of the soil, and affecting the normal growth of plant roots.

[0123] The present invention selects a vehicle to travel on different roads to obtain first test data; trains and optimizes a model based on the first test data to obtain an exhaust gas prediction model; and identifies first terrain data, road environment data, vehicle model data and behavior data of a road sub-area based on the exhaust gas prediction model to accurately obtain the vehicle's exhaust emission data.

[0124] The polluted sub-area and polluted sub-data are obtained through a pollution identification model;

[0125] The pollution identification model includes an exhaust data identification layer, a first pollution identification layer, a second pollution identification layer, a pollution distribution identification layer and a pollution area division layer;

[0126] The exhaust gas data recognition layer recognizes and divides the exhaust gas emission data according to the wind speed data and wind direction data in the environmental data to obtain the left exhaust gas emission data and the right exhaust gas emission data;

[0127] The first pollution identification layer identifies the environmental data, terrain data, and left tail gas emission data of the ground sub-area to obtain left pollution distribution data;

[0128] The second pollution identification layer identifies the environmental data, terrain data, and right tail gas emission data of the ground sub-area to obtain right pollution distribution data;

[0129] The pollution distribution identification layer obtains ground pollution distribution data according to the left pollution distribution data and the right pollution distribution data.

[0130] The pollution area division layer identifies the ground pollution distribution data through a clustering algorithm, divides the ground sub-areas according to the identification results, and obtains pollution sub-areas and pollution sub-data.

[0131] The present invention identifies and divides the exhaust emission data according to the wind speed data and wind direction data in the environmental data, and obtains the left exhaust emission data and the right exhaust emission data; obtains the left pollution distribution data according to the environmental data, terrain data, and the left exhaust emission data; obtains the right pollution distribution data according to the environmental data, terrain data, and the right exhaust emission data; and accurately identifies the road pollution data according to the left pollution distribution data and the right pollution distribution data. Then, the pollution sub-areas are divided according to the clustering algorithm, and the soil pollution identification of the pollution sub-areas is effectively improved.

[0132] A soil pollution test module is used to sample and test the soil in the polluted sub-area to obtain soil pollution data; the soil pollution data includes a first soil pollution index, a second soil pollution index, a third soil pollution index and a fourth soil pollution index; the first soil pollution index is a soil heavy metal content index; the second soil pollution index is a nitrogen oxide content index; the third soil pollution index is a hydrocarbon content index; and the fourth soil pollution index is a particulate matter index;

[0133] The particle index includes particle size and particle concentration.

[0134] The ecological anomaly identification module analyzes the plants in the polluted sub-area through remote sensing images and obtains ecological anomaly data based on plant image feature recognition;

[0135] The soil pollution early warning module obtains the soil pollution coefficient based on the pollution sub-data, soil pollution data and ecological anomaly data of the pollution sub-area; and issues early warning based on the soil pollution coefficient.

[0136] The soil pollution coefficient is obtained through a soil pollution identification model;

[0137] The soil pollution identification model includes a pollution data identification layer, a soil pollution identification layer and a soil pollution calibration layer;

[0138] The pollution data identification layer identifies the pollution sub-data according to the pollution threshold to obtain the pollution factor; clusters the pollution sub-areas according to the pollution factor to divide the pollution sub-areas into multiple pollution sub-area sets;

[0139] The soil pollution identification layer identifies the soil pollution coefficient of the polluted sub-area according to the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological anomaly data and the pollution factor;

[0140] The soil pollution calibration layer performs outlier identification and calibration according to the soil pollution coefficient of the polluted sub-region in the same polluted sub-region set; and outputs the soil pollution coefficient after outlier identification and calibration.

[0141] In the pollution data identification layer, the process of dividing the pollution sub-area set according to the clustering algorithm is as follows:

[0142] Get the pollution factor of the polluted sub-area, and set the data weight for each data dimension in the pollution factor;

[0143] According to the clustering algorithm, the data weights and pollution factors are identified and divided to obtain multiple data clusters; according to the data cluster division, a set of polluted sub-areas is obtained.

[0144] To verify the accuracy of the soil pollution identification model;

[0145] Model 1 is the soil pollution identification model of the present invention, including a pollution data identification layer, a soil pollution identification layer and a soil pollution calibration layer; Model 2 is based on Model 1, removing the soil pollution calibration layer; directly identifying the soil pollution coefficient of the polluted sub-area according to the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological abnormality data and the pollution factor, and outputting it; Model 3 is based on Model 2, not considering the influence of the pollution factors in the soil pollution identification layer on soil pollution, and only identifying through the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index and the ecological abnormality data;

[0146] After obtaining the recognition results of Model 1, Model 2 and Model 3, the similarity between the soil pollution coefficient of all polluted sub-areas and the actual soil pollution coefficient is taken as the recognition accuracy, and Table 2 is obtained.

[0147] Table 2 Soil pollution identification model verification data table

[0148] Model Accuracy Model 1 0.9342 Model 2 0.8953 Model 3 0.8795

[0149] From the data shown in Table 2, it can be seen that model 1 has the best recognition effect.

[0150] The present invention identifies polluted sub-data according to a pollution threshold value to obtain a pollution factor; divides a polluted sub-region into a plurality of polluted sub-region sets according to the pollution factor; identifies a first soil pollution index, a second soil pollution index, a third soil pollution index, a fourth soil pollution index, ecological abnormal data and a pollution factor of the polluted sub-region in the polluted sub-region set to obtain a soil pollution coefficient of the polluted sub-region; then identifies and calibrates abnormal values ​​of the soil pollution coefficient; and accurately obtains a soil pollution coefficient that reflects the degree of soil pollution in the polluted sub-region.

[0151] The method for giving early warning according to the soil pollution coefficient is: setting a soil pollution threshold, and identifying and giving early warning to the soil pollution coefficient according to the soil pollution threshold.

[0152] The present invention determines a ground monitoring area according to a road coverage area and an exhaust gas impact distance; divides the ground monitoring area according to terrain data and environmental data to obtain ground sub-areas; obtains exhaust gas emission data according to vehicle model data and driving data in the ground sub-areas; obtains pollution distribution data according to the exhaust gas emission data, terrain data and environmental data; divides the ground sub-areas according to the pollution distribution data to obtain polluted sub-areas and pollution sub-data; samples and tests the soil in the polluted sub-areas to obtain soil pollution data; analyzes plants in the polluted sub-areas to obtain ecological anomaly data; obtains a soil pollution coefficient according to the pollution sub-data, soil pollution data and ecological anomaly data of the polluted sub-areas; and accurately identifies and warns soil pollution on both sides of the road according to the soil pollution coefficient.

[0153] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An ecological remote sensing detection system for soil contaminated areas, characterized in that: include: A ground area division module identifies a road coverage area and determines a ground monitoring area according to the road coverage area and the exhaust gas impact distance; Acquire terrain data and environmental data of the ground monitoring area; divide the ground monitoring area into ground sub-areas according to the terrain data and environmental data of the ground monitoring area; The exhaust pollution identification module obtains exhaust emission data according to the model data and driving data of the vehicle in the ground sub-area; obtains pollution distribution data according to the exhaust emission data, terrain data and environmental data; divides the ground sub-area according to the pollution distribution data to obtain pollution sub-areas and pollution sub-data; A soil pollution testing module is used to sample and test the soil in the polluted sub-area to obtain soil pollution data; the soil pollution data includes a first soil pollution index, a second soil pollution index, a third soil pollution index and a fourth soil pollution index; the first soil pollution index is a soil heavy metal content index; the second soil pollution index is a nitrogen oxide content index; the third soil pollution index is a hydrocarbon content index; and the fourth soil pollution index is a particulate matter index; An ecological anomaly identification module analyzes the plants in the polluted sub-area through remote sensing images and obtains ecological anomaly data based on plant image feature identification; The soil pollution early warning module obtains the soil pollution coefficient based on the pollution sub-data, soil pollution data and ecological anomaly data of the pollution sub-area; and issues early warning based on the soil pollution coefficient.

2. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The method for determining the ground monitoring area is: Conduct exhaust gas impact distance test according to the road coverage area, and obtain the maximum impact distance as the exhaust gas impact distance; Determine the identification direction according to the road coverage area, and determine the left impact area and the right impact area according to the road coverage area, the identification direction and the exhaust impact distance; The ground monitoring area is determined based on the road coverage area, the left impact area, and the right impact area.

3. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The environmental data includes temperature data, humidity data, wind speed data, wind direction data, oxygen concentration and air pressure data; obtained through IoT sensors; The terrain data includes first terrain data, second terrain data and third terrain data; the first terrain data is terrain data of the road coverage area; the second terrain data is terrain data of the left impact area; and the third terrain data is terrain data of the right impact area.

4. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The process of dividing the ground sub-areas is as follows: The first terrain data is divided according to the clustering algorithm and the environmental data to obtain first segmented data; the second terrain data is divided according to the clustering algorithm and the environmental data to obtain second segmented data; the third terrain data is divided according to the clustering algorithm and the environmental data to obtain third segmented data; The ground monitoring area is divided according to the first segmented data, the second segmented data and the third segmented data to obtain ground sub-areas.

5. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The vehicle model data includes engine parameters, fuel type and actual vehicle weight; the vehicle behavior data includes speed data and steering data of the vehicle during the period of being in the ground sub-area.

6. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The exhaust emission data is obtained through an exhaust prediction model; The training process of the exhaust gas prediction model is as follows: Select a vehicle to drive on different roads to obtain first test data; The first test data includes test road terrain data, test road environment data, test vehicle model data, test vehicle behavior data and collected test exhaust emission data; The model is trained and optimized according to the first test data to obtain the exhaust gas prediction model.

7. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The polluted sub-area and polluted sub-data are obtained through a pollution identification model; The pollution identification model includes an exhaust data identification layer, a first pollution identification layer, a second pollution identification layer, a pollution distribution identification layer and a pollution area division layer; The exhaust gas data recognition layer recognizes and divides the exhaust gas emission data according to the wind speed data and wind direction data in the environmental data to obtain the left exhaust gas emission data and the right exhaust gas emission data; The first pollution identification layer identifies the environmental data, terrain data, and left tail gas emission data of the ground sub-area to obtain left pollution distribution data; The second pollution identification layer identifies the environmental data, terrain data, and right tail gas emission data of the ground sub-area to obtain right pollution distribution data; The pollution distribution identification layer obtains ground pollution distribution data according to the left pollution distribution data and the right pollution distribution data; The pollution area division layer identifies the ground pollution distribution data through a clustering algorithm, divides the ground sub-areas according to the identification results, and obtains pollution sub-areas and pollution sub-data.

8. The ecological remote sensing detection system for soil contaminated areas according to claim 1 is characterized by: The soil pollution coefficient is obtained through a soil pollution identification model; The soil pollution identification model includes a pollution data identification layer, a soil pollution identification layer and a soil pollution calibration layer; The pollution data identification layer identifies the pollution sub-data according to the pollution threshold to obtain the pollution factor; clusters the pollution sub-areas according to the pollution factor to divide the pollution sub-areas into multiple pollution sub-area sets; The soil pollution identification layer identifies the soil pollution coefficient of the polluted sub-area according to the first soil pollution index, the second soil pollution index, the third soil pollution index, the fourth soil pollution index, the ecological anomaly data and the pollution factor; The soil pollution calibration layer performs outlier identification and calibration according to the soil pollution coefficient of the polluted sub-region in the same polluted sub-region set; and outputs the soil pollution coefficient after outlier identification and calibration.

9. The ecological remote sensing detection system for soil contaminated areas according to claim 8, characterized in that: In the pollution data identification layer, the process of dividing the pollution sub-area set according to the clustering algorithm is as follows: Obtaining a pollution factor of a polluted sub-region, and setting a data weight for each data dimension in the pollution factor; The data weight and the pollution factor are identified and divided according to a clustering algorithm to obtain a plurality of data clusters; and a set of polluted sub-regions is obtained according to the data cluster division.

10. The ecological remote sensing detection system for soil contaminated areas according to claim 1, characterized in that: The method for giving early warning according to the soil pollution coefficient is: setting a soil pollution threshold, and identifying and giving early warning to the soil pollution coefficient according to the soil pollution threshold.

Images