An insect swarm density estimation algorithm based on centimeter-wave scanning radar

Abstract

The present invention provides an insect swarm density estimation algorithm based on centimeter-wave scanning radar. By dividing the three-dimensional space into independent resolution units and performing statistics on the detection units that detect the target, the spatial distribution and three-dimensional shape of the insect swarm are better obtained. The relationship between the echo intensity and the effective cross-sectional area of ​​the target reflection in the radar power formula is used to deduce the number of insect individuals in the detection unit. Combined with the distance and speed distribution diagram generated by the Doppler effect, the number of individuals in different detection units is verified, making the statistics of the overall number of insect swarms in the airspace more reasonable and ensuring the accuracy of the final density value estimation.

Description

Technical Field

[0001] The present invention relates to the technical field of digital signal processing, and in particular to an insect swarm density estimation algorithm based on centimeter-wave scanning radar. Background Art

[0002] Insect pests have long been a difficult problem to prevent in agriculture and forestry. Since the Chinese Academy of Agricultural Sciences proposed the "observation before trapping" strategy for insect swarms, agricultural and forestry units across the country have begun implementing it. However, insect monitoring has remained stagnant.

[0003] Because insect swarms are distributed widely in space, the number levels of insects in different areas are also different, and the size of a single insect is small, making it impossible to capture it accurately; at the same time, the measurement accuracy of many monitoring equipment does not meet the detection requirements of insects, resulting in the inability to identify single or multiple insects in the same distance space, making it extremely difficult to accurately grasp the swarm information such as the swarm size and swarm density. Summary of the Invention

[0004] The purpose of the present invention is to address the deficiencies of the above-mentioned existing technologies and provide an insect swarm density estimation algorithm based on centimeter-wave scanning radar to accurately capture the three-dimensional shape and volume of the insect swarm in the airspace and calculate the number of targets in each unit.

[0005] To achieve the above object, the present invention adopts the following technical solutions:

[0006] The present invention provides an insect swarm density estimation algorithm based on centimeter wave scanning radar, comprising the following steps:

[0007] S1. Utilize the high range resolution and high angular resolution of the radar to divide the three-dimensional space into independent spatial units;

[0008] S2. Using the rotation characteristics of the scanning radar, obtain radar echo data at different azimuths at the same pitch angle in a two-dimensional scanning manner;

[0009] S3, collect the radar echo data at all angles after the radar rotates in azimuth for one circle at each pitch working angle;

[0010] S4. Performing digital signal processing on the collected echo data according to a signal processing method of a pulse Doppler radar to obtain original video, background video, and detection video of each wave position;

[0011] S5. Setting a target detection threshold and performing target discrimination on the independent spatial unit to obtain a detection unit;

[0012] S6. Based on the echo intensity corresponding to the detection unit, the effective reflection cross-sectional area of ​​the target in the detection unit is inferred by using the radar power equation, which combines the echo intensity with the reflection cross-sectional area of ​​targets of different sizes at different distances, and the values ​​of azimuth, pitch, distance, and effective reflection cross-sectional area are recorded;

[0013] S7. Based on the known effective reflection cross-sectional area of ​​a single insect, the number of valid targets in different detection units and the total number of insects in the entire airspace can be calculated;

[0014] S8. Finally, use the speed-distance two-dimensional distribution map to perform data verification, eliminate outliers, and then use the density formula to obtain the average density value of the insect swarm in the current airspace and the local space insect swarm density value.

[0015] Furthermore, the bottom length of the independent space unit is R·θ α and the bottom width is R·θ β , the height is ΔD

[0016] Where R is the straight-line distance from the space unit to the radar;

[0017] θ α is the horizontal azimuth resolution angle of the radar antenna beam;

[0018] θ β is the vertical elevation resolution angle of the radar antenna beam;

[0019] The volume of the independent space unit is: ΔV = R·θ α ·R·θ β ΔD

[0020] Among them, ΔR is 1.5m.

[0021] Furthermore, the maximum detection distance R of the radar in the detection unit max and minimum detection signal-to-noise ratio S min The relationship is:

[0022]

[0023] Among them, P t is the radar transmitter power; A e is the effective receiving area of ​​the antenna; σ is the radar cross-section.

[0024] Furthermore, the corresponding relationship between the straight-line distance R corresponding to the detection unit and the effective cross-sectional area Δσ is:

[0025]

[0026] Among them, C r is a constant related to the radar system itself.

[0027] Furthermore, the effective cross-sectional area Δσ is:

[0028] Δσ=σ0·W e ·N t

[0029] Where σ0 is the reflection cross-sectional area of ​​a single insect;

[0030] We is the reflection attenuation coefficient of a single insect in different postures;

[0031] N t is the number of insects in the distance resolution unit.

[0032] Furthermore, the number of insects N in the detection unit at different straight-line distances R t The corresponding relationship between echo energy S and noise power N is:

[0033]

[0034] Furthermore, the number of insects in the detection unit and the volume of the detection unit are added to obtain the total number of insects N tAll for:

[0035]

[0036] The volume of space occupied by the insect in the airspace ΔV All for:

[0037]

[0038]

[0039]

[0040] The beneficial effects of the present invention are: through two-dimensional scanning, namely, pitch and azimuth; the exploration of the airspace is realized, and the narrow beam characteristics of the antenna are utilized to effectively avoid echo interference in other angular directions during the antenna scanning process;

[0041] Based on the principle that "the longer the wavelength, the stronger the penetrating power", the S band of centimeter waves was selected instead of millimeter waves. Since the wavelength of millimeter waves is relatively close to the body size of individual insects, millimeter waves are mostly used to detect single insects. Centimeter waves, on the other hand, have a certain penetrating power over insect swarms, ensuring that insects close to the radar antenna beam in a straight line can be detected while insects at a longer distance can also be detected.

[0042] Since the swarm is constantly moving, it may appear in different airspaces at different times. If the antenna scanning cycle is too long, the same target may be scanned multiple times in different areas. Based on the characteristics of the swarm being distributed at different altitudes during its migration, different pitch angles are selected to detect different altitudes, reducing the pitch detection time and solving the problem of long detection time at all pitch angles. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a block diagram of an insect swarm density estimation algorithm based on centimeter-wave scanning radar according to the present invention;

[0044] Figure 2 Schematic diagram of antenna beam transmission;

[0045] Figure 3 It is a schematic diagram of the three-dimensional resolution unit space;

[0046] Figure 4 is the spatial height distribution diagram corresponding to different pitch angles;

[0047] Figure 5 Design a graph for minimum detection signal-to-noise ratio;

[0048] Figure 6 This is the three-dimensional spatial distribution map of the insect swarm;

[0049] Figure 7 The insect swarm display image observed by the terminal software at a certain elevation angle of the position;

[0050] Figure 8 Display of a flock of birds observed by the terminal software at a certain elevation angle of the position. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] See also Figure 1 , an insect swarm density estimation algorithm based on centimeter-wave scanning radar, comprising the following steps:

[0053] S1. Utilize the high range resolution and high angular resolution of the radar to divide the three-dimensional space into independent spatial units;

[0054] S2. Using the rotation characteristics of the scanning radar, obtain radar echo data at different azimuths at the same pitch angle in a two-dimensional scanning manner;

[0055] S3, collect the radar echo data at all angles after the radar rotates in azimuth for one circle at each pitch working angle;

[0056] S4. Performing digital signal processing on the collected echo data according to a signal processing method of a pulse Doppler radar to obtain original video, background video, and detection video of each wave position;

[0057] S5. Setting a target detection threshold and performing target discrimination on the independent spatial unit to obtain a detection unit;

[0058] S6. Based on the echo intensity corresponding to the detection unit, the effective reflection cross-sectional area of ​​the target in the detection unit is inferred by using the radar power equation, which combines the echo intensity with the reflection cross-sectional area of ​​targets of different sizes at different distances, and the values ​​of azimuth, pitch, distance, and effective reflection cross-sectional area are recorded;

[0059] S7. Based on the known effective reflection cross-sectional area of ​​a single insect, the number of valid targets in different detection units and the total number of insects in the entire airspace can be calculated;

[0060] S8. Finally, use the speed-distance two-dimensional distribution map to perform data verification, eliminate outliers, and then use the density formula to obtain the average density value of the insect swarm in the current airspace and the local space insect swarm density value.

[0061] The bottom length of the independent space unit is R·θ α and the bottom width is R·θ β , the height is ΔD

[0062] Where R is the straight-line distance from the space unit to the radar;

[0063] θ α is the horizontal azimuth resolution angle of the radar antenna beam;

[0064] θ β is the vertical elevation resolution angle of the radar antenna beam;

[0065] The volume of the independent space unit is: ΔV = R·θ α ·R·θ β ΔD

[0066] Among them, ΔR is 1.5m.

[0067] The maximum detection distance R of the radar in the detection unit max and minimum detection signal-to-noise ratio S min The relationship is:

[0068]

[0069] Among them, P t is the radar transmitter power; A e is the effective receiving area of ​​the antenna; σ is the radar cross-section.

[0070] The corresponding relationship between the straight-line distance R corresponding to the detection unit and the effective cross-sectional area Δσ is:

[0071]

[0072] Among them, C r is a constant related to the radar system itself.

[0073] The effective cross-sectional area Δσ is:

[0074] Δσ=σ0·W e ·N t

[0075] Where σ0 is the reflection cross-sectional area of ​​a single insect;

[0076] We is the reflection attenuation coefficient of a single insect in different postures;

[0077] N t is the number of insects in the distance resolution unit.

[0078] The number of insects N in the detection unit at different straight-line distances R t The corresponding relationship between echo energy S and noise power N is:

[0079]

[0080] The number of insects in the detection unit and the volume of the detection unit are added together to obtain the total number of insects N tAll for:

[0081]

[0082] The volume of space occupied by the insect in the airspace ΔV All for:

[0083]

[0084] Among them, i is the data under the pitch angle, and j is the data under the horizontal 360 degrees.

[0085] The density value β of the insect group in the entire space within the airspace mean for:

[0086]

[0087] Example 1

[0088] The migration altitude of insect swarms is generally less than 500m. Different species have different distribution layers, some are distributed between 80-240m, some between 100-300m, some between 200-300m, and some between 400-500m. The migration altitude layers of common insect species are shown in Table 1.

[0089] Table 1 Swarm flight altitude

[0090]

[0091] However, if there is strong wind, the swarm can follow the airflow to fly to higher altitudes, up to 800m, 1000m, 1200m, or even 1800m, 2000m;

[0092] Therefore, multiple gears are designed in this method, including low-altitude area, mid-high-altitude area, and high-altitude area. The height ranges and pitch angles of different gears are shown in Table 2.

[0093] Table 2 Radar detection altitude range

[0094]

[0095]

[0096] Given that the maneuverability of multiple insects in a spatial unit at a certain distance during swarm flight is complex, and that it is difficult to detect tiny insects, this method proposes the concept of "making the big smaller and the small bigger," dividing the three-dimensional airspace into independent distance-resolution units for monitoring:

[0097] In order to avoid the radar antenna being interfered with by echoes at other angles during scanning, this method uses horizontal and vertical θ 3dB The 1.9° pencil-shaped narrow beam parabolic antenna ensures high gain in the main lobe direction while reducing the interference from the side lobe angle direction. The antenna gain diagram is shown in Figure 2 :

[0098] In order to avoid the influence of different movement properties of insects due to the large number of insects in the distance resolution unit, the distance resolution unit of the sampling should not be too large. The radar system with a distance resolution ΔD of 1.5m is selected in this method. The schematic diagram of the specific independent spatial unit is shown in Figure 3 :

[0099] The volume of an independent space unit ΔV = R·θ α ·R·θ β ΔD;

[0100] Since the horizontal and elevation angles of the antenna beam are both 1.9°, we can finally get the three-dimensional unit at distance R.

[0101] The distance resolution unit is an independent spatial unit. The echo energy it feeds back to the radar is equal to the vector sum of all insect individuals in the unit. The energy is proportional to the number of insects. By referring to the relationship between echo energy, target reflection cross-sectional area, and target distance in the radar equation, the number of insects in different distance units can be deduced:

[0102] Determine the working parameters and technical indicators of the insect radar system, as shown in Table 3;

[0103] Table 3 Radar operating parameters

[0104]

[0105] According to the selection method of different altitude layers of the swarm flight, the radar is completed at the pitch angles of 6°, 12°, 18°, 24°, 30°, 32°, 42°, 53°, and 64°, and a rapid horizontal rotation is performed for detection, such as Figure 4 As shown;

[0106] Collect echo data at different azimuths and elevations n , sorted in order according to the array arrangement [X1, X2, ..., X n ];

[0107] Perform radar pulse compression, moving target detection, and constant false alarm detection on the echo data, and then calculate the minimum detection threshold S min , the distance resolution unit where the target appears is identified (the distance resolution unit where the target exists is hereinafter referred to as the detection unit);

[0108] Minimum detection threshold S min The design principle is shown in Figure 5 ;

[0109] S min =C·N mean

[0110] Among them, C is the threshold detection coefficient, N mean is the average noise power.

[0111] To prevent the appearance of "outliers" in the echo caused by external interference, which would result in large echo energy in a certain range resolution unit and ultimately identify the wrong target; or the insect in the unit is too small and the echo energy is too weak to be detected;

[0112] This method uses the combination of "coherent integration and Doppler velocity" to increase the signal-to-noise ratio while also being able to separate targets in the same range unit according to different movement speeds.

[0113] Convert the one-dimensional range domain echo data into the two-dimensional Doppler velocity-range domain,

[0114] When the number of targets with different speeds and the number of detected insects N appear in a two-dimensional distribution at a certain distance unit t , when there is a large difference, this unit can be considered as an outlier and will not be tested;

[0115] The density value ρ in the detection unit after final verification is calculated;

[0116] Count the number of targets N in all detection units in the beam at the current azimuth and elevation t [N t1 、N t2 ,…N tn ], space volume ΔV[V1, V2, ... V n ] and the density value ρ[ρ1, ρ2, …ρ n ];

[0117] Similarly, count the current pitch angle, rotate one circle, and detect how many distance units have targets, as well as the number of targets calculated by each detection unit N. t and the corresponding spatial volume ΔV;

[0118] Similarly, the number of insect swarms N in all detection units after each pitch working angle is counted and the azimuth rotates one circle is t and spatial volume ΔV;

[0119] Since this method also obtains the density value ρ of each detection unit in the process of calculating the final airspace insect swarm density value, the density value is summed and averaged according to the distribution of the insect swarm flight altitude layer (at different pitch angles) to obtain the average insect swarm density ρ at different altitude layers. h ;

[0120] Similarly, once the density value of each detection unit is obtained, three-dimensional space mapping can be performed according to actual needs. By substituting the density value into the specific space unit, the distribution of the insect swarm in the airspace can be clearly understood. Figure 6 As shown:

[0121] The workflow of the specific swarm density estimation algorithm is detailed in Figure 1 , the actual radar altitude detection effect diagram is shown in Figure 7 、 Figure 8 .

[0122] The above-described embodiments merely illustrate the implementation methods of the present invention. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. An insect swarm density estimation algorithm based on centimeter wave scanning radar, characterized in that: The following steps are involved: S1. Utilize the high range resolution and high angular resolution of the radar to divide the three-dimensional space into independent spatial units; S2. Using the rotation characteristics of the scanning radar, obtain radar echo data at different azimuths at the same pitch angle in a two-dimensional scanning manner; S3, collect the radar echo data at all angles after the radar rotates in azimuth for one circle at each pitch working angle; S4. Performing digital signal processing on the collected echo data according to a signal processing method of a pulse Doppler radar to obtain original video, background video, and detection video of each wave position; S5. Setting a target detection threshold and performing target discrimination on the independent spatial unit to obtain a detection unit; S6. Based on the echo intensity corresponding to the detection unit, the effective reflection cross-sectional area of ​​the target in the detection unit is inferred by using the radar power equation, which combines the echo intensity with the reflection cross-sectional area of ​​targets of different sizes at different distances, and the values ​​of azimuth, pitch, distance, and effective reflection cross-sectional area are recorded; S7. Based on the known effective reflection cross-sectional area of ​​a single insect, the number of valid targets in different detection units and the total number of insects in the entire airspace can be calculated; S8. Finally, use the speed-distance two-dimensional distribution graph to perform data verification, remove outliers, and then use the density formula to obtain the average density value of the insect swarm in the current airspace and the local space insect swarm density value; The bottom length of the independent space unit is R·θ α and the bottom width is R·θ β , the height is ΔD; Where R is the straight-line distance from the space unit to the radar; θ α is the horizontal azimuth resolution angle of the radar antenna beam; θ β is the vertical elevation resolution angle of the radar antenna beam; The volume of the independent space unit is: ΔV = R·θ α ·R·θ β ΔD; Among them, ΔR is 1.5m; The maximum detection distance R of the radar in the detection unit max and minimum detection signal-to-noise ratio S min The relationship is: Among them, P t is the radar transmitter power; A e is the effective receiving area of ​​the antenna; σ is the radar cross section; The corresponding relationship between the straight-line distance R corresponding to the detection unit and the effective cross-sectional area Δσ is: Among them, C r is the constant related to the radar system itself; The effective cross-sectional area Δσ is: Δσ=σ0·W e ·N t ; Where σ0 is the reflection cross-sectional area of ​​a single insect; W e is the reflection attenuation coefficient of a single insect in different postures; N t is the number of insects in the distance resolution unit; The number of insects N in the detection unit at different straight-line distances R t The corresponding relationship between echo energy S and noise power N is: The number of insects in the detection unit and the volume of the detection unit are added together to obtain the total number of insects N tAll for: The volume of space occupied by the insect in the airspace ΔV All for: Among them, i is the data under the pitch angle, and j is the data under the horizontal 360 degrees; The density value β of the insect group in the entire space within the airspace mean for:

Images