A three-coordinate spherical tracing algorithm based on vector synthesis

Abstract

The gas emission source tracing method described in the present invention is based on the vector synthesis method and the variable step length formula. The drone records the concentration 1 at the initial position and determines the xyz axis. The concentration of the xyz axis at a unit distance is tested respectively, and an upwind unit vector is synthesized according to the three concentration values. An initial step length step1 is advanced in the direction of the vector, and the concentration 2 at the point is measured again. The variable step length conversion is performed according to the concentration change with 1, and the initial step length is covered as step2. The algorithm is continuously iterated to gradually approach the pollution source. This algorithm does not require synchronization of wind direction information during use, can effectively avoid falling into the local optimal problem, and allows the drone to be launched at a location far away from the emission source, so as to locate the location of the gas emission source in a shorter time.

Description

Technical Field

[0001] The invention relates to a gas emission source positioning method, and in particular to a three-coordinate spherical source tracing algorithm based on vector synthesis. Background Art

[0002] For the problem of gas emission source tracing, the current mainstream tracing algorithms are divided into two types: active olfactory method, which is mainly composed of simplex traversal algorithms such as Zigzag and Sprial, BAS algorithm, hill climbing algorithm and other intelligent search algorithms; static positioning method, which is guided by statistical theory and optimization theory and mainly uses mathematical methods. The active olfactory method takes a lot of time to find the smoke plume, and then combines the wind direction and wind speed information to search against the wind, and confirm the emission source according to some termination conditions. This type of method requires real-time detection of wind speed and direction, and because of the interference of ground buildings and the existence of turbulence, drones are prone to fall into local optimal solutions, which is inefficient and cannot be accurately traced. The tracing principle of the static positioning method is to use a preset sensor array and a gas diffusion model to solve the possible location of the gas emission source through mathematical methods, but due to the mathematical properties of the statistical theory itself, its accuracy is not high. For example, the Chinese patent publication number CN201510784619.0, a method for locating indoor odor sources based on a mobile robot, requires real-time monitoring of wind direction information during the odor source search process, and the tracing environment is a two-dimensional plane, which is not suitable for tracing emission sources in a three-dimensional space. For example, the Chinese patent publication number CN 110927342 B, a method for tracing the source of atmospheric pollutants based on a longhorn beetle whisker search algorithm, can only perform tracing operations on a preset height plane.

[0003] The present invention proposes a three-coordinate spherical tracing algorithm based on vector synthesis, which can effectively prevent the UAV from falling into local optimality during the search process without relying on wind direction and wind speed information. The tracing speed is significantly higher than that of the existing intelligent search algorithms, and can perform high-altitude tracing and accurately and quickly locate the position of the gas emission source. Summary of the invention

[0004] The technical problem to be solved by the present invention is how to develop a suitable gas emission source tracing algorithm and apply it to UAVs, combining the actual situation that the carbon emission source may exist in a huge range and gas emission source tracing without relying on wind direction and wind speed information. Therefore, we provide a three-coordinate spherical tracing algorithm based on vector synthesis. This method has the advantages of accurate and fast tracing, high search efficiency, optimal path and avoiding falling into local optimal solution.

[0005] In order to achieve the above object, the technical solution adopted by the present invention comprises the following steps:

[0006] Step 1: Determine the scope of emission sources;

[0007] Step 2: The drone takes off from any point around the possible range of the emission source, with the nose in any direction during takeoff;

[0008] Step 3: The drone climbs to a certain height and tests the gas concentration at that point;

[0009] Step 4: The drone determines the xyz axis and moves forward a standard distance in the direction of the coordinate axis to test the target gas concentration and synthesize a unit vector based on the gas concentration;

[0010] Step 5: Move forward an initial step in this direction and test the gas concentration at the current point again. Change the step size based on the difference between the two results, overwriting the initial step size.

[0011] Step 6: The drone repeats steps 4 and 5, iterates the three-coordinate spherical algorithm, and gradually approaches the emission source;

[0012] Step 7: When the drone moves back and forth in a small area near a certain point, it is considered that the emission source has been found.

[0013] Before tracing the source of gas emissions, there is no need to measure wind speed and direction in advance. The drone can be placed at any location within a large range around the emission source, with any nose direction and release height.

[0014] In step 4, the gas sensor measures each set of concentration data, and the data obtained is the gas concentration value C at one unit in the positive direction of the xyz axis of the current drone location. x , C y and C z .

[0015] The iterative process of the three-coordinate spherical tracing algorithm in step 6 is:

[0016] The current position of the drone is (x n ,y n , z n ), the current position concentration is C n .

[0017] The drone tests the concentration at (x+step, y, z)(x, y+step, z)(x, y, z+step) respectively, denoted as C nx , C ny and C nz .

[0018] The resultant vector (C nx -C n , C ny -C n , C nz -C n ) and take a step forward in that direction n .

[0019] The current position of the drone is (x n+1 ,y n+1 , z n+1 ), the current position concentration is C n+1 .

[0020] Calculate C n+1 -C n , bring in the variable step length formula:

[0021] step n+1 =step n ×k×(C n+1 -C n )

[0022] In the formula, k is the variable step size coefficient

[0023] Drones are tested separately (x+step n+1 ,y,z)(x,y+step n+1 ,z)(x,y,z+step n+1 ) is denoted as C (n+1)x , C (n+1)y and C (n+1)z .

[0024] The resultant vector (C (n+1)x -C n+1 , C (n+1)y -C n+1 , C (n+1)z -C n+1 ) and take a step forward in that direction n+1 . BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic diagram of the traceability process of the three-coordinate spherical traceability algorithm for vector synthesis;

[0026] Figure 2 It is the route map of the drone tracing process; DETAILED DESCRIPTION

[0027] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0028] Figure 1 The figure shows a schematic flow chart of the three-coordinate spherical tracing algorithm based on vector synthesis in this embodiment. The specific implementation process includes the following steps:

[0029] Step 1: Determine the scope of emission sources.

[0030] In this step, the emission source area is usually set as an industrial area, especially an area where facilities with huge carbon emissions such as power generation facilities exist.

[0031] Step 2: The drone can be launched at the nearest monitoring site and the nose can be placed in any direction.

[0032] The drone then climbs to a certain height to test the gas concentration at that point;

[0033] In this step, before tracing the source, there is no need to measure the wind direction and wind speed in advance, nor is there any need to move the drone and its supporting equipment around the search area.

[0034] Step 3: Start the iterative tracing algorithm, that is, the drone determines the xyz axis and moves forward a standard distance in the direction of the coordinate axis to test the target gas concentration. A unit vector is synthesized based on the gas concentration, and an initial step is moved forward in the direction. The gas concentration at the current point is tested again, and the step size is changed according to the difference between the two results to cover the initial step size.

[0035] The iterative process of the tracing algorithm in this step is:

[0036] The current position of the drone is (x n ,y n , z n ), the current position concentration is C n .

[0037] The drone tests the concentration at (x+step, y, z)(x, y+step, z)(x, y, z+step) respectively, denoted as C nx , C ny and C nz .

[0038] The resultant vector (C nx -C n , C ny -C n , C nz -C n ) and take a step forward in that direction n .

[0039] The current position of the drone is (x n+1 ,y n+1 , z n+1 ), the current position concentration is C n+1 .

[0040] Calculate C n+1 -C n , bring in the variable step length formula:

[0041] step n+1 =step n×k×(C n+1 -C n )

[0042] In the formula, k is the variable step size coefficient

[0043] Drones are tested separately (x+step n+1 ,y,z)(x,y+step n+1 ,z)(x,y,z+step n+1 ) is denoted as C (n+1)x , C (n+1)y and C (n+1)z .

[0044] The resultant vector (C (n+1)x -C n+1 , C (n+1)y -C n+1 , C (n+1)z -C n+1 ) and take a step forward in that direction n+1 .

[0045] Step 4: Determine whether step is less than the predetermined value;

[0046] Step 5: When step is less than the preset value, it is considered that the emission source has been found; Step 6: The drone returns.

[0047] The route map of the drone during the tracing process is as follows: Figure 2 shown.

Claims

1. A three-coordinate spherical tracing algorithm based on vector synthesis, The following steps are involved: Step 1: Determine the scope of emission sources; Step 2: The drone takes off from any point around the possible range of the emission source, with the nose in any direction during takeoff; Step 3: The drone climbs to a certain height and tests the gas concentration at that point; Step 4: The drone determines the xyz axis and moves forward a standard distance in the direction of the coordinate axis to test the target gas concentration and synthesize a unit vector based on the gas concentration; Step 5: Move forward an initial step in this direction and test the gas concentration at the current point again. Change the step size based on the difference between the two results to cover the initial step size. Step 6: The drone repeats steps 4 and 5, iterates the three-coordinate spherical algorithm, and gradually approaches the emission source; Step 7: When the drone moves back and forth in a small area near a certain point, it is considered that the emission source has been found; The iterative process of the three-coordinate spherical algorithm in step 6 is: The current position of the drone is (x n ,y n , z n ), the current position concentration is C n ; The drone tests the concentration at (x+step, y, z)(x, y+step, z)(x, y, z+step) respectively, denoted as C nx , C ny and C nz ; The resultant vector (C nx -C n , C ny -C n , C nz -C n ) and take a step forward in that direction n ; The current position of the drone is (x n+1 ,y n+1 , z n+1 ), the current position concentration is C n+1 ; Calculate C n+1 -C n , bring in the variable step length formula: step n+1 =step n ×k×(C n+1 -C n ) In the formula, k is the variable step length coefficient; Drones are tested separately (x+step n+1 ,y,z)(x,y+step n+1 ,z)(x,y,z+step n+1 ) is denoted as C (n+1)x , C (n+1)y and C (n+1)z ; The resultant vector (C (n+1)x -C n+1 , C (n+1)y -C n+1 , C (n+1)z -C n+1 ) and take a step forward in that direction n+1 .

Images