Insecticidal lamp Internet of Things deployment optimization method and system in combination with video monitoring

Abstract

The invention discloses an insecticidal lamp Internet of Things deployment optimization method and system combined with video monitoring, and the method comprises the steps: building an omnidirectional perception model of an insecticidal lamp and a direction perception model of a camera, carrying out the meshing of a farmland, dividing the farmland into a coverage region and an obstacle region through the irregularity of the farmland, and converting a problem into a secondary coverage problem, and designing a double-layer optimization algorithm for solving to realize an optimal deployment mode. The camera is integrated in the insecticidal lamp, and as a novel insecticidal lamp internet-of-things system, the system can detect intruders while killing pests.

Description

Technical Field

[0001] The present invention relates to the field of agricultural Internet of Things coverage, and in particular to an insecticidal lamp Internet of Things deployment optimization method and system combined with video monitoring. Background Art

[0002] Agriculture is the pillar of economic and social stability. However, sudden disasters caused by agricultural pests and diseases, especially migratory pests such as locusts, grasshoppers, and armyworms, are still very serious. At present, chemical pesticides are still the main means of preventing and controlling pests and diseases, which not only seriously pollute the soil, water, and air, causing pesticide residue problems, but also lead to a decrease in insect resistance and control effects. Solar insecticidal lamps are a technology that uses the photosensitive properties of insects to release a specific spectrum of attracting light sources that are sensitive to agricultural migratory pests. When pests fly to the metal mesh around the lamp, high voltage electricity is released to kill the pests. They have the advantages of being green, pollution-free, effective against adult insects, and having outstanding long-term control effects, and are highly valued by agricultural production enterprises and scientific research institutions. At the same time, with the vigorous development of wireless sensor networks, wireless communication modules are installed on insecticidal lamps, insecticidal lamp nodes are constructed, and large-scale deployment is carried out to form a new type of agricultural Internet of Things, which has become a new type of migratory pest monitoring and control technology.

[0003] However, with the increasing application of new agricultural IoT in agricultural production, there are more and more reports of theft and vandalism. In the absence of appropriate anti-theft and anti-vandalism methods, it is easy to cause significant economic losses. After theft and vandalism problems occur, the effectiveness of pest control is greatly reduced. Therefore, how to deploy insecticidal lamps in outdoor farmland with the lowest deployment cost of new agricultural IoT nodes, which can kill pests and detect intruders at the same time. Summary of the invention

[0004] The present invention aims to provide an optimization method and system for the deployment of insecticidal lamps combined with video surveillance. By utilizing the phototropism of insects, insecticidal lamps are deployed in farmland to eliminate pests and prevent pests. Cameras are deployed on some insecticidal lamp nodes to monitor intruders and prevent theft and damage. This new agricultural Internet of Things deployment optimization system can detect intruders while killing pests.

[0005] In order to solve the above technical problems, the specific solution adopted by the present invention is: an optimal deployment method with the lowest total cost of insecticidal lamps and cameras, including establishing an omnidirectional perception model of insecticidal lamps and a direction perception model of cameras, and then proposing a two-layer optimization algorithm to solve the coverage problem, which consists of solving two mixed integer programming models in sequence.

[0006] Further, the following steps are included:

[0007] S1: Establishing an omnidirectional perception model of the insecticidal lamp according to the coverage characteristics of the insecticidal lamp;

[0008] S2: Establishing a direction perception model of the camera according to the coverage characteristics of the camera;

[0009] S3: According to the irregularity of the farmland area, the farmland is gridded and divided into feasible areas and obstacle areas. Insecticide lamps and cameras are prohibited from being deployed in obstacle areas.

[0010] S4 designs a coverage model based on the omnidirectional perception model of insecticidal lamps established in S1 and the irregularity of farmland areas in S3. Insecticidal lamps can only be deployed in the grid center of feasible farmland areas, and cameras can only be deployed on insecticidal lamps;

[0011] S5: Design a two-layer optimization algorithm to solve the deployment position of the insecticidal lamp in the farmland starting from the convex corner and deploying from the outside to the inside. Secondly, the deployment position of the insecticidal lamp is used as input to calculate the camera position.

[0012] Furthermore, the coverage characteristics of the S1 insecticidal lamp establish the omnidirectional perception model of the insecticidal lamp as a disc perception model. The disc perception model is:

[0013]

[0014] In the formula, represents the Euler distance between the insecticidal lamp and the center of the grid, r s Indicates the effective radius of the insecticidal lamp.

[0015] Furthermore, the coverage characteristics of the S2 camera establish the direction perception model of the camera as a fan-shaped perception model. The fan-shaped perception model is:

[0016]

[0017] In the formula, and is the camera node c k Nodes with insecticidal lamps j The distance between the positive horizontal axis and o kl The angle in the counterclockwise direction.

[0018] Furthermore, the S3 irregular farmland consists of a feasible area and an obstacle area. The S5 two-level optimization algorithm consists of solving two mixed integer programming models in sequence.

[0019] Furthermore, the farmland is gridded in S3, and the farmland is divided into a feasible area and an obstacle area. Insecticide lamps and cameras are prohibited from being deployed in the obstacle area. According to the irregularity of the farmland area, it can be manually set to meet the requirements of different farmlands in practical applications. The farmland is modeled as a cross-connected grid point area. The center point of the cell in the feasible area is defined as a set P, and the insecticide lamp needs to cover all points in the set P. Note that the grid division standard is determined by the user. If high accuracy is required, K is set to a larger value at the cost of higher computational complexity, while a smaller K is set to achieve lower time complexity.

[0020] Furthermore, the omnidirectional perception model of the insecticidal lamp established in S4 and the irregularity of the farmland area in S3.

[0021] Furthermore, the model of insecticidal lamp covering farmland in S40 is a regional coverage model. The purpose of coverage is that the insecticidal lamp attracts insects. Based on the omnidirectional perception model in S1, any point P in the feasible area P i The coverage rate, that is, the probability that pests are captured by insecticidal lamp nodes, can be expressed as:

[0022]

[0023] Where λ i Indicates P i The probability of being captured by S, S j represents the jth insecticidal lamp node in the insecticidal lamp set S, λ ij represents the probability that the i-th node is covered by the j-th insecticidal lamp.

[0024] Therefore, the area coverage can be calculated as

[0025]

[0026] Where λ represents the coverage rate of the farmland by insecticidal lamps, p i represents the i-th grid node in the feasible region P, and K represents the total number of grids in the farmland.

[0027] Furthermore, the S41: camera coverage model of insecticidal lamps is a target coverage model, the purpose of which is to monitor the theft and destruction of insecticidal lamp nodes, so all nodes where insecticidal lamps are deployed should be covered by installed cameras. Based on the S2 direction perception model, any insecticidal lamp node s in the insecticidal lamp set S j The probability of being covered by the camera set can be expressed as:

[0028]

[0029] Where ρ j Indicates jThe probability of being covered by the camera in C, c k represents the kth camera in the camera set C, where C represents the camera set, c l It is consistent with the lth direction in the camera direction set O. jkl Indicates that the jth insecticidal lamp is c along the l direction k Probability of camera coverage

[0030] Therefore, the target coverage for ROI can be calculated as

[0031]

[0032] Where ρ represents the probability of the insecticidal lamp set being covered, s j represents the jth insecticidal lamp in the insecticidal lamp set S, ρ j represents the probability that the jth insecticidal lamp is covered, and M represents the total number of insecticidal lamps.

[0033] Furthermore, the S5 two-layer optimization algorithm solves the covering problem by solving two mixed integer programming models in sequence.

[0034] Furthermore, the main algorithm of the two-layer optimization algorithm solves the regional coverage problem and deploys insecticidal lamp nodes to eliminate pests, that is, under the constraint of regional coverage λ=1, the deployment position of insecticidal lamps is optimized.

[0035] The objective function is designed as follows:

[0036]

[0037] Where s j represents the jth insecticidal lamp in the insecticidal lamp set S, C s represents the cost of an insecticide lamp, x j Indicates whether insecticidal lamps are deployed on farmland grid node j.

[0038] Constrained by

[0039]

[0040] Where s i represents the i-th insecticidal lamp in the insecticidal lamp set S, w ij represents the probability that farmland grid point i is covered by insecticidal lamp node j, x j Indicates whether insecticidal lamps are deployed on farmland grid node j. i represents the i-th feasible area in the farmland P.

[0041] Furthermore, when solving the main problem, farmland is usually an irregular polygon due to the complex terrain. The feasible region is divided into grids, and we define the points in the feasible region as multiple groups. We start from the first convex vertex and deploy insecticidal lamp nodes in sequence to cover all convex vertices. After the first set of convex vertex sets is covered, the above deployment strategy is repeated on the remaining insecticidal lamp nodes to cover the remaining point groups, and the deployed insecticidal lamp nodes in the insecticidal lamp set S are updated until all points in the feasible region are covered.

[0042] The model design to solve the main problem is as follows:

[0043]

[0044] In the formula It represents the best deployment position of insecticidal lamp, p j represents the jth point in the grid p.

[0045] Furthermore, the sub-algorithm of the two-layer optimization algorithm solves the target coverage problem and deploys camera nodes to monitor insecticidal lamps, that is, optimizes the deployment position of the camera under the constraint of target coverage rate ρ=1.

[0046] The objective function is designed as follows:

[0047]

[0048] Where C k represents the kth insecticidal lamp in the insecticidal lamp set C, C c represents the cost of a camera, z k Indicates whether an insecticidal lamp is deployed on insecticidal lamp node k.

[0049] Furthermore, when solving the sub-problem, the fan-shaped sensing range of the camera is a directional sensing model that can only observe a certain angle. Since the deployment positions of the insecticidal lamp nodes are discrete, the number of observed insecticidal lamp nodes may vary with the placement angle of the camera. The optimal deployment of the camera requires not only decisions on the placement of the camera, but also decisions on the perception direction. An optimal deployment strategy that covers all insecticidal lamp nodes and minimizes the number of cameras is proposed. A polygon is formed by connecting the surrounding insecticidal lamp nodes together, starting from the first insecticidal lamp node, and deploying cameras in sequence to cover all insecticidal lamp nodes.

[0050] Furthermore, the deployed system can simultaneously kill pests and detect intruders with minimal total cost.

[0051] The optimal deployment objective function is as follows

[0052]

[0053] Where s j represents the jth insecticidal lamp in the insecticidal lamp set S, C s represents the cost of an insecticide lamp, x j Indicates whether insecticidal lamps are deployed on farmland grid node j. k represents the kth insecticidal lamp in the insecticidal lamp set C, C c represents the cost of a camera, z k Indicates whether an insecticidal lamp is deployed on insecticidal lamp node k.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows: when deploying insecticidal lamps, an algorithm of deploying from the outside to the inside starting from the convex corner is used to solve the regional coverage problem, so as to achieve full coverage of the feasible area of ​​the farmland with the least number of insecticidal lamps. Secondly, in the deployment of cameras to monitor insecticidal lamps, an outer polygon is constructed, and the camera is deployed toward the inside of the polygon. This achieves an optimal deployment of insecticidal lamps and cameras with the lowest total cost while detecting intruders and preventing the theft of insecticidal lamps. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 Schematic diagram of the method of deploying insecticidal lamps and cameras in irregular farmland

[0056] Figure 2 Schematic diagram of regional coverage solution

[0057] Figure 3 Schematic diagram of the target coverage solution.

[0058] Figure 4 Specific example diagram. DETAILED DESCRIPTION

[0059] The present invention is further described in detail below in conjunction with the accompanying drawings.

[0060] like Figure 1 The figure shows a flow chart of a method for deploying insecticidal lamps and cameras in irregular farmland provided by the present invention. As can be seen from the figure, the method includes the following steps:

[0061] S1: Establish an omnidirectional sensing model of insecticidal lamps based on their coverage characteristics. This node can kill insects that appear at any point within its sensing range with a probability of 1.

[0062] S2: Build a camera direction perception model based on the camera's coverage characteristics. The camera installed on the insecticidal lamp node observes itself and other insecticidal lamp nodes to prevent the insecticidal lamp from being stolen and damaged.

[0063] S3: According to the irregularity of the farmland area, the farmland is gridded, the farmland area is divided into feasible area and obstacle area, and a mathematical model is established.

[0064] S4: Establish a regional coverage model for insecticidal lamps covering farmland and a target coverage model for cameras covering insecticidal lamps.

[0065] S50: The main algorithm of the two-layer optimization algorithm is to solve the regional coverage problem and deploy insecticidal lamp nodes to eliminate pests. When the insecticidal lamp covers the feasible area, the feasible area is first processed to find the outermost convex corner point of the feasible area, and the deployment starts from the first convex corner point on the outermost side until all the outermost convex corners are covered. Then the remaining feasible areas are subjected to the same operation until the entire feasible area is fully covered by the insecticidal lamp.

[0066] S51: The double-layer optimization algorithm sub-algorithm solves the target coverage problem and deploys cameras to monitor all insecticidal lamp nodes. The optimal deployment of cameras requires not only decisions on the placement of the cameras, but also decisions on the perception direction of the cameras. In terms of deployment, the surrounding insecticidal lamp nodes are connected together to form a polygon, starting from the first insecticidal lamp node, and cameras are deployed in sequence to cover the optimal deployment of all insecticidal lamp nodes.

[0067] In step S1, according to the phototaxis of a certain type of pests, an insecticidal lamp with a suitable capture radius is found, an omnidirectional perception model is established, and a 2-tuple of the fixed capture distance of the insecticidal lamp and the deployment position of the insecticidal lamp is used to represent it.

[0068] In step S2, a direction perception model is established for the camera, using a 4-tuple of camera deployment position, perception direction, perception radius and perception angle to represent it.

[0069] In step S3, the farmland is divided into feasible areas and obstacle areas according to the irregularity of the farmland area, and the farmland is gridded to find the set of cell centers. The grid division standard is determined by the user. If high accuracy is required, K is set to a higher value, which will be at the expense of higher computational complexity. If K is set to a smaller value, lower time complexity can be obtained.

[0070] In step S4, two coverage models, area coverage and target coverage, are established. Figure 2 All farmlands shown are feasible areas and are evenly divided into 16 small squares. The 16 small squares are feasible areas, and the coverage goal is to use the least number of insecticidal lamps to cover 16 grid centers. In this example, a total of 4 insecticidal lamp nodes {s1, s2, s3, s4} are deployed and placed at the center of 4 small square grids to meet the area coverage of the feasible area. Two camera nodes {c1, c2} are deployed, respectively deployed on insecticidal lamp nodes s2 and s3, to meet the monitoring deployment of 4 insecticidal lamp nodes.

[0071] In step S50, the main algorithm of the double-layer optimization algorithm is used to solve the regional coverage problem. The algorithm consists of an inner and outer double loop. The outer loop inputs the uncovered grid set points, and obtains the outermost convex set in the feasible area through the convex point function. The inner loop starts to deploy insecticidal lamps from the first point of the convex set obtained in the outer loop, and deploys the insecticidal lamps on the grid nodes with the maximum coverage rate that can cover the first point of the convex set. According to this deployment method, the outermost grid convex corner set is completely covered. In this way, the insecticidal lamps are deployed from the outside to the inside along the convex corner deployment method until the entire farmland is covered. Specific examples are as follows Figure 3 As shown in the figure, it consists of a feasible area and an obstacle area, and the feasible area is divided into 63 small squares. The main algorithm performs the first outer loop to obtain the outermost convex point set {v1, v2, v3, v4, v5, v6, v7, v8}. This is followed by the inner loop of the main algorithm, which covers the convex point set obtained from the outer loop. For the grid center p1 closest to the first convex set point v1, in order to achieve the maximum coverage while covering p1, it is best to deploy the insecticidal lamp at the grid p2 position. This deployment method is used until the outermost convex set is completely covered. The first coverage result is as follows: Figure 3 As shown by the black circle in the figure. After the first cycle ends, the remaining feasible region set is updated and the above cycle is repeated until the entire feasible region is covered.

[0072] In step S51, a sub-algorithm of the double-layer optimization algorithm is used to solve the target coverage problem. The algorithm consists of a loop, which inputs the insecticidal lamp node to obtain the convex corner set of the outer layer of the insecticidal lamp set. The camera is deployed at the first point of the convex corner set, and the starting edge of the camera is pointed by the deployed camera to the nearest insecticidal lamp. This deployment ensures that the camera can at least cover itself and the nearest insecticidal lamp. The direction of the camera is toward the inside of the multi-convex polygon, which can avoid the camera facing the outside of the farmland, causing a waste of resources. Specific examples are as follows Figure 4 As shown in the figure, the set of insecticidal lamps in the figure is {s1, s2, s3, s4, s5}, which can form a pentagon. Camera c1 is deployed on the first insecticidal lamp node s1, with the starting direction along the insecticidal lamp node s2 closest to s1, and the direction is toward the inside of the polygon. It can be seen that the set of insecticidal lamps covered by c1 is {s1, s2, s5}. Then camera c2 is deployed on the insecticidal lamp node s3, with the starting direction along s4. It can be seen that the set of insecticidal lamps covered by c2 is {s3, s4}. The insecticidal lamps are all covered, and the camera deployment is complete.

[0073] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0074] 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. A method and system for optimizing the deployment of an insecticidal lamp Internet of Things combined with video surveillance, characterized in that: It includes establishing an omnidirectional coverage model and a directional coverage model, and then converting the problem into two problems: regional coverage and target coverage. By designing a two-layer optimization algorithm, a new type of agricultural Internet of Things node with the lowest deployment cost is realized. The system can kill pests and detect intruders at the same time.

2. According to the method and system for optimizing the deployment of the Internet of Things of insecticidal lamps combined with video surveillance in claim 1, it is characterized in that: The following steps are involved: S1: Establishing an omnidirectional perception model of the insecticidal lamp according to the coverage characteristics of the insecticidal lamp; S2: Establishing a direction perception model of the camera according to the coverage characteristics of the camera; S3: According to the irregularity of the farmland area, the farmland is gridded and divided into feasible areas and obstacle areas. Insecticide lamps and cameras are prohibited from being deployed in obstacle areas. S4 designs a coverage model based on the omnidirectional perception model of insecticidal lamps established in S1 and the irregularity of farmland areas in S3. Insecticidal lamps can only be deployed in the grid center of feasible farmland areas, and cameras can only be deployed on insecticidal lamps; S5: Design a two-layer optimization algorithm to solve the regional coverage problem and the target coverage problem in turn. When solving the deployment position of the insecticidal lamp in the farmland, deploy it from the convex corner from the outside to the inside, and then use the deployment position of the insecticidal lamp as input to calculate the camera position.

3. According to claim 2, a method and system for optimizing the deployment of an insecticidal lamp Internet of Things combined with video surveillance, characterized in that: The coverage characteristics of the S1 insecticidal lamp establish the omnidirectional perception model of the insecticidal lamp as a disc perception model. The disc perception model is: In the formula, represents the Euler distance between the insecticidal lamp and the center of the grid, r s Indicates the effective radius of the insecticidal lamp. The coverage characteristics of the S2 camera establish the direction perception model of the camera as a fan-shaped perception model. The fan-shaped perception model is: In the formula, and is the camera node c k Nodes with insecticidal lamps j The distance between the positive horizontal axis and o kl The angle in the counterclockwise direction. The S3 irregular farmland consists of a feasible area and an obstacle area. The S5 two-level optimization algorithm consists of solving two mixed integer programming models in sequence.

4. According to claim 2, a method and system for optimizing the deployment of an insecticidal lamp Internet of Things combined with video surveillance, characterized in that: In S3, the farmland is gridded and divided into feasible areas and obstacle areas. In the obstacle areas, insecticidal lamps and cameras are prohibited from being deployed. According to the irregularity of the farmland area, it can be manually set to meet the requirements of different farmlands in practical applications. The farmland is modeled as a cross-connected grid point area. The center point of the cell in the feasible area is defined as a set P, and the insecticidal lamp needs to cover all points in the set P. Note that the grid division standard is determined by the user. If K is set to a larger value, the coverage accuracy will be improved, but the calculation complexity will be increased. On the contrary, if K is set to a smaller value, lower time complexity will be achieved, and the coverage accuracy will be reduced.

5. According to claim 2, a method and system for optimizing the deployment of an insecticidal lamp Internet of Things combined with video surveillance, characterized in that: The omnidirectional perception model of the insecticidal lamp established in S4 and the irregularity of the farmland area in S3. S40: The model of insecticidal lamps covering farmland is a regional coverage model. The purpose of coverage is to attract insects with insecticidal lamps. Based on the S1 omnidirectional perception model, any point p i The coverage rate, denoted by p i The pests are killed by the insect lamp set S j The probability of node capture can be expressed as: Where λ i Indicates P i The probability of being captured by S, S j represents the jth insecticidal lamp node in the insecticidal lamp set S, λ ij represents the probability that the i-th node is covered by the j-th insecticidal lamp. Therefore, the area coverage can be calculated as Where λ represents the coverage rate of the farmland by insecticidal lamps, p i represents the i-th grid node in the feasible region P, and K represents the total number of grids in the farmland. S41: The model of cameras covering insecticidal lamps is a target coverage model, which aims to monitor the theft and destruction of insecticidal lamp nodes. Therefore, all nodes where insecticidal lamps are deployed should be covered by installed cameras. Based on the S2 direction perception model, any insecticidal lamp node s can be described as j The coverage of insecticidal lamp nodes s j The probability that the position is covered by the camera set can be expressed as: Where ρ j Indicates j The probability of being covered by the camera in C, c k represents the kth camera in the camera set C, where C represents the camera set, c l It is consistent with the lth direction in the camera direction set O. jkl Indicates that the jth insecticidal lamp is c along the l direction k Probability of camera coverage Therefore, the target coverage for ROI can be calculated as Where ρ represents the probability of the insecticidal lamp set being covered, s j represents the jth insecticidal lamp in the insecticidal lamp set S, ρ j represents the probability that the jth insecticidal lamp is covered, and M represents the total number of insecticidal lamps.

6. According to claim 2, a method and system for optimizing the deployment of an insecticidal lamp Internet of Things combined with video surveillance, characterized in that: The S5 two-layer optimization algorithm solves the covering problem by solving two mixed integer programming models in sequence.

7. According to claim 6, the method and system for optimizing the deployment of the Internet of Things of insecticidal lamps combined with video surveillance is characterized in that: For the main problem of the two-layer optimization algorithm, in the first stage of the algorithm, the goal of solving the main problem is to find the optimal number of insecticidal lamp nodes required to cover all feasible areas in the farmland, that is, to optimize the deployment position of the insecticidal lamp nodes under the constraint of area coverage λ=1. The objective function is designed as follows: Where s j represents the jth insecticidal lamp in the insecticidal lamp set S, C s represents the cost of an insecticide lamp, x j Indicates whether insecticidal lamps are deployed on farmland grid node j. Constrained by Where s i represents the i-th insecticidal lamp in the insecticidal lamp set S, w ij represents the probability that farmland grid point i is covered by insecticidal lamp node j, x j Indicates whether insecticidal lamps are deployed on farmland grid node j. i represents the i-th feasible area in the farmland P.

8. According to the method and system for optimizing the deployment of insecticidal lamps in the Internet of Things combined with video surveillance described in claim 7, due to the influence of complex terrain, when solving the main problem, farmland is usually an irregular polygon. The feasible area in the farmland is divided into grids to obtain a grid point set, and the outermost convex point set in the feasible area is obtained by a convex point function. Starting from the first convex vertex in the convex point set, the insecticidal lamp is deployed on the grid node with the maximum coverage rate that can cover the first point of the convex set, and the outermost grid convex corner set is completely covered in this deployment mode. After the first group of convex point sets are covered, the above deployment strategy is repeated on the remaining insecticidal lamp nodes to cover the remaining point groups, and the deployed insecticidal lamp nodes in the insecticidal lamp set S and the grid point sets that are not covered in the remaining feasible area are updated until the feasible area is fully covered. The model design to solve the main problem is as follows: In the formula It represents the best deployment position of insecticidal lamp, p j represents the jth point in the grid p.

9. According to claim 6, the method and system for optimizing the deployment of the Internet of Things of insecticidal lamps combined with video surveillance is characterized in that: For the subproblems of the two-layer optimization algorithm, in the second stage of the two-layer optimization algorithm, the goal of solving the subproblem is to find the optimal (i.e., minimum) number of camera nodes required to cover all insect-killing lamp nodes. In other words, the goal of the subproblem is to optimize the deployment position of the camera nodes under the constraint of the target coverage rate ρ=1. The objective function is designed as follows: Where C k represents the kth insecticidal lamp in the insecticidal lamp set C, C c represents the cost of a camera, z k Indicates whether an insecticidal lamp is deployed on insecticidal lamp node k.

10. According to the method and system for optimizing the deployment of insecticidal lamps in the Internet of Things combined with video surveillance described in claim 9, when solving the sub-problem, the fan-shaped sensing range of the camera is a directional sensing model that can only observe a certain angle. Since the deployment positions of insecticidal lamp nodes are discrete, the number of observed insecticidal lamp nodes may vary with the placement angle of the camera. The optimal deployment of the camera requires not only a decision on the placement position of the camera, but also a decision on the perception direction. An optimal deployment strategy is proposed to cover all insecticidal lamp nodes and minimize the number of cameras. By connecting the surrounding insecticidal lamp nodes together to form a polygon, starting from the first insecticidal lamp node, and deploying cameras in sequence to cover all insecticidal lamp nodes. The deployed system can kill pests while detecting intruders with the lowest total cost. The optimal deployment objective function is as follows: Where s j represents the jth insecticidal lamp in the insecticidal lamp set S, C s represents the cost of an insecticide lamp, x j Indicates whether insecticidal lamps are deployed on farmland grid node j. k represents the kth insecticidal lamp in the insecticidal lamp set C, C c represents the cost of a camera, z k Indicates whether an insecticidal lamp is deployed on insecticidal lamp node k.

Images