Planting path planning method and system based on internet of things

Abstract

The application relates to the technical field of image processing, and particularly discloses a planting path planning method and system based on an Internet of Things, which comprises the following steps: acquiring normalized slope and NDVI values at each point in a farmland; clustering each point according to a comprehensive distance between the points to obtain a group of optimal clustering clusters; dividing the farmland into multiple sub-regions by using a Venn diagram algorithm; and planning paths for all the sub-regions by using a ploughing coverage algorithm to obtain optimal planting paths. The planting path planning method and system based on the Internet of Things can avoid repeated work at the edge of the farmland, improve work efficiency, and meet the work requirements of regions with different slopes and different crop densities.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more specifically to a planting path planning method and system based on the Internet of Things. Background Technology

[0002] The Internet of Things (IoT) refers to connecting objects to a network through information sensing devices and according to agreed-upon protocols. These objects exchange and communicate information through communication media to achieve functions such as intelligent identification, positioning, tracking, and monitoring. IoT technology can be used to remotely control agricultural machinery and automate planting operations, such as automated seedling raising, spraying, and fertilization, thereby improving agricultural production efficiency and reducing labor intensity.

[0003] To achieve full coverage of farmland, reduce repetitive operations, and lower fuel consumption and wear on agricultural machinery, it is necessary to rationally plan the movement paths of agricultural machinery. Among these methods, the ox-plowing coverage algorithm is a widely used full-coverage path planning algorithm in agriculture. This algorithm ensures that every area is covered by simulating the movement of oxen plowing through the farmland. Specifically, the farmland to be planned is divided into a series of simple cells, and a movement pattern is found to ensure that each cell is visited at least once.

[0004] In existing ox-plowing covering algorithms, cells are typically regular in shape, such as rectangles, and the row spacing of the movement path is fixed. However, actual farmland is usually irregular in shape and has slopes. Regularly shaped cells cannot achieve full coverage of the farmland edges, resulting in overlapping areas in the turning paths between adjacent cells at the farmland edges. This not only lengthens the movement path and reduces work efficiency but also leads to repetitive work. Furthermore, using fixed row spacing cannot meet the needs of areas with varying slopes and crop density, resulting in row spacing that is too large or too small, leading to missed or repeated work. Summary of the Invention

[0005] This invention provides a planting path planning method and system based on the Internet of Things, aiming to solve the technical problems in the prior art, such as path repetition due to irregular edges, low operation efficiency, and the inability of fixed path row spacing to meet the needs of different farmland types.

[0006] This invention provides a planting path planning method based on the Internet of Things, comprising the following steps:

[0007] Obtain the normalized slope and NDVI value at various points within the farmland;

[0008] Clustering is performed on the points based on the comprehensive distance between them, resulting in an optimal cluster.

[0009] Wherein, the comprehensive distance is the sum of the products of the Euclidean distance and the feature distance between two points and their corresponding weights; the feature distance is positively correlated with the square of the difference between the slopes between the corresponding two points and the square of the difference between the NDVI values; the weight of the Euclidean distance is inversely correlated with the standard deviation of the slope of each point in the farmland, and the weight of the feature distance is the difference between 1 and the weight of the corresponding Euclidean distance.

[0010] The farmland is divided into multiple sub-regions using the Vinonic diagram algorithm. The seed point of the Vinonic diagram algorithm is the cluster center of each cluster, and the combined distance from a point in each sub-region to the seed point in the corresponding sub-region is less than the combined distance to the other seed points.

[0011] The ox-plowing mulching algorithm is used to plan paths for all sub-regions to obtain the optimal planting path;

[0012] Among them, the path row spacing in each sub-region is positively correlated with the standard path row spacing and negatively correlated with the corresponding row spacing adjustment coefficient; the row spacing adjustment coefficient is positively correlated with the mean slope of each point in the corresponding sub-region and negatively correlated with the mean NDVI value of each point in the corresponding sub-region.

[0013] In the above scheme, by dividing the farmland into multiple sub-regions in a more refined manner and using the ox-plowing mulching algorithm for path planning in each sub-region, full coverage of each sub-region can be achieved, thus avoiding the repetition of paths when turning at the edges of the farmland. Furthermore, by correcting the standard path row spacing to obtain the path row spacing for each sub-region, it can meet the path planning needs of areas with different slopes and crop density.

[0014] Preferably, the characteristic distance d between point a and point b within the farmland a,b for:

[0015]

[0016] In the formula, P a Let P be the slope at point a. b Let I be the slope at point b. a Let I be the NDVI value of point a. b Let be the NDVI value of point b.

[0017] In the above scheme, the characteristic distance between two points is characterized by the difference in slope and the difference in NDVI value, making the characteristic distance more accurate.

[0018] Preferably, the weight of the Euclidean distance is the normalized value of the inverse of the standard deviation of the slope at each point in the farmland.

[0019] Preferably, the optimal cluster is the group of clusters with the highest clustering score; wherein, the clustering score Q of the i-th cluster is... i for:

[0020]

[0021] In the formula, B i,j W is the mean of the combined distances between each point in the j-th cluster of the i-th cluster and the cluster centers of the other clusters. i,j Let n be the mean of the combined distances between each point in the j-th cluster within the i-th cluster and its cluster center. i is the total number of clusters in the i-th cluster group.

[0022] In the above scheme, the clustering score is measured by inter-cluster distance and intra-cluster distance. The larger the inter-cluster distance and the smaller the intra-cluster distance, the higher the clustering score and the better the corresponding cluster. This makes the calculation simple and easy to understand.

[0023] Preferably, the row spacing adjustment coefficient α of the k-th sub-region k for:

[0024]

[0025] In the formula, Let be the average slope of all points within the k-th sub-region. P is the mean NDVI value of all points in the k-th sub-region. max I represents the maximum slope at all points within the farmland. max This represents the maximum NDVI value at each point within the farmland.

[0026] In the above scheme, the average slope and the average NDVI value of each point in each sub-region are used to characterize the row spacing adjustment coefficient of the corresponding sub-region. The influence of slope and crop density on the path row spacing is fully considered, making the calculation of path row spacing in each sub-region more accurate and reasonable.

[0027] Preferably, the difference between the path row spacing of 1 and the corresponding row spacing adjustment coefficient in each sub-region is multiplied by the standard path row spacing.

[0028] Preferably, the optimal planting path is the planting path corresponding to the minimum value of the objective function of each planting path; wherein, the objective function R of the m-th planting path is... m for:

[0029]

[0030] In the formula, T m Let T be the total operation time for the m-th planting path. max S represents the maximum total operation time. m Let S be the area of ​​the path repetition of the m-th planting path, and S be the total area of ​​farmland.

[0031] In the above scheme, the quality of the planting path is measured by the total operation time and the path repetition area. The smaller the total operation time and the smaller the path repetition area, the better the planting path. This method is simple to calculate and easy to understand.

[0032] Preferably, the slope of each point in the farmland is calculated using a digital elevation model of the farmland; the NDVI value of each point in the farmland is calculated using multispectral images of the farmland.

[0033] Among the above schemes, the digital elevation model (DEM) has advantages such as high accuracy, good visualization, and ease of slope calculation. Multispectral imagery can provide information from different spectral bands, thus facilitating the calculation of NDVI values.

[0034] Preferably, using a mixture of A * The algorithm smooths out the turns of the optimal planting path.

[0035] In the above scheme, smoothing the turns of the planting path can simplify calculations, improve the efficiency of the algorithm, and reduce the energy consumption of vehicles when turning.

[0036] The present invention also provides an Internet of Things (IoT)-based planting path planning system, including a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the IoT-based planting path planning method described above.

[0037] The beneficial effects are:

[0038] The present invention uses the Veno diagram algorithm to finely divide farmland into multiple sub-regions, and then applies the ox-plowing mulching algorithm to each sub-region, ensuring full coverage at the edges of the farmland and preventing overlapping areas in turning paths, thereby improving operational efficiency. Furthermore, the standard path row spacing is adjusted according to the different slopes and NDVI values ​​in different sub-regions, ensuring that the adjusted row spacing meets the planting needs of different sub-regions. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the steps of the IoT-based planting path planning method according to an embodiment of the present invention.

[0040] Figure 2 This is a structural block diagram of an IoT-based planting path planning system according to an embodiment of the present invention. Detailed Implementation

[0041] The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the invention, and should not be construed as limiting the invention.

[0042] like Figure 1As shown, according to a first aspect of the present invention, a planting path planning method based on the Internet of Things is provided, comprising the following steps:

[0043] S1. Obtain the normalized slope and NDVI value at each point in the farmland.

[0044] The slope of each point within the farmland is calculated using a digital elevation model (DEM). A DEM is a digital simulation of the ground topography using limited topographic elevation data; it is a physical ground model representing ground elevation as an ordered array of numerical values. DEMs are typically represented by an array of regularly spaced grid points on the ground, each with a corresponding elevation value. These elevation values ​​are usually obtained through methods such as topographic mapping, aerial photogrammetry, satellite remote sensing, or ground surveying, using acquisition equipment including lidar and high-resolution cameras. These acquisition devices facilitate data collection, which in turn facilitates data processing in subsequent steps. For example, the slope of each point within the farmland can be calculated using a fitted plane method based on the DEM. Alternatively, the DEM can be input into geographic information system (GIS) software to obtain the slope at each point within the farmland.

[0045] NDVI (Normalized Difference Vegetation Index) is used to monitor and assess vegetation cover and growth status. NDVI is calculated by analyzing the red and near-infrared light bands in multispectral imagery acquired by satellite or aircraft. NDVI is the ratio of the difference between the near-infrared and red light bands to the sum of the near-infrared and red light bands. A higher NDVI value indicates denser vegetation, while a lower NDVI value indicates sparser or absent vegetation.

[0046] In this step, the purpose of normalizing the slope and NDVI value is to improve the comparability of different data types. Normalization process unifies data with different units and dimensions into the same numerical range, so that different indicators can be compared and analyzed comprehensively, eliminating errors caused by different data dimensions.

[0047] The location of each point within the farmland can be represented using coordinates. For example, by establishing a rectangular coordinate system with the location of the agricultural machinery as the origin, the edge of the farmland where the machinery is located as the horizontal axis, and the vertical axis perpendicular to the edge of the farmland where the machinery is located as the vertical axis, the coordinates of each point within the farmland, as well as the corresponding slope and NDVI value, can be obtained.

[0048] S2. Cluster the points according to the comprehensive distance between them to obtain an optimal cluster.

[0049] The comprehensive distance is the sum of the products of the Euclidean distance and the feature distance between two points, respectively, multiplied by their corresponding weights. The feature distance characterizes the difference in slope and NDVI value between the two points. This is because actual farmland is not flat and has slopes, and the density of crops varies at different points within the farmland. Density can be characterized by the NDVI value; a larger NDVI value indicates denser crops, and a smaller NDVI value indicates sparser crops. Therefore, when clustering points within the farmland, it is necessary to consider not only the actual straight-line distance between the two points, i.e., the Euclidean distance, but also the feature distance between them. Therefore, step S2 also includes the following steps:

[0050] S21. Obtain the combined distance between two points within the farmland.

[0051] To obtain the comprehensive distance between two points, it is necessary to obtain the feature distance, Euclidean distance, and corresponding weights between the two points. Therefore, step S21 also includes the following steps:

[0052] S211. Calculate the characteristic distance between two points.

[0053] The characteristic distance is positively correlated with the square of the difference in slope between the two points and the square of the difference in NDVI values. This is because the larger the square of the difference in slope and the square of the difference in NDVI values ​​between the two points, the greater the difference between the two points, i.e., the larger the characteristic distance. Conversely, the smaller the square of the difference in slope and the square of the difference in NDVI values ​​between the two points, the smaller the difference between the two points, i.e., the smaller the characteristic distance.

[0054] In one embodiment, the characteristic distance d between point a and point b within the farmland is... a,b for:

[0055]

[0056] In the formula, P a Let P be the slope at point a. b Let I be the slope at point b. a Let I be the NDVI value of point a. b Let be the NDVI value of point b.

[0057] In step S211, the characteristic distance between two points is characterized by the difference in slope and the difference in NDVI value, making the calculation of the characteristic distance more accurate.

[0058] S212. Calculate the Euclidean distance between two points.

[0059] Euclidean distance represents the actual straight-line distance between two points. It is the square root of the sum of the squares of the differences between the corresponding coordinates of the two points.

[0060] S213. Obtain the weights of the feature distance and Euclidean distance between the two points respectively.

[0061] The weight of the Euclidean distance is inversely correlated with the standard deviation of the slope at each point within the farmland, while the weight of the characteristic distance is the difference between 1 and the corresponding weight of the Euclidean distance. This is because a larger standard deviation of the slope at each point within the farmland indicates a greater dispersion of slopes at each point, meaning the farmland is more uneven. Using Euclidean distance for clustering will produce a larger error, therefore, Euclidean distance should be assigned a smaller weight. Conversely, a larger standard deviation of the slope at each point within the farmland necessitates using the characteristic distance to characterize the distance between two points, thus requiring a larger weight for the characteristic distance.

[0062] In one embodiment, the weight of the Euclidean distance is the reciprocal normalized value of the standard deviation of the slope at each point in the farmland. The weight of the feature distance is the difference between 1 and the corresponding weight of the Euclidean distance.

[0063] In step S213, the standard deviation of each slope reflects the dispersion of each slope, and thus reflects the flatness of the farmland. This assigns different weights to Euclidean distance and characteristic distance, making the weight setting more reasonable and the calculation of the comprehensive distance between two points more accurate.

[0064] S214. Obtain the combined distance between the two points.

[0065] In one embodiment, the combined distance D between point a and point b within the farmland is... a,b for:

[0066]

[0067] In the formula, d a,b Let d be the characteristic distance between point a and point b within the farmland. ′ a,b Let σ be the Euclidean distance between points a and b within the farmland. P Let be the standard deviation of the slope at all points within the farmland, and norm() be the standard normalization function.

[0068] In step S21, different weights are assigned to Euclidean distance and feature distance based on the flatness of the farmland, thereby obtaining the corresponding comprehensive distance. This comprehensive distance not only reflects the true straight-line distance between two points, i.e., Euclidean distance, but also reflects the feature distance between two points, making it more reasonable and accurate to use the comprehensive distance for clustering in subsequent steps.

[0069] S22. Calculate the cluster score of a group of clusters obtained in each clustering, and take the group of clusters with the highest cluster score as the optimal cluster.

[0070] During clustering, clustering parameters need to be set, such as the number of clusters. Setting different numbers of clusters will result in multiple different cluster groups. In this step, different numbers of clusters are set to obtain multiple different cluster groups, and then the clustering scores of each group are compared. The cluster group with the highest clustering score is selected as the optimal cluster group.

[0071] Clustering scores are typically measured using inter-cluster distance and intra-cluster distance. A higher inter-cluster distance and a lower intra-cluster distance result in a higher clustering score and a better cluster. Specifically, inter-cluster distance represents the average distance between each point within a given cluster and the cluster centers of the other clusters, while intra-cluster distance represents the average distance between each point within a given cluster and its cluster center.

[0072] In one embodiment, the clustering score Q of the i-th cluster group is... i for:

[0073]

[0074] In the formula, B i,j W is the mean of the combined distances between each point in the j-th cluster of the i-th cluster and the cluster centers of the other clusters. i,j Let n be the mean of the combined distances between each point in the j-th cluster within the i-th cluster and its cluster center. i is the total number of clusters in the i-th cluster group.

[0075] After obtaining the cluster score of each cluster using the cluster scoring formula, the cluster with the highest cluster score is taken as the optimal cluster.

[0076] In step S2, the points in the farmland are clustered based on the comprehensive distance between the two points. This can take into account the influence of Euclidean distance and characteristic distance, and cluster points with similar slopes and similar NDVI values ​​into the same cluster. As a result, the division of the clusters is more reasonable, which is beneficial to the division of farmland in subsequent steps.

[0077] S3. Use the Veno diagram algorithm to divide the farmland into multiple sub-regions.

[0078] In step S3, the seed points of the Venn diagram algorithm are the cluster centers of each cluster, and the combined distance from a point in each sub-region to the seed point in the corresponding sub-region is less than the combined distance to the other seed points.

[0079] The Veno diagram algorithm, an existing technique, is a method for dividing a plane into multiple sub-regions. Each sub-region is generated from a seed point, and points within the sub-region are closer to their corresponding seed point via Euclidean distance than to other seed points. In existing techniques, the distances between points within a sub-region and the seed point are measured using Euclidean distance, which represents the actual straight-line distance between two points. This does not consider the slope of the farmland or the density of crops. Therefore, directly dividing farmland into sub-regions using the existing Veno diagram algorithm is unreasonable. Furthermore, the existing Veno diagram algorithm can only group points with similar actual straight-line distances into the same sub-region, but it cannot group points with similar slopes or similar NDVI values ​​into the same sub-region.

[0080] In step S3, the cluster center of each cluster is used as the seed point, and the corresponding sub-region is generated by the combined distance between the seed point and the other points, so that the division of farmland is more reasonable, taking into account not only the Euclidean distance between two points, but also the slope and NDVI value between two points.

[0081] S4. Use the ox-plowing mulching algorithm to plan paths for all sub-regions and obtain the optimal planting path.

[0082] The ox-plowing mulching algorithm is an existing technology that simulates the way oxen plow fields. This algorithm divides the farmland into multiple cells, ensuring that each cell is covered. In existing ox-plowing mulching algorithms, the cells are typically rectangular, and the row spacing of the movement paths is fixed. However, actual farmland is usually irregular in shape and has slopes. Rectangular cells are difficult to fully cover the irregular edges of the farmland, resulting in overlapping areas in the turning paths between adjacent cells at the farmland edges. This not only lengthens the movement path and reduces work efficiency but also leads to repetitive work. Furthermore, using fixed row spacing cannot meet the needs of areas with varying slopes and crop density, resulting in row spacing that is too large or too small, leading to missed or repeated work.

[0083] In step S4, the irregular farmland is divided into multiple sub-regions. The ox-plowing mulching algorithm is used for path planning in each sub-region, so that each sub-region can achieve full coverage, thereby achieving full coverage of the entire farmland. This avoids overlapping areas at the turning paths at the edge of the farmland and improves work efficiency.

[0084] Because the slope and crop density vary across different sub-regions, a fixed row spacing cannot meet the planting needs of different sub-regions. Therefore, it is necessary to adjust the row spacing for each sub-region. Specifically, the row spacing within each sub-region is positively correlated with the standard row spacing and negatively correlated with the corresponding row spacing adjustment coefficient. This row spacing adjustment coefficient is positively correlated with the mean slope of each point within the corresponding sub-region and negatively correlated with the mean NDVI value of each point within the corresponding sub-region.

[0085] The standard path row spacing is the row spacing within a level field, and it can be obtained empirically. By adjusting the standard path row spacing using a row spacing adjustment factor, the path row spacing for each corresponding sub-region can be obtained. The greater the average slope of each point within a sub-region, the more likely it is to result in an excessively large path row spacing; therefore, a larger row spacing adjustment factor is needed to reduce the standard path row spacing. Conversely, the smaller the average NDVI value of each point within a corresponding sub-region, indicating sparser crops, the smaller the path row spacing is required to prevent the standard path row spacing from being too large, leading to insufficient fertilization and spraying of crops, resulting in malnutrition.

[0086] Therefore, obtaining the optimal planting path in step S4 includes the following steps:

[0087] S41. Obtain the path row spacing for each sub-region.

[0088] The path spacing of each sub-region is related to the corresponding line spacing adjustment coefficient, therefore step S41 also includes the following steps:

[0089] S411. Calculate the row spacing adjustment coefficient for each sub-region.

[0090] In one embodiment, the row spacing adjustment factor α of the k-th sub-region k for:

[0091]

[0092] In the formula, Let be the average slope of all points within the k-th sub-region. P is the mean NDVI value of all points in the k-th sub-region. max I represents the maximum slope at all points within the farmland. max This represents the maximum NDVI value at various points within the farmland.

[0093] In step S411, the average slope and the average NDVI value of each point in each sub-region are used to characterize the row spacing adjustment coefficient of the corresponding sub-region. The influence of slope and crop density on the path row spacing is fully considered, making the calculation of path row spacing in each sub-region more accurate and reasonable.

[0094] S412. Obtain the path row spacing of the corresponding sub-region based on the row spacing adjustment coefficient of each sub-region.

[0095] The product of the difference between the path row spacing of 1 and the corresponding row spacing adjustment coefficient in each sub-region and the standard path row spacing.

[0096] Specifically, the path spacing H of the k-th sub-region k for:

[0097] H k =H0×(1-α) k );

[0098] In the formula, H0 is the standard path line spacing, and α k This is the row spacing adjustment factor for the k-th sub-region.

[0099] In step S41, the row spacing of the standard path is adjusted by the row spacing adjustment coefficient to obtain the row spacing of the corresponding sub-region. This can meet the planting conditions of sub-regions with different slopes and crop density, and avoid omissions in operations or crop malnutrition.

[0100] S42. Based on the path row spacing of each sub-region, the ox-plowing mulch algorithm is used to plan the path for all sub-regions and obtain the optimal planting path.

[0101] Because the ox-plowing mulching algorithm is used for path planning in each sub-region, and the shape and size of each sub-region are different, different planting paths can be obtained. For example, 50 path planning iterations can be performed, and the planting path corresponding to the minimum value of the objective function of the planting path is the optimal planting path.

[0102] The quality of a planting path can be measured by the total operation time and the area of ​​path repetition. The shorter the total operation time and the smaller the area of ​​path repetition, the better the planting path.

[0103] In one embodiment, the objective function R of the m-th planting path is... m for:

[0104]

[0105] In the formula, T m Let T be the total operation time for the m-th planting path. max S represents the maximum total operation time. m Let S be the area of ​​the path repetition of the m-th planting path, and S be the total area of ​​farmland.

[0106] The objective function value of each planting path is calculated, and the planting path corresponding to the minimum value of the objective function is taken as the optimal planting path.

[0107] In step S42, to improve the turning performance of the planting path, a mixture of A is used.* The algorithm smooths out the turns of the optimal planting path.

[0108] After obtaining the optimal planting path according to the method of the present invention, the optimal planting path is input into the control terminal of the agricultural machinery to carry out automated operation of the agricultural machinery.

[0109] This invention employs global planning, which involves planning an optimal path from the starting point to the destination for agricultural machinery based on farmland information, given a known farmland area. This is because farmland areas have few or no obstacles, and global planning can ensure coverage, improve planning efficiency, and reduce the use of computing resources.

[0110] In the IoT-based planting path planning method of this invention, the points within the farmland are first clustered based on comprehensive distance to obtain an optimal cluster. Then, using the cluster centers as seed points, the farmland is divided into multiple sub-regions using the Veno diagram algorithm. Finally, the ox-plowing mulching algorithm is used to plan paths for all sub-regions to obtain the optimal planting path. Because the farmland is divided into multiple sub-regions, the edges of the farmland can be fully covered when using the ox-plowing mulching algorithm for path planning, preventing overlapping areas in turning paths and thus improving operational efficiency. Furthermore, the modified row spacing used within each sub-region can meet the planting needs of different sub-regions.

[0111] like Figure 2 As shown, according to a second aspect of the present invention, an Internet of Things (IoT)-based planting path planning system is also provided. The system includes a memory and a processor, wherein the processor executes a computer program stored in the memory to implement the IoT-based planting path planning method described in the first aspect of the present invention.

[0112] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0113] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented by computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0114] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A planting path planning method based on the Internet of Things, characterized in that, Includes the following steps: Obtain the normalized slope and NDVI value at various points within the farmland; Clustering is performed on the points based on the comprehensive distance between them, resulting in an optimal cluster. Wherein, the comprehensive distance is the sum of the products of the Euclidean distance and the feature distance between two points and their corresponding weights; the feature distance is positively correlated with the square of the difference between the slopes between the corresponding two points and the square of the difference between the NDVI values; the weight of the Euclidean distance is inversely correlated with the standard deviation of the slope of each point in the farmland, and the weight of the feature distance is the difference between 1 and the weight of the corresponding Euclidean distance. The farmland is divided into multiple sub-regions using the Vinonic diagram algorithm. The seed point of the Vinonic diagram algorithm is the cluster center of each cluster, and the combined distance from a point in each sub-region to the seed point in the corresponding sub-region is less than the combined distance to the other seed points. The ox-plowing mulching algorithm is used to plan paths for all sub-regions to obtain the optimal planting path; Among them, the path row spacing in each sub-region is positively correlated with the standard path row spacing and negatively correlated with the corresponding row spacing adjustment coefficient; the row spacing adjustment coefficient is positively correlated with the mean slope of each point in the corresponding sub-region and negatively correlated with the mean NDVI value of each point in the corresponding sub-region.

2. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The characteristic distance d between points a and b within the farmland a,b for: In the formula, P a Let P be the slope at point a. b Let I be the slope at point b. a Let I be the NDVI value of point a. b Let be the NDVI value of point b.

3. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The weight of the Euclidean distance is the normalized value of the inverse of the standard deviation of the slope at each point in the farmland.

4. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The optimal cluster is the cluster with the highest clustering score; Wherein, the clustering score Q of the i-th cluster group i for: In the formula, B i,j W is the mean of the combined distances between each point in the j-th cluster of the i-th cluster and the cluster centers of the other clusters. i,j Let n be the mean of the combined distances between each point in the j-th cluster within the i-th cluster and its cluster center. i is the total number of clusters in the i-th cluster group.

5. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The row spacing adjustment factor α for the k-th sub-region k for: In the formula, Let be the average slope of all points within the k-th sub-region. P is the mean NDVI value of all points in the k-th sub-region. max I represents the maximum slope at all points within the farmland. max This represents the maximum NDVI value at each point within the farmland.

6. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The product of the difference between the path row spacing of 1 and the corresponding row spacing adjustment coefficient in each sub-region and the standard path row spacing.

7. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The optimal planting path is the planting path corresponding to the minimum value of the objective function of each planting path; Wherein, the objective function R of the m-th planting path is m for: In the formula, T m Let T be the total operation time for the m-th planting path. max S represents the maximum total operation time. m Let S be the area of ​​the path repetition of the m-th planting path, and S be the total area of ​​farmland.

8. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, The slope of each point in the farmland was calculated using the farmland's digital elevation model; the NDVI value of each point in the farmland was calculated using the farmland's multispectral imagery.

9. The planting path planning method based on the Internet of Things according to claim 1, characterized in that, Using Mixed A * The algorithm smooths out the turns of the optimal planting path.

10. A planting path planning system based on the Internet of Things, comprising a memory and a processor, characterized in that, The processor executes the computer program stored in the memory to implement the Internet of Things-based planting path planning method as described in any one of claims 1 to 9.

Images