Forest stand space structure optimization method considering neighborhood index and particle swarm optimization

By combining particle swarm algorithm and neighborhood index in the optimization of stand structure, the problem of single and nonlinear conflicts in the existing technology is solved, the precise optimization of the stand space structure is achieved, and the intelligent level of forest management is improved. stand quality.

CN118940904APending Publication Date: 2024-11-12RES INST OF FOREST RESOURCE INFORMATION TECHN CHINESE ACADEMY OF FORESTRY
View PDF 0 Cites 6 Cited by

Patent Information

Application Number
CN202411053862.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-01
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The existing forest stand structure optimization method ignores the potential impact of the retained wood properties and spatial distribution characteristics after harvesting on forest ecosystem functions in the logging strategy, and the optimization goal is single, with nonlinearity and discontinuity between multiple dimensions. Conflict makes it difficult to achieve accurate improvement in the quality of stands.

Method used

The stand space structure optimization method that considers neighborhood index and particle swarm algorithm is adopted, and the tree attribute data is recorded through field measurement, angle scale, size ratio, mixing degree and other indicators are calculated, and particle position is updated in combination with particle swarm algorithm initialization and iterative update. Build a comprehensive evaluation index for the spatial structure of the forest stands to achieve multi-objective optimization, accurately locate and harvest trees.

Benefits of technology

It has achieved accurate optimization of the spatial structure of the forest stand, improved the intelligence level of forest management, and can accurately capture the distribution of trees with small mixed degrees, large angle scales and large sizes in different business scenarios, reduce the degree of competition, and improve the overall stand Quality, better than traditional linear optimization models.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940904A_ABST
    Figure CN118940904A_ABST
Patent Text Reader

Abstract

The invention discloses a forest stand space structure optimization method considering a neighborhood index and a particle swarm algorithm, and belongs to the technical field of forest management and structure optimization in forestry scientific research. The tree height, diameter at breast height, crown breadth, height under branches, tree age, tree species, gradient, altitude and canopy density information of all investigated trees are recorded in a conventional field measurement mode, a forest stand space structure comprehensive evaluation index is constructed, and a forest stand structure covers forest distribution pattern, mixing degree and size differentiation and is used for forest stand structure evaluation and development succession prediction. By combining a particle swarm algorithm, initializing the position of a particle swarm, continuously iteratively updating the position of particles in the particle swarm and searching an optimal solution set, after each iteration, the comprehensive index of the forest stand space structure is calculated, whether the forest stand space structure reaches a set optimization target is judged, and whether the forest stand structure reaches an optimal optimization effect is finally judged by combining a cutting limit.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of forest management and structure optimization in forestry scientific research, and relates to a forest stand spatial structure optimization method considering neighborhood index and particle swarm algorithm. Background Art

[0002] Stand structure covers forest distribution pattern, degree of intermixing, and size differentiation, which can be used to evaluate stand structure and predict its development and succession. Considering the spatial structural characteristics of stands in the process of forest planning is considered the key to sustainable forest management. Neighborhood indices that consider the position of individual trees and the interaction between adjacent trees have received widespread attention, including size ratio, intermixing degree, angular scale, etc. It is of great significance to evaluate forest spatial structure through domain indicators, formulate corresponding strategies to change tree species diversity, optimize the spatial distribution of trees, reduce the degree of competition, and thus improve the functional quality of forest ecosystems. It is of great significance to use neighborhood indicators of different dimensions to formulate forest planning strategies.

[0003] As a management measure in forest planning, felling can not only promote tree growth and improve forest structure, but also maintain biodiversity and prevent pests and diseases and fires. In single-tree forest felling, since felling is restricted by many factors, which trees are removed and with what strategy is the key. Previous studies focused on using trunk quality, tree species diversity, adjacent relationships between trees, economic standards, etc. as selection criteria for felled trees to construct felling models, which has certain limitations and ignores the potential impact of the properties of retained trees after felling and their spatial distribution characteristics on the function of forest ecosystems. Summary of the invention

[0004] In view of the problems of the prior art, the present invention provides a forest stand spatial structure optimization method considering neighborhood index and particle swarm algorithm based on field survey data, structural multi-objective optimization algorithm and comprehensive decision index of forest stand spatial structure.

[0005] A stand spatial structure optimization method considering neighborhood index and particle swarm algorithm comprises the following steps: using conventional field measurement methods to record the tree height, breast diameter, crown width, height under branches, tree age, tree species, slope, altitude and canopy density information of all investigated trees; using handheld real-time dynamic measuring instruments to record the longitude and latitude coordinates of trees; comprehensively considering the influence of horizontal spatial distribution of trees, competition and species diversity on the stand spatial structure; according to the main influencing factors of forest spatial structure and combined with collected tree attribute data, three stand structure indicators of angular scale, size ratio and intermixing degree are calculated; at the same time, three objective functions are used as constraint conditions of the evaluation method to construct a comprehensive evaluation index of stand spatial structure; the stand structure covers forest distribution pattern, intermixing degree and size differentiation, which is used for stand structure evaluation and development succession prediction; by combining with particle swarm algorithm, the particle swarm position is initialized, the position of particles in the particle swarm is continuously iterated and updated, and the optimal solution set is found; after each iteration, the comprehensive index of stand spatial structure is calculated, and it is judged whether the stand spatial structure reaches the set optimization target; combined with felling quota, it is finally judged whether the stand structure reaches the optimal optimization effect.

[0006] The advantages of the present invention are: to make up for the single optimization target in traditional structural optimization, the optimization targets in multiple dimensions often conflict with each other, showing the characteristics of nonlinearity and discontinuity, to achieve accurate improvement of forest stand quality, to further accelerate the application of forest stand spatial structure optimization felling simulation technology, and to provide certain technical support for improving the level of intelligent forest management.

[0007] The multi-objective forest stand spatial structure optimization method proposed in the present invention is simple, efficient, and applicable to different management scenarios. It fully considers the relationship between forest structure and multiple management objectives, and can realize the comprehensive evaluation and analysis of forest stand horizontal spatial structure, competition, and biodiversity before and after logging. Its accuracy fully meets the application needs of grassroots forestry.

[0008] The present invention can accurately locate the distribution of trees with low intermixture, large angular scale, and large size ratio, and achieve accurate and efficient positioning and felling of felled trees. Trees with lower intermixture values ​​are more likely to be felled, trees with higher size ratios are more likely to be felled, and the number of trees with higher or moderate angular scale indices is relatively small. This shows that the method of the present invention can accurately capture individual trees with higher competitive advantages, lower intermixture, and horizontal distribution clustering, such as trees with higher size ratios or trees with clustered spatial patterns, and achieve accurate felling. The model we proposed can accurately locate and fell these trees that may have a negative impact on the forest stand structure, tree species composition, or the intensification of tree competition.

[0009] Compared with the traditional linear optimization model, the model of the present invention has the advantages of fast finding of the optimal solution set, low redundancy and good optimization effect in terms of the efficiency and effect of optimizing the spatial structure of the forest stand. In the experiment of five natural sample plots, based on the data of each tree inspection, the multi-group particle swarm algorithm, particle swarm algorithm and Monte Carlo were used to simulate felling, and the multi-objective particle swarm algorithm model was superior to the basic particle swarm model and Monte Carlo. For uniform distribution, random distribution and clustered distribution, the average value of the relative increase of the index under the multi-objective particle swarm model is slightly higher than that of the particle swarm model, and much higher than that of the Monte Carlo model. At the same time, the average number of iterations of the multi-group particle swarm algorithm is much less than the number of simulations of the Monte Carlo model, and is also slightly less than that of the ordinary particle swarm model. When the multi-objective particle swarm model integrates multi-group target strategies, the number of iterations is obviously better than the ordinary particle swarm algorithm. The method of the present invention has advantages in the optimization of multi-dimensional forest stand spatial structure.

[0010] The model of the present invention can accurately and efficiently solve the felling intensity of the sample plots under different competition, mixed and site conditions, and realize the optimization of the forest stand structure under the optimal felling intensity. In the model of the present invention, as the thinning intensity increases, the mixed degree, angle scale, and size ratio show a similar reverse U-shaped trend, and the maximum value of the relative increase ratio usually appears at the T2 intensity, indicating that under different thinning intensities, the overall forest stand spatial structure is affected to varying degrees, and it is determined that removing 15% of the trees is the best thinning intensity for optimizing the forest stand structure.

[0011] The model of the present invention has achieved a certain degree of optimization for the degree of intermixing, size ratio, angular scale and forest stand comprehensive evaluation index to varying degrees, and the overall quality of the forest stand spatial structure has been improved. After the forest stand is optimized by the model proposed by the present invention, the degree of competition is reduced, the degree of intermixing is increased, and the distribution is closer to random distribution or uniform distribution. The proposed forest stand spatial structure multi-objective optimization method has a good optimization effect.

[0012] The multi-objective structural method based on particle swarm algorithm is simple, efficient, and applicable to different management scenarios. It fully considers the relationship between forest structure and multiple management objectives, and achieves the overall optimization of the spatial structure, competition, and biodiversity of the forest stand before and after logging. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. As shown in the figure:

[0014] Figure 1 It is a technical flow chart of the present invention.

[0015] Figure 2 This is a flowchart of the forest stand spatial structure optimization of the present invention.

[0016] Figure 3 It is a frequency distribution diagram of the mixed degree, angular scale and size ratio of felled wood of the present invention.

[0017] Figure 4 It is a graph showing changes in various indicators before and after felling according to the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Example 1: A method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm, such as Figure 1 As shown, conventional field measurement methods were used to record information such as tree height, breast diameter (≥5cm), crown width, height under branches, tree age, tree species, slope, altitude, and canopy density of all surveyed trees.

[0020] The longitude and latitude coordinates of trees were recorded using a handheld real-time dynamic measuring instrument.

[0021] Taking into account the impact of horizontal spatial distribution of trees, competition and species diversity on the spatial structure of the forest stand, the three stand structure indices of angular scale, size ratio and mixed degree were calculated according to the main influencing factors of forest spatial structure and combined with the collected tree attribute data. At the same time, the three objective functions were used as constraints of the evaluation method.

[0022] Construct a comprehensive evaluation index for forest stand spatial structure.

[0023] Stand structure covers forest distribution pattern, degree of intermixing and size differentiation, and can be used for stand structure evaluation and prediction of its development and succession by considering the position of individual trees and the neighborhood index of the interaction between adjacent trees, including size ratio, degree of intermixing, angular scale, etc.

[0024] The spatial structure of forests is evaluated through domain indicators, and corresponding strategies are formulated to change tree species diversity, optimize forest spatial distribution, and reduce competition, thereby improving the functional quality of forest ecosystems. By combining the particle swarm algorithm, the particle swarm position is initialized, and the position of particles in the particle swarm is continuously updated iteratively to find the optimal solution set. After each iteration, the comprehensive index of the forest stand spatial structure is calculated, and it is judged whether the forest stand spatial structure has reached the set optimization goal. Combined with the felling limit, it is finally judged whether the forest stand structure has achieved the optimal optimization effect.

[0025] like Figure 2 As shown in the figure, the forest space corresponds to the solution of the algorithm, the tree coordinates correspond to the coordinates of the particles, and the optimization model function is used as the fitness function, which converts the forest structure problem into an optimization process of the algorithm. The final output of the algorithm is the target tree that needs to be adjusted or recorded in the forest stand.

[0026] Import the spatial location information and individual tree attribute data of all trees, such as DBH, tree height and crown width. Start the simulated felling. First, initialize the stand structure optimization model, then analyze the stand spatial structure characteristics, obtain the calculated stand structure index, and then calculate the stand spatial structure comprehensive evaluation index. Use the particle swarm optimization algorithm to calculate felled trees and retained trees. Among them, when using the particle swarm optimization algorithm, the generated particles are divided into multiple subgroups, and the initial position and speed of each particle are randomly generated. The speed and position are continuously updated iteratively. Each iteration calculates the comprehensive evaluation index value of the stand spatial structure, and outputs the trees that need to be felled and the trees that need to be retained. Finally, calculate the comprehensive evaluation index of the stand spatial structure after felling, and combine the felling quota target to check whether the stand management effect meets the set value. If the set management target is met and the comprehensive evaluation index of the stand spatial structure has no significant change, stop searching and output the target tree, and end the optimization. Otherwise, if the set value is not met, continue the optimization, return and continue to search for the best solution.

[0027] like Figure 3 As shown in the figure, the distribution of the mixed degree, angular scale, and size ratio of the harvested trees were analyzed. The distribution of the spatial structure index of the harvested trees in the five sample plots (Dahu, Longhu, Dashan, Lutou and Wuyun) was different.

[0028] In all five sample sites (Dahu, Longhu, Dashan, Lutou and Wuyun), the distribution of the mixedness index showed that trees with lower values ​​were more likely to be harvested, and the average index values ​​of harvested trees ranged from 0.2 to 0.36. The number of trees with high or moderate mixedness indices was relatively small.

[0029] The size ratio distribution showed that trees with higher values ​​were more likely to be harvested, and the mean values ​​of the index for harvested trees ranged from 0.7 to 0.82, with relatively fewer trees having low or moderate size ratios.

[0030] The distribution of the angular scaling index showed that trees with higher values ​​were more likely to be harvested, and the mean values ​​of the index for harvested trees ranged from 0.6 to 0.7. The number of trees with high or moderate angular scaling indices was relatively small.

[0031] Trees with higher size ratios or clustered tree spatial patterns were more likely to be harvested, and individual trees with higher competitive advantage, lower intermixing degree, and horizontal distribution clustering were more likely to be harvested, indicating that these trees may have a negative impact on stand structure, tree species composition, or the intensification of tree competition.

[0032] like Figure 4 As shown in the figure, in all natural plots (Da Hu, Long Hu, Da Shan, Lu Tou and Wu Yun), the degree of intermixing increased significantly with the increase of thinning intensity.

[0033] However, the values ​​of size ratio and angular scale decreased significantly between two consecutive thinning intensities in all plots.

[0034] For the relative increase ratio of intermixture degree, size ratio and angular scale, it was observed that with the gradual increase of thinning intensity, the three showed a similar inverse U-shaped trend, and the maximum relative increase ratio usually occurred at T2 intensity, indicating that under different thinning intensities, the overall spatial structure of the forest stand was affected to varying degrees, and removing 15% of the trees (T2 intensity) was the optimal thinning intensity for optimizing the forest stand structure.

[0035] The degree of intermixing, size ratio and angular scale have been optimized to a certain extent, such as the degree of competition has been reduced, the degree of intermixing has been increased, and the distribution is closer to random distribution or uniform distribution. From the results, it can be seen that the proposed multi-objective optimization method for stand spatial structure has a good optimization effect. The multi-objective structure method based on particle swarm algorithm is simple, efficient, and applicable to different management scenarios. It fully considers the relationship between forest structure and multiple management objectives, and realizes the overall optimization of stand horizontal spatial structure, competition, and biodiversity before and after felling.

[0036] Example 2: Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown in the figure, a forest stand spatial structure optimization method considering neighborhood index and particle swarm algorithm mainly involves forest management simulation technology in forest management and multi-objective optimization technology in computer science.

[0037] A stand spatial structure optimization method considering neighborhood index and particle swarm algorithm includes two aspects: construction of comprehensive evaluation index of stand spatial structure and optimization of stand spatial structure based on particle swarm algorithm. The steps are as follows:

[0038] Step 1. Construction of comprehensive evaluation index of forest stand spatial structure:

[0039] Step 1.1) Stand structure index:

[0040] In order to ensure the stability and health of the forest, it is necessary to restore and protect the number of trees, and timely evaluate and adjust the spatial structure of the forest, so as to build a healthy and stable forest management system. In the present invention, representative artificial mixed forest observation plots are selected for plot investigation, and the investigation factors include single tree position coordinates, breast diameter, tree height, crown width, etc. The adjacent trees of the trees in the plot are calculated, and the stand structure index of the stand is calculated. The angular scale, mixed degree, and size ratio are selected to construct a comprehensive evaluation index of the stand spatial structure.

[0041] The stand spatial structure index includes three indicators: angular scale, size ratio and mixed degree. The formula is as follows:

[0042]

[0043]

[0044] Among them, W i is the angular scale of the central tree i, n is the number of neighboring trees, and Z ij is the value of the angular dimension of the center tree i relative to the adjacent tree j.

[0045] When the angle between the center tree i and the adjacent tree j is less than the standard angle, Z ij =1, otherwise, Z ij =0.

[0046] U i is the size ratio of the center tree i, n is the number of adjacent trees, k ij It is the value of the ratio of the size of the central tree i to the adjacent tree j.

[0047] When the neighbor tree j is smaller than the center tree i, k ij =0, otherwise, k ij =1.

[0048] M ij is the degree of intermixing of the central tree i, n is the number of adjacent trees, v ij is the value of the degree of intermixing of the central tree i relative to the adjacent tree j. When the adjacent tree j and the central tree i are of the same species, v ij =0, otherwise, v ij =1.

[0049] The number n of adjacent trees selected by the present invention is 4.

[0050] Step 1.2) Construction of comprehensive evaluation index of forest stand spatial structure:

[0051] According to the main influencing factors of forest spatial structure, three objective functions, namely angular scale, size ratio, and intermixed degree, are selected as the constraints of the evaluation method. The purpose of constructing a multi-objective optimization model is to optimize the forest stand structure. The output of the model is the target tree that needs to be regulated. Among them, trees with low intermixed degree, large size ratio, and large angular scale need to be adjusted. The forest stand spatial structure optimization model is:

[0052]

[0053]

[0054] L i =f(M i , W i , U i )

[0055]

[0056] Among them, i represents the number of the central tree, j represents the number of the adjacent tree, and M i , W i , U i Indicates the degree of intergrowth, angular scale and size ratio of the target trees; ω m ,ω w ,ω u They are the weights corresponding to the three optimization constraint objectives: mixing degree, angular scale and size ratio.

[0057] L i Represents the comprehensive evaluation index of the stand spatial structure of central tree i.

[0058] L represents the comprehensive evaluation index of the overall forest stand spatial structure of the sample plot.

[0059] N represents the total number of trees; They represent the standard deviation of the degree of intermixing, the angular scale, and the size ratio, respectively.

[0060] The maximize and minimize here represent finding the maximum and minimum values ​​respectively.

[0061] Step 2. Multi-objective optimization of forest stand spatial structure:

[0062] The forest space corresponds to the solution of the algorithm, the tree coordinates correspond to the coordinates of the particles, and the optimization model function is used as the fitness function, which converts the forest structure problem into an optimization process of the algorithm. For ordinary particle swarms, N particles are randomly initialized, and then these particles are divided into subgroups through the K-means algorithm. The particle with the shortest distance is selected as the center. The final output of the algorithm is the target tree that needs to be adjusted or recorded in the forest stand:

[0063] Step 2.1: Import the spatial location information and individual tree attribute data of all trees, such as DBH, tree height and crown width.

[0064] Step 2.2: Randomly generate an initial population of size N. Divide the generated particles into multiple sub-populations, and randomly generate the initial position and velocity of each particle.

[0065] Step 2.3: Calculate the Euclidean distance from each particle to the center of the group, adjust the position of the particle so that it falls on the nearest group, and calculate the individual fitness value based on the initial particle.

[0066] Step 2.4: Update the speed and position, adjust the real-time position of the current particle, and calculate the value of the comprehensive evaluation index of the forest stand spatial structure.

[0067] Step 2.5: Calculate the distance between particle groups and insert or delete particle groups based on the distance.

[0068] Step 2.6: For the inserted new particle group, initialize its particles.

[0069] Step 2.7: Determine the change of the comprehensive evaluation index L of the forest stand spatial structure. If the value of the comprehensive evaluation index L of the forest stand spatial structure has not changed, stop searching and output the target tree. Otherwise, return and continue searching for the best solution.

[0070] Example 3: Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown in the figure, a stand spatial structure optimization method considering neighborhood index and particle swarm algorithm is used, and an example of multi-objective optimization of stand spatial structure in Hunan State Lutou Forest Farm is presented.

[0071] Step 1. Data collection:

[0072] Five coniferous and broad-leaved mixed forest plots were established in the state-owned forest farm. Each plot was 20m×30m in size and a 3m buffer zone was set up. The tree height, breast diameter (≥5cm), crown width and species of all surveyed trees were recorded using conventional field measurement methods.

[0073] The latitude and longitude coordinates of the trees were recorded using a handheld RTK.

[0074] The spatial distribution of the five natural plots and the statistical characteristics of the trees are shown in Table 1.

[0075] Table 1 Basic statistical characteristics of the five study plots

[0076]

[0077]

[0078] Step 2. Construction of comprehensive evaluation index of forest stand spatial structure:

[0079] Taking into account the impact of horizontal spatial distribution of trees, competition and species diversity on the spatial structure of forest stands, the present invention selects three objective functions, namely, angular scale, size ratio and intermixing degree, as the constraints of the evaluation method according to the main influencing factors of forest spatial structure. A comprehensive evaluation index of forest spatial structure is constructed, and the model output is the target tree that needs to be regulated. Trees with low intermixing degree, large breast diameter size ratio and large angular scale are the key to affecting the overall structure of the forest stand and need to be adjusted. The dynamic multi-objective optimization model of forest spatial structure is:

[0080]

[0081] Where: i represents the number of the central tree, M i , Wi-Fi, U i Respectively represent the degree of intergrowth, angular scale and size ratio of the target trees; ω m ,ω w ,ω u are the weights corresponding to the three optimization constraint targets of intermixing degree, angular scale and size ratio. L represents the comprehensive evaluation index of the overall forest stand spatial structure of the sample plot. N represents the total number of trees: They represent the standard deviation of the degree of intermixing, the angular scale, and the size ratio, respectively.

[0082] The larger the angular scale, the more concentrated the trees are distributed; the smaller it is, the more evenly the trees are distributed; the larger the size ratio, the greater the competition; the greater the impact of surrounding trees on the trees; the larger the mixedness index, the higher the tree species richness; therefore, trees with a smaller comprehensive evaluation index should be cut down.

[0083] Step 3. 3D forest classification model:

[0084] Based on the data of each tree measurement in five natural plots, multi-population particle swarm algorithm, particle swarm algorithm and Monte Carlo were used to simulate felling. The M0-PS0 model outperformed the basic PS0 model and Monte Carlo. For uniform distribution, random distribution and clustered distribution, the average relative increase under the M0-PS0 model reached 31.94%, 29.51% and 24.71%, respectively, which was slightly higher than the PSO model and much higher than the RD-TH model. At the same time, the average number of iterations (AITE) of the multi-population particle swarm algorithm was about 26 to 43 times, which was much less than the 10,000 simulations of the RD-TH model and slightly less than the PSO model. When the M0-PS0 model was compared with the PS0 model, when the M0-PS0 model integrated multi-population target strategies, ALIV and AITE were obviously better than the PS0 model. Obviously, as shown in Table 2, the proposed method has advantages in the optimization of multi-dimensional forest stand spatial structure.

[0085] Table 2 Optimization effect of forest stand spatial structure of different models

[0086]

[0087] M0-PS0: multi-population particle swarm algorithm; PS0: ordinary particle swarm algorithm; RD-TH: Monte Carlo-based felling model; The average value of relative increase ratio; AITE: average number of iterations; ALIV: average value of comprehensive evaluation index of forest stand spatial structure; RUNN: number of felling runs.

[0088] Step 4. Tree growth related model:

[0089] In all natural plots, the degree of intermixture increased significantly with the increase of cutting intensity. However, for all plots, between two consecutive thinning intensities, the size ratio and angular scale values ​​decreased significantly with the increase of thinning intensity. Regarding the relative increase ratio of intermixture, size ratio, and angular scale, three similar inverse U-shaped trends were observed. As the thinning intensity gradually increased, the maximum value of the relative increase usually occurred at a cutting intensity of 15%, indicating that the overall stand spatial structure was affected to different degrees under different cutting intensities.

[0090] Step 5. Dynamic visualization simulation of forest management:

[0091] The distribution of intermixedness, angular scale, and size ratio of harvested trees was analyzed. The distribution of spatial structure index of harvested trees in the five plots was different. In all five plots, the distribution of intermixedness index showed that trees with lower values ​​were more likely to be harvested, with an average relative frequency of 33.7%. However, the number of trees with high or moderate intermixedness index was relatively small. Similarly, trees with higher size ratio or clustered spatial pattern of trees were more likely to be harvested, with the average relative frequency of size ratio and angular scale being 47.3% and 35.6%, respectively. Individual trees with higher competitive advantage, lower intermixedness, and horizontal distribution clustering were more likely to be harvested, indicating that these trees may have a negative impact on stand structure, tree species composition, or the intensification of tree competition.

[0092] It can be seen from the example results that the multi-objective optimization method for the spatial structure of the forest stand proposed in the present invention has a good optimization effect. The multi-objective structure method based on the particle swarm algorithm is simple, efficient, and suitable for different management scenarios. It fully considers the relationship between forest structure and multiple management objectives, and realizes the overall optimization of the spatial structure, competition, and biodiversity of the forest stand before and after felling, which can meet the application needs of grassroots forestry.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm, characterized in that: The following steps are included: the tree height, breast diameter, crown width, height under branches, tree age, tree species, slope, altitude and canopy density of all surveyed trees were recorded using conventional field measurement methods. Use a handheld real-time dynamic measuring instrument to record the longitude and latitude coordinates of the trees. Taking into account the impact of horizontal spatial distribution of trees, competition and species diversity on forest spatial structure, the three forest stand structure indices of angular scale, size ratio and intermixed degree were calculated according to the main influencing factors of forest spatial structure and the collected tree attribute data. At the same time, the three objective functions were used as constraints of the evaluation method. Construct a comprehensive evaluation index of forest stand spatial structure. Stand structure includes forest distribution pattern, degree of intergrowth and size differentiation, which is used for stand structure evaluation and development succession prediction. By combining the particle swarm algorithm, the particle swarm position is initialized, the position of the particles in the particle swarm is continuously updated iteratively, and the optimal solution set is found. After each iteration, the comprehensive index of the forest stand spatial structure is calculated, and it is judged whether the forest stand spatial structure reaches the set optimization goal. Combined with the felling limit, it is finally judged whether the forest stand structure has achieved the optimal optimization effect.

2. The method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm according to claim 1, characterized in that: Contains the following steps: The forest space corresponds to the solution of the algorithm, the coordinates of the trees correspond to the coordinates of the particles, and the optimization model function is used as the fitness function. The forest spatial structure problem is converted into an optimization process of the algorithm. The final output of the algorithm is the target tree that needs to be adjusted or recorded in the forest stand. Import the spatial location information and individual tree attribute data of all trees, such as breast diameter, tree height and crown width, and start simulated felling. First, initialize the stand structure optimization model, then analyze the spatial structure characteristics of the stand, obtain the calculated stand structure index, and then calculate the comprehensive evaluation index of the stand spatial structure. Use the particle swarm optimization algorithm to calculate the felled trees and retained trees. When using the particle swarm optimization algorithm, divide the generated particles into multiple subgroups, and randomly generate the initial position and speed of each particle. Continuously iterate and update the speed and position. Calculate the comprehensive evaluation index value of the stand spatial structure once each iteration, and output the trees that need to be felled and the trees that need to be retained. Calculate the comprehensive evaluation index of the stand spatial structure after felling. Combined with the felling limit target, check whether the stand management effect meets the set value. If the set management target is met and the comprehensive evaluation index of the stand spatial structure has no significant change, stop searching and output the target tree, and end the optimization. Otherwise, if the set value is not met, continue the optimization, return and continue to search for the best solution.

3. The method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm according to claim 1, characterized in that: The steps of constructing a comprehensive evaluation index of forest stand spatial structure are as follows: selecting representative artificial mixed forest observation plots for plot investigation, the investigation factors include single tree position coordinates, breast diameter, tree height, crown width, calculating the neighboring trees in the plot, calculating the forest stand structure index, selecting angle scale, mixed degree, size ratio to construct a comprehensive evaluation index of forest stand spatial structure, According to the main influencing factors of forest spatial structure, three objective functions, namely, angular scale, size ratio and intermixed degree, are selected as the constraints of the evaluation method. The multi-objective optimization model is constructed to optimize the forest stand structure. The output of the model is the target tree that needs to be regulated. Among them, trees with small intermixed degree, large size ratio and large angular scale need to be adjusted. The optimization model of forest stand spatial structure is: L i =f(M i ,W i ,U i ) Among them, i represents the number of the central tree, j represents the number of the adjacent tree, and M i , W i , U i Indicates the degree of intergrowth, angular scale and size ratio of the target trees; ω m ,ω w ,ω u are the weights corresponding to the three optimization constraints of mixed degree, angular scale and size ratio, L i represents the comprehensive evaluation index of the spatial structure of the forest stand of the central tree i, L represents the comprehensive evaluation index of the spatial structure of the overall forest stand of the plot, and N represents the total number of trees; They represent the standard deviation of the degree of mixing, angular scale and size ratio respectively, where maximize and minimize represent the solution of the maximum and minimum values ​​respectively.

4. The method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm according to claim 1, characterized in that: The steps of multi-objective optimization with stand spatial structure are as follows: the forest space corresponds to the solution of the algorithm, the tree coordinates correspond to the coordinates of the particles, the optimization model function is used as the fitness function, and the forest structure problem is converted into an optimization process of the algorithm. For ordinary particle swarms, N particles are randomly initialized, and then these particles are divided into subgroups through the K-means algorithm. The particle with the shortest distance is selected as the center. The final output of the algorithm is the target tree that needs to be adjusted or recorded in the stand.

5. The method for optimizing forest stand spatial structure by considering neighborhood index and particle swarm algorithm according to claim 4, characterized in that: The steps are as follows: Step 2.1, import the spatial location information and individual tree attribute data of all trees, such as breast diameter, tree height and crown width, Step 2.2: Randomly generate an initial population of size N, divide the generated particles into multiple sub-populations, and randomly generate the initial position and velocity of each particle. Step 2.3, calculate the Euclidean distance from each particle to the center of the group, adjust the position of the particle so that it falls on the nearest group, and calculate the individual fitness value based on the initial particle, Step 2.4: Update the speed and position, adjust the real-time position of the current particle, and calculate the value of the comprehensive evaluation index of the forest stand spatial structure. Step 2.5: Calculate the distance between particle groups and insert or delete particle groups according to the distance. Step 2.6: For the newly inserted particle group, initialize its particles. Step 2.7, determine the change of the comprehensive evaluation index L of the forest stand spatial structure. If the value of the comprehensive evaluation index L of the forest stand spatial structure does not change, stop searching and output the target tree; otherwise, return and continue searching for the best solution.

Citation Information

Cited By

  • Forest stand mixing optimization method and system based on tree cluster units

    CN120197779A

  • Method for improving carbon sequestration amount of target area

    CN120198242A

  • Methods for increasing carbon sequestration in target areas

    CN120198242B

  • Ststand density optimization method based on RTK space positioning

    CN120257668A

  • Larix glarch forest cultivation decision-making system based on multi-objective optimization

    CN120494558A