Larch forest cultivation decision system based on multi-objective optimization

Abstract

The present application relates to the technical field of artificial forest, specifically to a larch forest cultivation decision system based on multi-objective optimization, which comprises a spatial layout evaluation module, an intermediate cutting condition judgment module, a planting distance adjustment evaluation module, a climate adaptability analysis module and a stand trend prediction module. Through measuring tree height, diameter at breast height and crown size, combined with spatial position data, the system can more accurately evaluate the growth environment and mutual relationship of trees, real-time analysis of crown intersection area and growth rate allows immediate determination of the necessity of intermediate cutting, effectively reduces unnecessary tree felling, guarantees the health and productivity of the stand, real-time analysis of planting distance adjustment optimizes forest layout, improves forest ecological benefits and quality, climate adaptability analysis strengthens the adaptability evaluation of different tree species under specific climate conditions, ensures the suitability of tree species and environmental matching, supports biodiversity, improves the sustainable utilization and ecological function of resources, and realizes the balance between production and protection.

Description

Technical Field

[0001] This invention relates to the field of plantation technology, and in particular to a decision-making system for larch forest cultivation based on multi-objective optimization. Background Technology

[0002] In the field of plantation technology, the focus is on constructing the optimal forest spatial structure to cultivate healthy, stable, and high-quality forests. The key lies in meeting the multifunctional needs of modern forest management and promoting the optimization of forest spatial structure through precise and efficient timber harvesting decisions. Research on the spatial structure optimization of larch plantations in North China is crucial for improving forest productivity, promoting soil and water conservation and biodiversity protection, and is of great significance for improving the forest ecological environment.

[0003] Among them, the multi-objective optimization larch forest cultivation decision system uses multi-objective optimization technology to formulate and optimize larch forest cultivation strategies, aiming to balance the goals of forestry production and ecological protection. The system finds the optimal solution among multiple objectives (such as production costs, growth rate, timber quality and environmental impact). Its application is mainly used to provide scientific operational advice to forest farm managers, help formulate more effective forest management and logging plans, and achieve sustainable use of resources and maximization of ecological benefits.

[0004] Existing technologies rely on experience or relatively crude data processing, which limits the accuracy and efficiency of decision-making. The lack of real-time and accurate data analysis can easily lead to untimely or inappropriate thinning decisions, affecting the health and productivity of forest stands. Existing technologies lack flexibility in adjusting tree spacing and fail to respond quickly to the impact of climate change on tree growth, resulting in insufficient ability of forest stands to adapt to extreme weather events, which in turn affects the ecological benefits of forest land. In addition, existing technologies fail to fully consider changes in specific climatic conditions in tree species selection, which limits the effective protection of biodiversity and the balance between forestry production and ecological protection goals. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a multi-objective optimization-based decision-making system for larch forest cultivation.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a multi-objective optimization-based larch forest cultivation decision-making system, the system comprising:

[0007] The spatial layout assessment module uses data on larch tree height, diameter at breast height (DBH), and crown size to measure the physical distance between adjacent trees, calculate the crown overlap area, assess the spatial distribution of trees, determine the number of neighboring trees, and obtain the neighboring tree density index.

[0008] The thinning condition determination module identifies the neighboring trees of each tree based on the neighboring tree density index, collects data on the canopy interlacing area and growth rate, compares it with the benchmarks of canopy interlacing and growth rate, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier.

[0009] Based on the thinning requirement identifier, the planting spacing adjustment assessment module determines the tree number that needs intervention, measures the actual distance between trees, selects the standard planting spacing as a reference, analyzes the degree of deviation from the current spacing, assesses whether the planting spacing needs to be replanned or thinning is carried out, and generates planting spacing adjustment data.

[0010] Based on the planting distance adjustment data, the climate adaptability analysis module obtains the climate conditions required for larch forest cultivation, compares the data with the tree species' heat resistance, drought resistance, cold resistance and carbon sequestration potential, identifies tree species suitable for larch forest cultivation under the current climate conditions, and generates a tree species adaptability rating.

[0011] The present invention is improved in that the adjacent forest density index includes forest distance, coverage, and density; the thinning demand identifier includes growth competition intensity, disturbance level, and thinning priority; the planting distance adjustment data includes distance deviation, planting distance compliance, and planting distance adjustment demand; and the tree species adaptability rating includes climate adaptability, survival potential, and growth potential.

[0012] The present invention is improved in that the spatial layout evaluation module includes:

[0013] The canopy overlap calculation submodule, based on larch tree height, diameter at breast height (DBH), and canopy size data, obtains the center coordinates of each tree according to the spatial location data of the pine forest, using the following formula:

[0014]

[0015] Calculate the crown overlap area AS between each pair of adjacent trees. overlap , where rs i rs is the crown radius of the i-th tree, referring to the distance from the center of the trunk to the edge of the crown. j ds is the radius of the crown of the j-th tree. ij π is the straight-line distance between the center points of the i-th and j-th trees, and π is the mathematical constant pi.

[0016] The diameter at breast height (DBH) difference comparison submodule collects corresponding DBH data based on the crown overlap area and the number of each larch tree, evaluates the DBH difference between adjacent trees, and filters out trees that fall within the DBH difference benchmark range to obtain the DBH matching degree.

[0017] The neighboring tree density determination submodule assesses the spatial distribution of trees based on the diameter at breast height matching degree, counts the average number of neighboring trees in each area, and obtains the neighboring tree density index.

[0018] The present invention is improved in that the thinning condition determination module includes:

[0019] The density over-standard identification submodule obtains the coordinate data of each tree and its surrounding trees based on the neighboring tree density index, calculates the tree density around each target tree, compares it with the density benchmark, identifies the area of ​​trees with excessive density, and obtains the cross-identification result.

[0020] The growth rate assessment submodule calls the interleaving recognition results, collects the annual change in tree diameter at breast height, compares it with the average growth data of the same tree species in the region, identifies whether the individual growth rate is lower than the average level, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier.

[0021] The present invention is improved in that the planting spacing adjustment evaluation module includes:

[0022] The tree numbering and identification submodule collects the numbering information of the identified trees in the forest land based on the thinning requirement identifier. By comparing the numbering label data with the tree distribution image, it determines the tree numbers of densely distributed or non-standard thinning areas and generates a set of trees that need intervention.

[0023] The spacing offset analysis submodule calls the set of trees that need intervention, detects the planting distance between numbered trees, analyzes the degree of offset from the current tree spacing based on the standard planting spacing, and obtains the tree spacing offset data.

[0024] The planting distance intervention assessment submodule uses the following formula based on the tree spacing offset data:

[0025]

[0026] Calculate the impact value (IH) of planting spacing intervention to assess whether planting spacing needs to be replanned or thinning carried out, and generate planting spacing adjustment data. ΔH is the average spacing offset of the g-th tree. s ρ is the standard spacing offset baseline value, ρ is the probability of tree overlap within the numbered area, CH is the intervention cost factor for thinning operations, and L is the standard spacing offset baseline value. g N is the topographic correction factor for the g-th tree. IH It represents the total number of trees.

[0027] The present invention is improved in that the climate adaptability analysis module includes:

[0028] Based on the vegetation spacing adjustment data, the climate parameter extraction submodule extracts the annual average temperature, annual precipitation and corresponding monthly variation data of the current region. Combined with the original climate records, it calculates the annual temperature change rate, interannual precipitation variation and climate change range, identifies and removes precipitation extremes and temperature anomalies, and establishes a regional climate parameter set.

[0029] The tree species rating generation submodule compares the heat resistance, drought resistance, cold resistance and carbon sequestration potential of larch trees with the set of regional climate parameters, selects tree species that are in the suitable range under the current climate conditions, and classifies them into three levels: high, medium and low, to generate a tree species adaptability rating.

[0030] The present invention has an improvement, wherein the system further includes:

[0031] The stand trend prediction module selects target tree species for planting based on the tree species adaptability rating, collects data on the tree diameter at breast height (DBH) growth rate and crown expansion rate, and combines this data with climate prediction data for future periods to conduct simulation analysis, assess the potential change trends of stand density and structure, and obtain stand change assessment results.

[0032] The results of the stand change assessment include information on growth dynamics and changes in stand structure.

[0033] The present invention is improved in that the stand trend prediction module includes:

[0034] The tree species selection submodule analyzes the tree diameter at breast height (DBH) growth rate and crown expansion rate based on the tree species adaptability rating, compares them with climate prediction data for future time periods, selects target tree species for planting, identifies the impact on growth data, and obtains an adaptability screening score.

[0035] The structural trend assessment submodule evaluates the potential trends in stand density and structure based on the aforementioned adaptability screening score, using the following formula:

[0036]

[0037] The stand change assessment results TB were obtained, of which DB y VB is the average density of the y-th tree species. y It is the growth potential coefficient of the yth tree species, GB y PB is the current diameter at breast height (DBH) growth rate of tree species in category y. y RB represents the expected growth rate of diameter at breast height (DBH) for the y-th tree species corresponding to climate, RB represents the fluctuation range of the annual variation growth rate of the target tree species, and MB represents... u It is the age structure ratio of the u-th forest layer, AB u It is the average age of the trees at the u-th level, n tb It represents the total number of tree species, m tb It refers to the number of age structure layers in the forest stand.

[0038] The present invention is improved in that the spatial layout evaluation module executes multi-objective collaborative optimization logic, and uses the linear combination of the canopy overlap area and the density index of adjacent trees as a constraint condition;

[0039] The spatial competition index is obtained by multiplying the density and coverage, and the coverage threshold is set to 1.2 times the canopy radius.

[0040] The distance between the tree crown radius and the center point is calculated using normalized units in meters.

[0041] The present invention is improved in that the neighboring tree density index is calculated using the Hegyi competition index to assess the intensity of competition between trees;

[0042] The aforementioned canopy interlacing benchmark refers to the critical value of canopy closure, with an interference limit of 70%-80%. Exceeding this limit indicates that the trees are too dense and thinning should be considered.

[0043] The standard planting spacing is 2Mx2M or 2Mx3M.

[0044] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0045] In this invention, by measuring tree height, diameter at breast height (DBH), and crown size, and combining this with spatial location data, the system can more accurately assess the growth environment and interrelationships of trees. Real-time analysis of crown overlap area and growth rate allows for immediate determination of the necessity of thinning, effectively reducing unnecessary logging, ensuring forest health and productivity, and real-time analysis of planting distance to adjust and optimize forest stand layout, thereby improving forest ecological benefits and quality. Climate adaptability analysis strengthens the assessment of the adaptability of different tree species under specific climatic conditions, ensuring the suitability of tree species to the environment, supporting biodiversity, enhancing the sustainable use of resources and ecological functions, and achieving a balance between production and conservation. Attached Figure Description

[0046] Figure 1 This is a system flowchart of the present invention;

[0047] Figure 2 This is a flowchart of the spatial layout evaluation module in this invention;

[0048] Figure 3 This is a flowchart of the intermediate valve condition determination module of the present invention;

[0049] Figure 4 This is a flowchart of the planting spacing adjustment and evaluation module in this invention;

[0050] Figure 5 This is a flowchart of the climate adaptability analysis module in this invention;

[0051] Figure 6 This is a flowchart of the forest stand trend prediction module in this invention. Detailed Implementation

[0052] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0053] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0054] Example

[0055] Please see Figure 1 This invention provides a technical solution: a multi-objective optimization-based larch forest cultivation decision system comprising:

[0056] The spatial layout assessment module is based on the tree height, diameter at breast height (DBH), and crown size data of larch. According to the spatial location data of pine forest, it measures the physical distance between adjacent trees, calculates the crown overlap area, compares the data between trees based on the difference in DBH, assesses the spatial distribution of trees, and determines the number of neighboring trees to obtain the neighboring tree density index.

[0057] The thinning condition determination module identifies the neighboring trees of each tree based on the density index of neighboring trees, collects data on the canopy overlap area and growth rate, compares it with the benchmarks of canopy overlap and growth rate, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier.

[0058] The planting spacing adjustment assessment module identifies the trees that need intervention based on the thinning demand identifier, measures the actual distance between the trees, selects the standard planting spacing as a reference, analyzes the degree of deviation from the current spacing, assesses whether the planting spacing needs to be replanned or thinning needs to be carried out, and generates planting spacing adjustment data.

[0059] The climate adaptability analysis module obtains the climate conditions required for larch forest cultivation based on planting distance adjustment data, compares the data with the heat resistance, drought resistance, cold resistance and carbon sequestration potential of tree species, identifies tree species suitable for larch forest cultivation under the current climate conditions, and generates tree species adaptability ratings.

[0060] The stand trend prediction module selects target tree species for planting based on tree species adaptability rating, collects data on tree diameter at breast height (DBH) growth rate and crown expansion rate, and combines this data with climate prediction data for the next 3 to 5 years to conduct simulation analysis, assess the potential change trends of stand density and structure, and obtain stand change assessment results.

[0061] The spatial layout assessment module executes multi-objective collaborative optimization logic, using a linear combination of canopy overlap area and adjacent tree density as a constraint.

[0062] The spatial competition index is obtained by multiplying the density and coverage, and the coverage threshold is set to 1.2 times the canopy radius.

[0063] Both the canopy radius and the distance to the center point are calculated using normalization in meters.

[0064] The Hegyi competition index was used to calculate the density index of neighboring trees to assess the intensity of competition between trees.

[0065] The Hegyi competition index is specifically:

[0066]

[0067] Among them, CI i DBH is the competition index of target tree i. i The diameter at breast height (DBH) of target tree i j It is the diameter at breast height (DBH) of the neighboring tree j, D ij It is the horizontal distance between trees i and j.

[0068] The canopy interleaving benchmark refers to the critical value of canopy closure. The disturbance limit is 70%-80%. If it exceeds this, it means that the trees are too dense and thinning should be considered.

[0069] The standard planting spacing is 2Mx2M or 2Mx3M.

[0070] The indicators of adjacent forest density include tree distance, coverage, and density; the indicators of thinning demand include growth competition intensity, disturbance level, and thinning priority; the data on planting distance adjustment include distance deviation, planting distance compliance, and planting distance adjustment demand; the tree species adaptability rating includes climate adaptability, survival potential, and growth potential; and the results of the stand change assessment include growth dynamics and information on changes in stand structure.

[0071] Please see Figure 2 The spatial layout assessment module includes:

[0072] The canopy overlap calculation submodule, based on larch tree height, diameter at breast height (DBH), and canopy size data, obtains the center coordinates of each tree according to the spatial location data of the pine forest, using the following formula:

[0073]

[0074] Calculate the crown overlap area AS between each pair of adjacent trees. overlap , where rs i rs is the crown radius of the i-th tree, referring to the distance from the center of the trunk to the edge of the crown. j ds is the radius of the crown of the j-th tree. ij π is the straight-line distance between the center points of the i-th and j-th trees, and π is the mathematical constant pi.

[0075] First, data on the height, diameter at breast height (DBH), and crown width of each larch tree within the target area were collected. Through on-site measurement, data such as a tree height of 18.5 meters, DBH of 32.4 centimeters, and crown diameter of 5.6 meters could be obtained. The crown radius was calculated by dividing the crown diameter by 2, resulting in 2.8 meters. Simultaneously, the spatial coordinates were measured, for example, X = 12.0 meters and Y = 8.0 meters, as the spatial center coordinates. Next, the center point coordinates and crown radius of each larch tree in the sample plot were recorded, and the trees were paired. The Euclidean distance formula was used to calculate the straight-line distance between the center points of any pair of trees. If the coordinates of T1 and T2 are (12.0, 8.0) and (15.0, 10.0) respectively, the distance between them is calculated as follows:

[0076]

[0077] The corresponding canopy radius is rs i = 2.8 meters, rs j =2.5 meters. Substitute the above parameters into the formula for the overlapping area of ​​the tree canopy to calculate the canopy area:

[0078] π·(2.8 2 +2.5 2 =3.1416·(7.84+6.25) =3.1416·14.09≈44.27;

[0079] Calculate the distance product and radius term:

[0080] ds 12 ·(rs i +rs j )=3.61·(2.8+2.5)=3.61·5.3≈19.13;

[0081] The overlapping area is:

[0082] AS overlap=44.27 - 19.13 = 25.14;

[0083] The result is the area of ​​the overlapping canopy region between the two larch trees T1 and T2, which is 25.14 square meters. This area describes the degree of overlap between the two tree canopies on the horizontal plane and can serve as the basic data input for subsequent spatial structure assessment and forest interaction analysis.

[0084] The diameter at breast height (DBH) difference comparison submodule collects corresponding DBH data based on the crown overlap area and the number of each larch tree, evaluates the DBH difference between adjacent trees, and filters out trees that fall within the DBH difference benchmark range to obtain the DBH matching degree.

[0085] For tree pairs whose canopy overlap area has been calculated, they are numbered and extracted accordingly. If there is canopy overlap between T1 and T2 calculated previously, then the diameter at breast height (DBH) data of T1 and T2 need to be obtained. According to the sample data, the DBH of T1 is 32.4 cm and that of T2 is 29.7 cm. The difference in DBH between the two trees is calculated using the absolute value difference expression, i.e., |D i -D j Substituting the data, we get |32.4-29.7|=2.7 cm. Then, we set a baseline range for diameter at breast height (DBH) difference. In general forestry research, a difference within 3.0 cm is considered a similar configuration. To ensure rationality, this baseline value is set to ≤3.0 cm after statistical analysis of field observation samples, referencing the distribution pattern of natural competition intensity among tree species in different forest stands. We further compare the difference of 2.7 cm with the baseline value. Tree pairs that meet the condition are selected as "DBH matching" combinations. If the difference of a group is 3.6 cm, it will be excluded. Finally, all tree pairs that meet the matching conditions are marked as qualified tree pairs, forming a DBH matching dataset. This dataset will be used in the next module to locate their spatial proximity structural relationships.

[0086] The neighboring tree density determination submodule assesses the spatial distribution of trees based on the diameter at breast height (DBH) matching degree, counts the average number of neighboring trees in each area, and obtains the neighboring tree density index.

[0087] The program retrieves the selected tree pairs with matching diameter at breast height (DBH). A neighborhood search radius is set around each tree, typically 5 meters, based on the average tree spacing in a larch forest. Within this radius, it searches for other trees with overlapping canopies and matching DBH. For example, if tree T1 is the center and there are three trees (T2, T4, and T5) within 5 meters, their neighbor count is 3. This process is repeated for all eligible trees. The total number of neighbor trees for each tree in the sample is then calculated and divided by the number of eligible samples. Assuming a total of 93 neighbor trees and 30 sample trees, the average number of neighbor trees is 93 / 30 = 3.1. Finally, the program retrieves the area data of the analysis region. For example, if the sample plot area is 900 square meters, the density index is calculated as 3.1 × 30 / 900 = 0.103 trees / square meter, which represents the number of neighboring individuals with potential spatial overlap and similar growth status within a unit area. This can be further used as a reference value for spatial competition density. During the process, the "statistical" operation must be completed strictly according to the spatial radius and screening results. The unit of the density result must be consistent with the area unit. This result shows that under the condition of relatively small difference in diameter at breast height and spatial proximity, the number of matching neighboring trees within a unit area of ​​forest land is 0.103 trees / square meter, which can be used as a quantitative indicator for assessing the spatial structure status.

[0088] Please see Figure 3 The thinning condition determination module includes:

[0089] The density over-standard identification submodule obtains the coordinate data of each tree and its surrounding trees based on the density index of neighboring trees, calculates the density of trees around each target tree, compares it with the density benchmark, identifies the area of ​​trees with excessive density, and obtains the cross-identification result.

[0090] Spatial coordinate data of each tree was obtained through geographic surveying and mapped onto a two-dimensional plane for spatial distance calculation. Then, a 5-meter radius neighborhood was established around each tree, and the number of trees within this neighborhood was counted using a geometric distance judgment method. For example, if a target tree is numbered T002 and located at coordinates (16.5, 23.0), a 5-meter radius buffer zone was established around this point. The Euclidean distance between the target tree and other trees was measured one by one. If the distance between the target tree and other trees was less than or equal to 5 meters, the number of neighboring trees was recorded as 8. If the preset density benchmark was 6 trees / 5 meters, then the density of neighboring trees of T002 exceeded the benchmark. According to this judgment standard, the tree was marked as an individual with excessive density. Simultaneously, the two-dimensional projection boundary of the canopy was extracted for the target tree and its neighboring trees, and a graphic overlay method was used. To determine the overlapping area of ​​tree canopies, the target tree canopy shape is overlapped with the outline of each neighboring tree canopy, and the overlapping area is calculated. For example, if the target tree canopy is 12 square meters and the overlapping area of ​​neighboring trees is 4.5 square meters, then the overlapping ratio is 4.5 / 12 = 0.375, or 37.5%. If the overlap area benchmark is set at 30%, this value exceeds the set standard. Therefore, this individual simultaneously meets the density and overlapping conditions. The density benchmark value and the overlap area benchmark value are set based on historical survey data. The median density is 5 trees, and the benchmark is set up to 6 trees. The overlap area benchmark is obtained by statistically analyzing the overlap ratio of multiple samples, with an average of 0.28, which is rounded up to 0.30 as a fixed reference. After completing the above determination for multiple sample trees, individuals that simultaneously meet the density and overlapping conditions are selected to obtain the overlap identification result.

[0091] The growth rate assessment submodule calls the interleaving recognition results, collects the annual change in tree diameter at breast height, compares it with the average growth data of the same tree species in the region, identifies whether the individual growth rate is lower than the average level, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier.

[0092] Select target trees identified as having density overlap interference and read their diameter at breast height (DBH) data from year 1 to year 3. For example, tree number T002 has a DBH of 23.5 cm in year 1 and 24.2 cm in year 3. The average annual growth is (24.2 - 23.5) / 2 = 0.35 cm / year. Then, retrieve the average annual growth data of samples of the same tree species and age in the same area, for example, 0.40 cm / year. Calculate the growth deviation as 0.40 - 0.35 = 0.05 cm / year. If the criterion for slow growth is set as a 20% reduction from the average value... If the growth rate is less than 0.32 cm / year, it is considered a low growth rate. According to the above standard, if the growth of an individual is lower than this value, for example, the tree numbered T004, whose diameter at breast height increased from 21.0 cm to 21.5 cm, has an annual growth rate of (21.5-21.0) / 2 = 0.25 cm / year, it is judged as a low-growth individual. Combined with its intercropping identification record, if the intercropping area ratio is 38%, which is also higher than the intercropping benchmark value of 30%, then the tree meets both the intercropping and insufficient growth conditions. It is further determined that it meets the requirement for thinning, and this type of individual is marked as a thinning object, generating a thinning requirement identifier.

[0093] Please see Figure 4 The planting spacing adjustment assessment module includes:

[0094] The tree numbering and identification submodule is based on the thinning requirement identifier. It collects the numbering information of the identified trees in the forest land, compares the numbering data with the tree distribution image, determines the tree numbers of densely distributed or non-standard thinning areas, and generates a set of trees that need intervention.

[0095] The process involves acquiring the identification numbers of all target trees within the forest area. High-resolution image sequences are obtained using a drone equipped with a visible light and near-infrared imaging system. These images are then combined with historical geographic annotation data to extract the spatial locations and corresponding identification numbers of all trees in the target area. Subsequently, the acquired images are analyzed by region, spatially aligning the tree crown features in the images with the existing identification data. If the coordinates of a point in the image deviate from the coordinates recorded in the identification table by more than 0.5 meters, it is marked as an identification number requiring verification. This process can be manually verified to confirm the correctness of the tree identification numbers. Image recognition analyzes the center coordinates of each identified tree and uses a 5-meter detection radius to determine if there are adjacent identified trees. If more than three identified trees are found within this radius, the area is considered a densely populated area. For example, if identified trees L001 are detected within a 5-meter radius... If trees L002, L003, L004, and L005 are identified as densely distributed, then L001 to L005 should be marked as trees requiring thinning. On the other hand, for cases that do not meet the standard thinning requirements, such as the distance between two trees L006 and L007 being less than the set lower limit of 1.5 meters or exceeding the upper limit of 3.5 meters, they should also be marked as abnormal spacing numbers. In actual judgment, if the distance between L006 and L007 is 1.2 meters, then the pair of trees does not meet the lower limit requirement; if the distance between L008 and L009 is 3.8 meters, then it exceeds the upper limit and should also be included in the abnormal number set. After all the above number judgments are completed, it is necessary to further remove number data that have occlusion, duplicate numbers, or abnormal distribution in the image. The output number set after completion is the set of trees requiring intervention.

[0096] The spacing offset analysis submodule calls the tree set that needs intervention, detects the planting distance between numbered trees, analyzes the degree of offset from the current tree spacing based on the standard planting spacing, and obtains the tree spacing offset data.

[0097] After calling the aforementioned tree set requiring intervention, the actual planting distance between each numbered tree is checked. This distance can be calculated using the GPS coordinates between any two trees. In practice, let the coordinates of tree A and tree B be (100.2, 23.1) and (100.3, 23.1), respectively. The distance can be calculated using the spherical coordinate distance formula to be approximately 11.1 meters. However, to suit small-scale scenarios, a simplified calculation using planar Euclidean distance can also be used. The result is approximately 11.1 meters with a 0.1-degree difference. In practice, if meters are used as the unit, the original coordinates are usually converted to meters by extracting the scale coefficient from a topographic map of equal scale before calculation. For each pair of numbered trees, the deviation between the actual planting distance and the standard planting distance is the distance offset value. For example, let the standard planting distance be 2... If the actual distance is 2.5 meters, the offset value of the numbered pair is 0.0 meters. If the actual distance is 1.7 meters, the offset is 0.3 meters. The offset values ​​of all numbered tree pairs can be uniformly processed by taking the absolute value to prevent positive and negative differences from canceling each other out. Then, they are summed and divided by the number of numbered pairs to obtain the average offset degree of the entire area. If there are 5 groups of numbered tree offset values ​​of 0.3, 0.2, 0.4, 0.1, and 0.5 meters respectively, the total offset is 1.5 meters, and the average offset is 1.5 divided by 5 equals 0.3 meters. In order to remove outliers, an abnormal offset threshold should be set. If a group of offset values ​​is higher than three times the standard deviation of the average value, such as an offset of 1.8 meters, it can be judged as data anomaly and removed from subsequent calculations. The set of all offset data obtained is the tree spacing offset data.

[0098] The planting distance intervention assessment submodule uses the following formula based on the tree spacing offset data:

[0099]

[0100] Calculate the impact value (IH) of planting spacing intervention to assess whether planting spacing needs to be replanned or thinning carried out, and generate planting spacing adjustment data. ΔH represents the average spacing offset of the g-th tree, indicating the average deviation of the actual planting spacing of the g-th tree compared to the standard planting spacing. s ρ is the standard spacing offset baseline value, used to assess whether the actual spacing offset exceeds the acceptable range; ρ is the probability of tree overlap within the numbered area, measuring the probability that trees will overlap due to excessive density within the target area; CH is the intervention cost factor for thinning operations; L... g N is the topographic correction factor for the g-th tree, used to adjust the impact of spacing offset based on the topographic features (such as slope, soil type, etc.) of the tree's location. IH It is the total number of trees;

[0101] Based on the tree spacing offset data, the actual planting offset of each tree needs to be assessed. By comprehensively considering the offset difference between the trees and the standard spacing, forest topography factors, and the risk of tree overlap in densely populated areas, the overall intervention impact value is calculated. The average spacing offset of the g-th tree is calculated from the difference between the planting distance of the numbered tree and all its neighboring trees and the standard spacing. The absolute value is then averaged. In the example, the offsets between trees L101 and L102 and L103 are 0.3 meters and 0.2 meters, respectively. Therefore, the average offset is (0.3 + 0.2) / 2 = 0.25 meters. ΔH s The standard offset reference value is set at 0.2 meters, based on the acceptable error for forestry layout. g : Terrain correction coefficient, used to reflect the weighted impact of terrain complexity on spacing offset assessment. If the g-th tree is located in a slope area or gravelly soil zone, its value can be set to 1.1; if it is located in a gentle sandy soil zone, its value is 0.9. ρ Tree overlap probability, reflects the degree of canopy overlap in the target area, which is obtained in remote sensing image analysis by the ratio of overlapping area to total canopy area. For example, if image analysis shows an overlap ratio of 25%, then ρ = 0.25. CH Intervention cost factor, set according to the average thinning cost in the operation area. If the total cost of labor and equipment is 240 yuan / time, then CH = 240. N IH =3 Total number of trees in the target area. Let the parameters corresponding to the three trees (numbered L101, L102, L103) be as follows:

[0102]

[0103] ΔH s =0.2 meters, ρ=0.25, CH=240;

[0104] Calculate the weighted average of the difference between the offset value and the baseline value for each tree:

[0105] |0.25-0.2|·1.05=0.05·1.05=0.0525;

[0106] |0.35-0.2|·1.00=0.15·1.00=0.15;

[0107] |0.15-0.2|·0.95=0.05·0.95=0.0475;

[0108] Summing and averaging:

[0109]

[0110] Calculate overlapping intervention terms:

[0111] ρ·CH=0.25·240=60;

[0112] Calculate the impact value of the intervention:

[0113] IH = 0.0833 + 60 = 60.0833;

[0114] The current planting spacing intervention impact value is 60.0833. This value represents the comprehensive adjustment degree of the current planting pattern after considering spatial offset, terrain correction and overlapping intervention factors. If this value is higher than the preset adjustment threshold (e.g., threshold 50), planting spacing re-delineation or thinning operations should be carried out.

[0115] Please see Figure 5 The climate adaptability analysis module includes:

[0116] The climate parameter extraction submodule extracts the annual average temperature, annual precipitation and corresponding monthly variation data of the current region based on the vegetation spacing adjustment data. Combined with the original climate records, it calculates the annual temperature change rate, interannual precipitation variation and climate change range, identifies and removes precipitation extremes and temperature anomalies, and establishes a regional climate parameter set.

[0117] The forestry distribution characteristics and planting distance standards of the target area were determined. For example, in a larch forest area at 125.4°E and 43.2°N, the standard distance was set at 1.8 meters. Based on the forest distribution and meteorological station layout, the annual average temperature, total annual precipitation, and monthly climate data for this area from 1970 to 2020 were extracted. The monthly average temperature and total precipitation for each of the 12 months were summarized using the annual average method, resulting in an annual average temperature of 6.2℃ and an annual precipitation of 680 mm. The above annual average temperature data were then analyzed. The variation range between adjacent years was calculated using the difference quotient method, and the temperature variation rate was obtained as 0.45℃ / year. Then, the annual precipitation was calculated in the same way, and the interannual variation of precipitation was found to be 120mm. Extreme value samples of monthly precipitation in the annual data series that were greater than the mean + 2 standard deviations or less than the mean - 2 standard deviations were identified. A total of 13 outliers were identified, which were directly removed and replaced by the nearest neighbor averaging method. Finally, the above temperature, precipitation and their variation range data were integrated and uniformly classified into the five regional climate parameters to establish a regional climate parameter set.

[0118] The tree species rating generation submodule compares the heat resistance, drought resistance, cold resistance and carbon sequestration potential of larch trees with the regional climate parameter set, selects tree species that are in the suitable range under the current climate conditions, and divides them into three levels: high, medium and low, to generate tree species adaptability rating.

[0119] Data parameters of common larch tree species are retrieved item by item, and four indicators—heat resistance, drought resistance, cold resistance, and carbon sequestration potential—are extracted sequentially. For example, tree species A has an upper limit of heat resistance of 30℃, a lower limit of cold resistance of -38℃, a drought resistance threshold of 450mm of annual precipitation, and a carbon sequestration potential of 4.2t / year. These are matched with regional climate parameters, including an average annual temperature of 6.2℃, an extreme temperature range of -31℃ to 29℃, and an annual precipitation of 680mm. If any parameter exceeds the tree species' adaptability boundary, it is marked as unsuitable. If all parameters are within the adaptability boundary, it is further classified according to the size of the carbon sequestration potential: a carbon sequestration potential value greater than 4.5t / year is high-level, between 3.0t / year and 4.5t / year is medium-level, and less than 3.0t / year is low-level. After completing the classification of each tree species, the rating labels and threshold ranges are recorded, and a structured tree species adaptability rating table is output to generate a tree species adaptability rating.

[0120] Please see Figure 6 The stand trend prediction module includes:

[0121] The tree species selection submodule analyzes the tree diameter at breast height (DBH) growth rate and crown expansion rate based on the tree species adaptability rating, compares them with climate prediction data for future time periods, selects target tree species for planting, identifies the impact on growth data, and obtains an adaptability screening score.

[0122] Based on historical meteorological data and regional climate simulation data, the local average annual temperature is set between 16℃ and 22℃, and the precipitation is set between 800mm and 1200mm as an adaptive climate reference range. Then, data on the diameter at breast height (DBH) growth rate and crown expansion rate of major tree species in the target area, such as Chinese fir, Masson pine, and oak, are obtained. For example, the DBH of a Masson pine tree measured in 2022 and 2023 were 12.4cm and 13.2cm respectively, so the growth rate is (13.2-12.4) / 1 = 0.8cm / year, and the crown width is 3.2m and 3.6m respectively, corresponding to an expansion rate of (3.6-3.2) = 0.4m / year. Simultaneously, the predicted average local temperature for the next five years (2025–2029) is retrieved as 19.5℃, and the predicted average annual precipitation is 1050mm. The optimal growth conditions for Chinese fir are compared to 18–20℃ and 1000–1000mm. The tree species was judged to have a high degree of climate adaptability based on a diameter at breast height (DBH) growth of 200 mm. Next, the deviation between the DBH growth and crown width expansion data and the climate parameters was analyzed. The absolute value of the deviation for each indicator was calculated, and a scoring penalty was set based on the deviation. For example, if the standard DBH growth rate was 0.9 cm / year and the actual measurement was 0.8 cm / year, the deviation was 0.1 cm, and 2 points were deducted for every 0.1 cm, for a total deduction of 2 points. Similarly, if the standard crown width expansion rate was 0.5 m / year and the actual measurement was 0.4 m / year, the deviation was 0.1 m, and 2 points were also deducted. Regarding climate deviations, a temperature deviation of 0.5℃ and a precipitation deviation of 50 mm resulted in deductions of 3 and 2 points respectively. The total score was 100 - (2 + 2 + 3 + 2) = 91 points. Under the preset adaptability screening score threshold of 85 points, this tree species scored 91 points, exceeding the threshold, and was therefore included in the target tree species range, resulting in an adaptability screening score of 91 points.

[0123] The structural trend assessment submodule evaluates the potential trends in stand density and structure based on the adaptability screening score, using the following formula:

[0124]

[0125] The stand change assessment results TB were obtained, of which DB y VB represents the average density of tree species of the yth class, indicating the average number of trees of a specific species per unit area within a given region. y It is the growth potential coefficient of tree species in category y, reflecting the maximum growth potential of this tree species under ideal conditions, GB y PB represents the current diameter at breast height (DBH) growth rate of tree species in category y, describing the annual rate of DBH growth for this tree species under current environmental conditions. y RB represents the expected growth rate of diameter at breast height (DBH) for tree species of type y under specific climatic conditions. RB is the annual variation in growth rate of the target tree species, measuring the variability and instability of the growth rate of the target tree species between different years. uAB represents the age structure proportion of the u-th forest layer, indicating the proportion of the u-th age layer in the forest stand. u It is the average age of the trees at the u-th level, n tb It represents the total number of tree species, m tb It refers to the number of strata in the age structure of the forest stand;

[0126] Extract the parameters for each target tree species. Assume there are three categories of selected tree species, numbered y = 1, 2, 3, with the following parameters:

[0127] Tree species 1: DB1 = 320 trees / ha, VB1 = 1.2, GB1 = 0.85 cm / year, PB1 = 0.80 cm / year;

[0128] Tree species 2: DB2 = 280 trees / ha, VB2 = 1.1, GB2 = 0.75 cm / year, PB2 = 0.72 cm / year;

[0129] Tree species 3: DB3 = 300 trees / ha, VB3 = 1.3, GB3 = 0.90 cm / year, PB3 = 0.88 cm / year;

[0130] Calculate the terms of the numerator:

[0131] The deviation item for tree species 1 is (GB1-PB1). 2 =(0.85-0.80) 2 =0.0025, substituting, we get 320·1.2·0.0025 = 0.96;

[0132] The deviation item for tree species 2 is (GB2-PB2). 2 =(0.75-0.72) 2 =0.0009, substituting, we get 280·1.1·0.0009 = 0.2772;

[0133] The deviation item for tree species 3 is (GB3-PB3). 2 =(0.90-0.88) 2 =0.0004, substituting gives 300·1.3·0.0004 = 0.156;

[0134] Summing the above three terms, we obtain the total value in the numerator:

[0135] 0.96 + 0.2772 + 0.156 = 1.3932;

[0136] Then calculate the parameters in the denominator. If the age structure has two levels, numbered u = 1, 2, its parameters are as follows:

[0137] Age group 1: MB1 = 0.45, AB1 = 12 years, product is 0.45·12 = 5.4;

[0138] Age group 2: MB2 = 0.55, AB2 = 18 years, the product is 0.55·18 = 9.9;

[0139] The sum is:

[0140]

[0141] Assuming the annual variation growth rate fluctuates by RB = 0.15, its square root is:

[0142]

[0143] Add it to the calculated value above:

[0144] Denominator = 15.3 + 0.387 = 15.687;

[0145] Substitute the numerator and denominator into the formula:

[0146]

[0147] The result indicates that the current forest stand structure change trend is in a low-fluctuation state. The value TB = 0.0888 can be used as the basis for subsequent structure optimization threshold comparison. According to the set benchmark value range, if TB < 0.1, it means that the forest stand structure change is stable, and the current structure change trend belongs to the stable range.

[0148] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A decision-making system for larch forest cultivation based on multi-objective optimization, characterized in that, The system includes: The spatial layout assessment module uses data on larch tree height, diameter at breast height (DBH), and crown size to measure the physical distance between adjacent trees, calculate the crown overlap area, assess the spatial distribution of trees, determine the number of neighboring trees, and obtain the neighboring tree density index. The thinning condition determination module identifies the neighboring trees of each tree based on the neighboring tree density index, collects data on the canopy interlacing area and growth rate, compares it with the benchmarks of canopy interlacing and growth rate, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier. Based on the thinning requirement identifier, the planting spacing adjustment assessment module determines the tree number that needs intervention, measures the actual distance between trees, selects the standard planting spacing as a reference, analyzes the degree of deviation from the current spacing, assesses whether the planting spacing needs to be replanned or thinning is carried out, and generates planting spacing adjustment data. Based on the planting distance adjustment data, the climate adaptability analysis module obtains the climate conditions required for larch forest cultivation, compares the data with the tree species' heat resistance, drought resistance, cold resistance and carbon sequestration potential, identifies tree species suitable for larch forest cultivation under the current climate conditions, and generates a tree species adaptability rating. The stand trend prediction module selects target tree species for planting based on the tree species adaptability rating, collects data on the tree diameter at breast height (DBH) growth rate and crown expansion rate, and combines this data with climate prediction data for future periods to conduct simulation analysis, assess the potential change trends of stand density and structure, and obtain stand change assessment results. The stand change assessment results include information on growth dynamics and changes in stand structure. The forest stand trend prediction module includes: The tree species selection submodule analyzes the tree diameter at breast height (DBH) growth rate and crown expansion rate based on the tree species adaptability rating, compares them with climate prediction data for future time periods, selects target tree species for planting, identifies the impact on growth data, and obtains an adaptability screening score. The structural trend assessment submodule evaluates the potential trends in stand density and structure based on the aforementioned adaptability screening score, using the following formula: ; The stand change assessment result TB was obtained, among which, It is the average density of the y-th type of tree species. It is the growth potential coefficient of the y-th tree species. This represents the current growth rate of diameter at breast height (DBH) for tree species of type y. is the expected growth rate of diameter at breast height (DBH) for the y-th tree species corresponding to climate, and RB is the fluctuation range of the annual variation growth rate of the target tree species. It is the age structure ratio of the u-th forest layer. It is the average age of the trees in the u-th layer. It is the total number of tree species. It refers to the number of age structure layers in the forest stand.

2. The larch forest cultivation decision-making system based on multi-objective optimization according to claim 1, characterized in that, The adjacent forest density indicators include forest distance, coverage, and density; the thinning demand indicators include growth competition intensity, disturbance level, and thinning priority; the planting distance adjustment data includes distance deviation, planting distance compliance, and planting distance adjustment demand; and the tree species adaptability rating includes climate adaptability, survival potential, and growth potential.

3. The larch forest cultivation decision-making system based on multi-objective optimization according to claim 1, characterized in that, The spatial layout evaluation module includes: The canopy overlap calculation submodule, based on larch tree height, diameter at breast height (DBH), and canopy size data, obtains the center coordinates of each tree according to the spatial location data of the pine forest, using the following formula: ; Calculate the crown overlap area between each pair of adjacent trees. in, It is the crown radius of the i-th tree, referring to the distance from the center of the trunk to the edge of the crown. It is the radius of the crown of the j-th tree. It is the straight-line distance between the center points of the i-th tree and the j-th tree. It is pi; The diameter at breast height (DBH) difference comparison submodule collects corresponding DBH data based on the crown overlap area and the number of each larch tree, evaluates the DBH difference between adjacent trees, and filters out trees that fall within the DBH difference benchmark range to obtain the DBH matching degree. The neighboring tree density determination submodule assesses the spatial distribution of trees based on the diameter at breast height matching degree, counts the average number of neighboring trees in each area, and obtains the neighboring tree density index.

4. The larch forest cultivation decision-making system based on multi-objective optimization according to claim 1, characterized in that, The thinning condition determination module includes: The density over-standard identification submodule obtains the coordinate data of each tree and its surrounding trees based on the neighboring tree density index, calculates the tree density around each target tree, compares it with the density benchmark, identifies the area of ​​trees with excessive density, and obtains the cross-identification result. The growth rate assessment submodule calls the interleaving recognition results, collects the annual change in tree diameter at breast height, compares it with the average growth data of the same tree species in the region, identifies whether the individual growth rate is lower than the average level, determines whether the tree meets the thinning conditions, and generates a thinning requirement identifier.

5. The larch forest cultivation decision system based on multi-objective optimization according to claim 1, characterized in that, The planting spacing adjustment assessment module includes: The tree numbering and identification submodule collects the numbering information of the identified trees in the forest land based on the thinning requirement identifier. By comparing the numbering label data with the tree distribution image, it determines the tree numbers of densely distributed or non-standard thinning areas and generates a set of trees that need intervention. The spacing offset analysis submodule calls the set of trees that need intervention, detects the planting distance between numbered trees, analyzes the degree of offset from the current tree spacing based on the standard planting spacing, and obtains the tree spacing offset data. The planting distance intervention assessment submodule uses the following formula based on the tree spacing offset data: ; Calculate the impact value of planting distance intervention The assessment determines whether planting spacing needs to be replanned or thinning carried out, generating planting spacing adjustment data. It is the average spacing offset value of the g-th tree. It is the standard spacing offset reference value. It represents the probability of overlapping trees within the numbered area. It is the intervention cost factor of thinning operations. It is the topographic correction factor for the g-th forest. It represents the total number of trees.

6. The larch forest cultivation decision system based on multi-objective optimization according to claim 1, characterized in that, The climate adaptability analysis module includes: Based on the vegetation spacing adjustment data, the climate parameter extraction submodule extracts the annual average temperature, annual precipitation and corresponding monthly variation data of the current region. Combined with the original climate records, it calculates the annual temperature change rate, interannual precipitation variation and climate change range, identifies and removes precipitation extremes and temperature anomalies, and establishes a regional climate parameter set. The tree species rating generation submodule compares the heat resistance, drought resistance, cold resistance and carbon sequestration potential of larch trees with the set of regional climate parameters, selects tree species that are in the suitable range under the current climate conditions, and classifies them into three levels: high, medium and low, to generate a tree species adaptability rating.

7. The larch forest cultivation decision system based on multi-objective optimization according to claim 2, characterized in that, The spatial layout evaluation module executes multi-objective collaborative optimization logic, using a linear combination of the canopy overlap area and the density of adjacent trees as a constraint condition; The spatial competition index is obtained by multiplying the density and coverage, and the coverage threshold is set to 1.2 times the canopy radius. The distance between the tree crown radius and the center point is calculated using normalized units in meters.

8. The larch forest cultivation decision system based on multi-objective optimization according to claim 1, characterized in that, The adjacent tree density index is calculated using the Hegyi competition index to assess the intensity of competition between trees. The aforementioned canopy interlacing benchmark refers to the critical value of canopy closure, with an interference limit of 70%-80%. Exceeding this limit indicates that the trees are too dense and thinning should be considered. The standard planting spacing is 2Mx2M or 2Mx3M.

Images