Method for optimizing spatial distribution of garden plants

Abstract

The invention discloses an optimization method for spatial distribution of garden plants, which comprises the following steps: generating a high-precision 3D model of a garden terrain by adopting a plant spatial adaptive layout algorithm and utilizing a laser radar or unmanned aerial vehicle technology, generating a weight matrix according to soil types, illumination intensity and drainage characteristics, and applying a dynamic weight Voronoi algorithm in the terrain model to optimize the spatial distribution of the garden plants. The method comprises the following steps: generating a planting range of each plant, establishing a root distribution model by using an underground imaging technology, simulating distribution dynamic of moisture and nutrients among root systems in combination with underground hydrological data, collecting sensor data in real time by using a wireless network, designing a dynamic irrigation scheme, and optimizing the planting density and irrigation range of the plants. Data such as terrain and climate are integrated into a virtual garden, the virtual garden is ensured to be consistent with an actual state through interaction of a sensor and a digital twinborn model, plant growth, ecological benefit changes and disaster risks are predicted by using the digital model, and a user is allowed to modify and optimize the garden layout in real time through a visual interface.

Description

Technical Field

[0001] The invention relates to the field of ecology, and in particular to a method for optimizing the spatial distribution of garden plants. Background Art

[0002] For comparison, the aquatic plant optimization adjustment method represented by the Chinese invention patent CN118734605B mainly optimizes the distribution of aquatic plants, and makes adjustments through spatial distribution data extraction, survival characteristics analysis, and abnormal trend coupling analysis. Although certain achievements have been made in the field of aquatic community plant optimization, when the method is applied to the optimization of spatial distribution of garden plants, there are the following significant deficiencies:

[0003] First, the technical methods in the public documents are mainly applicable to the ecological optimization of aquatic plants, and lack compatibility with diverse vegetation in garden environments. There are many types of garden plants, including trees, shrubs and herbaceous plants, and different types of plants vary greatly in terms of light, soil fertility, water requirements, etc. However, the invention mainly adjusts the spatial distribution based on the characteristics of aquatic plants, and fails to take into account the diversity needs of terrestrial plants. In addition, there are obvious terrain undulations in the garden environment, such as highlands, slopes and low-lying areas, while the aquatic environment is usually relatively flat, which makes it difficult for the method to effectively adapt to complex terrain conditions. Secondly, the method of the public document focuses on the coupling analysis of abnormal trends of aquatic plants, but lacks in-depth optimization for underground resource competition and plant root distribution. The underground root system of garden plants is the core of resource competition, and the utilization of underground water and nutrients directly affects the growth and ecological balance of plants. However, the patent method does not involve a root competition model, nor does it consider the root coordination mechanism between plants. This optimization method that ignores root factors leads to uneven resource distribution, especially in garden environments with high plant density or limited soil resources. The spatial distribution optimization of this method is limited to static data analysis and lacks consideration of the dynamic growth process of plants. The growth of garden plants is a dynamic process, and its ecological adaptability will be continuously adjusted with time and environmental changes. The optimization process of the public document is based on static distribution data and characteristic analysis, without introducing time variables and establishing a dynamic prediction model. This static optimization method is difficult to adapt to changes in environmental conditions during plant growth, such as seasonal changes in light intensity, fluctuations in soil moisture, and changes in space requirements caused by plant growth. In summary, the public document has achieved certain results in the optimization of aquatic plants, but there are significant deficiencies in the wide application of spatial distribution optimization of garden plants. These problems include lack of adaptability to diversified vegetation, ignoring underground resource competition and dynamic growth processes, failure to introduce digital twin technology, lack of multi-objective optimization capabilities, poor interactivity, and lack of real-time data support. To solve these problems, the present invention technically integrates dynamic modeling, digital twin system, highly interactive visualization platform, and multi-objective optimization algorithm to provide a more scientific, efficient, and intelligent solution for the spatial distribution of garden plants, thereby significantly improving the accuracy and sustainability of garden management. Summary of the invention

[0004] The purpose of the present invention is to provide a method for optimizing the spatial distribution of garden plants, thereby solving some of the drawbacks and deficiencies pointed out in the background technology. The invention solves the above-mentioned technical problems by adopting the following technical solutions:

[0005] Methods for optimizing the spatial distribution of garden plants include:

[0006] S1, plant space adaptive layout algorithm;

[0007] S1.1, Terrain modeling: Use LiDAR or UAV technology to generate a high-precision 3D model of the garden terrain;

[0008] S1.2, Dynamic weight calculation: Generate a weight matrix based on soil type, light intensity and drainage characteristics;

[0009] S1.3, variable Voronoi partitioning: Apply the dynamic weighted Voronoi algorithm in the terrain model to generate the planting range of each plant;

[0010] S2, spatial coordination model of roots among plants;

[0011] S2.1. Root system simulation modeling: Use underground imaging technology (such as electrical resistance imaging or geological radar) to establish a root distribution model;

[0012] S2.2, Water and nutrient flow analysis: Combined with groundwater hydrological data, simulate the dynamic distribution of water and nutrients among the root system;

[0013] S3, intelligent irrigation and monitoring system embedding;

[0014] S3.1, Data collection and transmission: Use wireless networks to collect sensor data in real time;

[0015] S3.2, Irrigation optimization: Design dynamic irrigation plans based on plant water requirements and real-time data;

[0016] S3.3, Feedback and correction: Optimize plant planting density and irrigation range through data analysis;

[0017] S4, digital twin garden system;

[0018] S4.1. Digital twin model generation: integrating terrain, climate, plant growth and other data into the virtual garden;

[0019] S4.2, Real-time data synchronization: through the interaction between sensors and digital twin models, ensure that the virtual garden is consistent with the actual state;

[0020] S4.3, Simulation and prediction: Use digital models to predict plant growth, changes in ecological benefits and disaster risks;

[0021] S4.4, User Interaction: Through the visual interface, users are allowed to modify and optimize the garden layout in real time;

[0022] Furthermore, the plant space adaptive layout algorithm includes dynamic weighted environmental factor modeling, which expresses the influence of different ecological factors on plant distribution as a nonlinear function. Unlike traditional weighting methods, this model can dynamically respond to changes in factor values ​​in space and generate appropriate weight distribution. The weight function is defined as:

[0023]

[0024] Among them, W(x,y) represents the comprehensive weight of the position (x,y), which is used to reflect the ecological suitability of the area; L(x,y) represents the light intensity at the position (x,y), which is an important factor affecting plant growth; S(x,y) represents the soil fertility at the position (x,y), which is related to the absorption capacity of the plant root system; S0 represents the target soil fertility value, which is used to evaluate the difference between the actual fertility and the target value; W(x,y) also represents the soil moisture content at the position, and its periodic changes are modeled by the sin function; γ1, γ2, and γ3 are the influence weights of light, soil fertility, and moisture, respectively, which are used to adjust the relative importance of each factor; α1 is the nonlinear influence range of controlling light intensity; β1 is the fluctuation effect of regulating the periodic changes of moisture on the weight.

[0025] Furthermore, the plant space adaptive layout algorithm includes the construction of a variable Voronoi diagram, which adjusts the influence range of each planting point through dynamic weights to make it highly matched with the environmental suitability. The boundary of the Voronoi unit is determined by the following improved formula:

[0026]

[0027] where d i Represents the weighted distance from the position (x, y) to the i-th planting point; (x i ,y i ) represents the coordinate position of the ith planting point; (x, y) represents the spatial position of the current assessment; W(x, y) is the dynamic weight, which represents the impact of environmental suitability on the current position; the denominator 1+W(x, y) is used to adjust the Euclidean distance so that the distance of high-weight (strong environmental suitability) areas is shortened and the distance of low-weight areas is lengthened.

[0028] Furthermore, the plant space adaptive layout algorithm includes multi-objective optimization and layout iteration, and a multi-objective optimization function is designed to adjust the distribution position of the planting points through an adaptive iterative algorithm to ensure that both the ecology and aesthetics of the plant distribution are taken into account;

[0029] The optimization objective function is defined as:

[0030]

[0031] Where F represents the overall optimization target value, which measures the comprehensive benefit of the entire layout; φ1: represents the ecological benefit contribution of plant growth, such as carbon absorption, soil conservation, etc.; W(x, y) is the dynamic weight, which represents the ecological suitability of the position (x, y); R(x, y) is the resource competition function, which is used to quantify the degree of competition for resources at this location among neighboring plants; λ1 is the nonlinear contribution of adjusting the dynamic weight to the ecological benefit; φ2 is used to balance the negative impact of resource competition on the overall layout; Ω represents the range of the planting area, which is used to integrate and calculate the comprehensive benefit of the entire space.

[0032] Furthermore, the establishment of the spatial coordination model of plant roots includes obtaining underground root spatial distribution data through underground imaging technology, combining dynamic prediction models and resource optimization algorithms, and controlling the efficient coordination between plant roots and the optimal utilization of underground resources.

[0033] The root density distribution R(x,y,z) is modeled as follows:

[0034]

[0035] Where R(x,y,z) represents the root density at the position (x,y,z), in units of roots / cubic meter, and is used to describe the spatial distribution of underground roots; R0 represents the maximum root density, which depends on the plant species and the growth characteristics of the root system; z is the underground depth, in meters, and describes the distribution of roots at different depths; κ1 represents the depth influence factor, which controls the attenuation rate of the root system with increasing depth; the larger the value, the shallower the root distribution in the depth direction; T(x,y) is the soil temperature distribution function, in degrees Celsius, and is used to describe the soil temperature at the position (x,y); T0 is the optimum temperature for plant root growth, in degrees Celsius; κ2 is the temperature difference adjustment coefficient, which reflects the sensitivity of temperature changes to root distribution; the larger the value, the more significant the effect of temperature differences on root distribution; exp(-k2·|T(x,y)-T0|) is used to simulate the inhibitory effect of temperature differences on root distribution.

[0036] Furthermore, the establishment of the spatial coordination model of the root systems among plants includes a dynamic root growth prediction model:

[0037] R t (x,y,z,t)=R(x,y,z)·(1-e -μ·t )·(1+δ·sin(ω·t+φ))

[0038] Among them, R t(x,y,z,t) is the root density at position (x,y,z) at time y, in units of roots / cubic meter; R(x,y,z)): is the basic value of the static root density model, calculated by the above formula; μ is the root growth rate factor, which reflects the root expansion speed in time; the larger the value, the faster the root expansion; t is time, in years, describing the growth stage of the plant; δ is the amplitude factor of root density fluctuation, which describes the regulatory effect of seasonal changes on root density; the larger the value, the more significant the fluctuation range; ω is the seasonal change frequency factor, in rad / year, reflecting the periodicity of root fluctuations; φ represents the phase offset, which is used to adjust the starting time of seasonal changes; (1-e -μ·t ) is used to simulate the cumulative effect of root growth over time; and (1+δ·sin(ω·t+φ)) simulates the cyclical effect of seasonal fluctuations on root density.

[0039] Furthermore, the spatial coordination model of the root systems among plants includes an optimization model for underground resource competition, which controls the optimization of the layout and density of planting sites by minimizing resource competition. The objective function of resource competition optimization is defined as:

[0040]

[0041] Among them, F is the objective function, which represents the comprehensive intensity of underground resource competition, and the optimization goal is to minimize it; C(x, y, z) is the intensity of root resource competition, which is defined as the superposition value of root densities of different plants:

[0042]

[0043] R i (x,y,z),R j (x, y, z) represents the root density of the i-th plant and the j-th plant respectively; D(x, y, z) is the spatial distance between roots, describing the distance between different plants at position (x, y, z); λ is the distance adjustment coefficient, which controls the degree of distance mitigation of resource competition; the larger the value, the stronger the mitigation effect of distance on competition; Ω is the spatial range of the planting area; d V It is a spatial volume element, which is used to integrate and calculate the resource competition intensity of the entire planting area.

[0044] Furthermore, the digital twin system includes building a highly dynamic and intelligent garden management platform by integrating multi-source data including terrain, climate, plant growth and other information and real-time Internet of Things technology; the system uses virtual reality VR and augmented reality AR technology to ensure that garden design and management plans can be synchronized in real time in virtual and real environments.

[0045] Furthermore, the digital twin system includes virtual and reality synchronous modeling: by combining three-dimensional modeling with real-time data updating, a virtual garden model V (x, y, z, t) is established, and the model can be dynamically adjusted according to changes in the actual environment.

[0046] Virtual synchronization technology: Through the virtual engine and physical simulation technology, the actual changes of the garden can be fed back to the virtual model in real time;

[0047] V(x,y,z,t)=R(x,y,z,t)+Δ(t)

[0048] Among them, V(x,y,z,t) represents the spatial coordinates (x,y,z) and the state of the virtual garden model at time t; R(x,y,z,t) represents the preliminary garden model generated based on the initial terrain and climate data, the spatial distribution and attributes of the model; Δ(t) represents the deviation corrected by real-time data.

[0049] Furthermore, the digital twin system includes a highly interactive visualization platform: through virtual reality VR and augmented reality AR technology, a highly interactive platform is provided, where users can view garden design plans in real time and optimize the plans in virtual scenes; users can adjust planting plans, irrigation systems, etc. in real time to view the ecological impact of different plans;

[0050] Interactive platform design: Garden managers can enter the virtual garden model through virtual reality equipment and make intuitive adjustments to plant distribution, ecological layout, irrigation system, etc. The system makes the best choice through real-time simulation and feedback.

[0051] F interact =∫ Ω [C(G(t),E(t))-L(D)]d A

[0052] Among them, F interact is the interactive optimization objective function, which indicates the target value of the optimization scheme after user interaction; Ω is the spatial scope of garden design, which indicates the area affected by all design parameters; C(G(t),E(t)) is the value of garden ecological service, which depends on the plant growth function G(t) and the environmental data function E(t); L(D) is the resource consumption function; d A It is the area element of each space unit, used to optimize the garden space.

[0053] The following are the beneficial effects of the present invention:

[0054] Through dynamic weight modeling, comprehensive consideration of environmental factors such as light, soil fertility, and water, a suitable plant distribution layout is generated to ensure that plants grow in the optimal growth environment, improving plant health and growth efficiency. Through the root space coordination model and resource competition optimization algorithm, the resource competition between plant roots is effectively alleviated, and the utilization efficiency of underground water and nutrients is improved, thereby reducing resource waste and reducing plant mortality. The multi-objective optimization algorithm is combined with a dynamic growth prediction model, while maximizing ecological benefits (such as carbon absorption, soil conservation, etc.), it significantly reduces irrigation and maintenance costs and promotes the ecological sustainable development of gardens. Based on the digital twin system, the present invention can dynamically reflect the actual state of the garden through real-time data updates. With the help of virtual reality (VR) and augmented reality (AR) technology, users can adjust the planting layout, irrigation system, and resource allocation in real time to ensure that garden management and design solutions are always in the best state. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a flow chart of the method for optimizing the spatial distribution of garden plants.

[0056] Figure 2 This is a flow chart of the plant space adaptive layout algorithm of the present invention.

[0057] Figure 3 A flow chart is established for the spatial coordination model of root systems among plants of the present invention.

[0058] Figure 4 This is a flow chart of the digital twin system of the present invention. DETAILED DESCRIPTION

[0059] The specific implementation modes of the present invention will be described in detail below in conjunction with the accompanying drawings.

[0060] Combined with Figure 1, an optimization method for the spatial distribution of garden plants, the method includes multiple steps: first, the plant space adaptive layout algorithm (S1) is the core algorithm, which uses a dynamic calculation model to optimize the layout of plants in real time based on the garden terrain, environmental factors and plant growth requirements. The algorithm can dynamically adjust and optimize the position of plants according to the growth status, spatial distribution and environmental characteristics of plants in the garden, thereby improving the ecological benefits and resource utilization of the garden. Specifically, the first step of the plant space adaptive layout is terrain modeling (S1.1), which uses laser radar technology (LiDAR) or sensors carried by drones to accurately scan the garden terrain and generate a high-precision three-dimensional terrain model. LiDAR technology can accurately obtain the height information of ground objects, ground contours and vegetation coverage by irradiating the ground with laser beams and reflecting them back to the sensor. These data, combined with high-resolution images and videos taken by drones, can construct a high-precision three-dimensional garden model that accurately expresses the terrain characteristics of the garden, such as undulations, slopes, and gullies. Next, dynamic weight calculation (S1.2) is a key step in optimizing plant distribution. In this process, the system assigns dynamic weight values ​​to each geographical area based on environmental factors such as soil type, light intensity, and drainage characteristics in the garden. Soil type determines the nutrient and water carrying capacity of different areas, light intensity affects the photosynthesis requirements of plants, and drainage characteristics determine the moisture content of the soil and the growth conditions of plant roots. Finally, variable Voronoi partitioning (S1.3) is applied to the garden terrain model to generate the planting range of each plant. The Voronoi diagram is a commonly used spatial partitioning method that can divide the space into several areas based on a given seed point, and the distance from each point in each area to the seed point is closer than the distance to other seed points. In the layout of garden plants, the traditional Voronoi diagram uses a fixed weight distribution, but this method does not take into account the dynamic resource distribution required for plant growth, so there is a resource competition problem in practical applications. To solve this problem, the variable Voronoi partitioning method introduces dynamic weights and adjusts the Voronoi diagram according to the weight value of each area to form a flexible spatial partitioning structure that adapts to environmental changes. In this improved Voronoi diagram, the planting range of each plant is not only affected by the distance, but also adjusted according to the environmental suitability of the area.

[0061] The second key step is the spatial coordination model of the roots among plants (S2). The root system is the main way for plants to obtain water and nutrients, and the distribution of the root system and its relationship directly affect the growth and health of the plant. Therefore, in the layout of garden plants, reasonable spatial coordination of the root system is crucial. Root simulation modeling (S2.1) is the first step in the model. It relies on underground imaging technology (such as electrical resistance imaging or geological radar) to accurately obtain the distribution of underground roots. Electrical resistance imaging technology can reflect the existence and distribution of roots by measuring the changes in the conductivity of soil or underground water. Geological radar can accurately depict the underground root structure by emitting high-frequency electromagnetic waves and detecting the reflected signals of underground objects. These technologies can obtain detailed information about the depth, extension range and root density of plant roots, and then generate an accurate three-dimensional root distribution model. Using these root distribution data, the model can more accurately predict the demand of plant roots for underground water and nutrients, and further optimize the spatial layout of garden plants on this basis. Next, water and nutrient flow analysis (S2.2) is the second step of the model, which combines groundwater data to simulate the dynamic distribution of water and nutrients among roots. By combining groundwater data (such as soil permeability, water content, nutrient concentration, etc.) with the root distribution model, the system can analyze the distribution of water and nutrients among different plant roots in real time, simulate the flow process of water and nutrients, and further evaluate the resource competition and interaction among plants. Through this analysis, the system can identify areas with uneven water and nutrient flow, predict competition and resource scarcity among plant roots, and dynamically adjust the layout of plants to reduce root overlap between different plants and avoid uneven plant growth or death due to competition for resources.

[0062] The third key step is the embedding of the intelligent irrigation and monitoring system (S3). This part implements more precise irrigation management and plant monitoring by combining sensor technology, wireless communication networks, data analysis algorithms and control systems, thereby achieving efficient use of water resources and optimizing plant growth. Specifically, data acquisition and transmission (S3.1) is the basis of the entire system. It uses wireless networks and sensors to collect environmental data and plant growth conditions in the garden in real time. For example, sensors can measure environmental variables such as soil moisture, temperature, light, and meteorological data in real time, and transmit these data to the central data processing platform through wireless communication networks (such as Wi-Fi, LoRa, etc.). Sensors can not only capture soil moisture and temperature changes at different locations in the garden, but also monitor the growth status of plants and real-time changes in their water requirements, thereby providing accurate data support for subsequent irrigation decisions. With the distribution of sensors and the real-time nature of data transmission, the ecological status of the entire garden can be fully monitored and analyzed through the data platform, ensuring that the system can always respond in a timely manner according to plant needs. Irrigation optimization (S3.2) is to design a dynamic irrigation plan that adapts to the needs of garden plants by analyzing the water requirements of plants and real-time environmental data. The water requirement of plants is affected by many factors, including plant species, root depth, soil type, climate conditions, and soil moisture. Through the real-time data collected, the system can adjust the irrigation amount and irrigation time according to the weather and soil conditions in different time periods. Feedback and correction (S3.3) is an indispensable part of the optimization process. The system optimizes the plant planting density and irrigation range through continuous data analysis. By analyzing the sensor data, the system can identify uneven water distribution in the garden, areas with excess or insufficient soil moisture, and uneven plant growth. Based on this feedback information, the system can intelligently adjust the plant planting density or irrigation area to optimize the allocation of resources.

[0063] The fourth key step is the digital twin garden system (S4). Digital twin technology is a system that reflects the real world in real time by creating a virtual model. It integrates multi-dimensional data such as geography, climate, and plant growth of the actual garden to form a digital virtual model synchronized with the actual garden, thereby providing accurate decision-making support for garden management and optimization. Specifically, digital twin model generation (S4.1) is the first step of the system. It generates a highly accurate digital twin model by integrating data such as the terrain, climate conditions, and plant growth status of the actual garden into the virtual garden. Terrain data includes information such as the three-dimensional structure, slope, and soil type of the garden. Climate data covers environmental factors such as temperature, precipitation, humidity, and light. Plant growth data includes the growth status, water demand, and root distribution of different plants. These data can be collected and input into the virtual garden through sensors, remote sensing technology, drones, and historical climate data, thereby establishing a highly accurate virtual model, so that the digital twin garden can fully reflect the environmental and ecological characteristics of the actual garden. Next, real-time data synchronization (S4.2) is the key link to ensure that the virtual garden is consistent with the actual state. By installing various sensors (such as temperature and humidity sensors, soil moisture sensors, meteorological sensors, etc.) in the garden and transmitting the real-time data collected by the sensors to the digital twin system through wireless networks, the state of the virtual garden can be dynamically updated as the real environment changes. In this way, the virtual garden can not only display the current state of the garden, but also adjust its performance according to real-time data to provide more accurate garden management support. Simulation and prediction (S4.3) is another important function of the digital twin garden system. It uses virtual models to simulate and predict plant growth, changes in ecological benefits, and disaster risks. By simulating factors such as different plant growth cycles, climate conditions, and soil types, the system can predict future plant growth and its impact on the environment, such as carbon sink effects, cooling effects, and other ecological benefits. At the same time, the system can also simulate various potential ecological risks, such as drought, floods, pests and diseases, and propose corresponding prevention or optimization measures based on the simulation results. For example, if the system predicts that there will be excess water in a certain area in the future, resulting in plant root rot, the system can adjust the irrigation plan of the area in advance or replan the plant layout to reduce ecological risks. Finally, user interaction (S4.4) is another important function of the digital twin garden system. Through the visual interface, users can view information such as plant growth, climate change, and resource allocation in the garden in real time, and can modify and optimize the garden layout according to actual needs. Through virtual reality (VR) or augmented reality (AR) technology, users can not only observe the current status of the garden intuitively, but also adjust the position, quantity, type and other parameters of the plants through simple operations. The system will provide real-time feedback on the modified effect, allowing users to test and optimize the garden layout in a virtual environment.

[0064] Embodiment 1:

[0065] Combined with Figure 2 , set to design the distribution of garden plants in the city park, the goal is to optimize the planting layout of plants according to environmental factors such as light, soil fertility and water in different areas, to ensure that each plant grows under the best environmental conditions. First, it is necessary to collect environmental data from different areas of the park, including light intensity (L(x,y)), soil fertility (S(x,y)) and water content (W(x,y)). The data collected by sensors and remote sensing technology are as follows:

[0066] The light intensity (L(x,y)) varies greatly in different areas of the garden. The strongest light appears in the open area in the center of the park, which is about 1000 lux, while the light intensity in the shaded area drops to about 200 lux.

[0067] Soil fertility (S(x,y)) also varies greatly in different areas. The soil fertility value on the east side of the park is 8 (the standard soil fertility value is 10), while the poorer area on the west side is only 4.

[0068] The distribution of soil moisture (W(x,y)) varies periodically, with low-lying areas with poor drainage on the west side having higher moisture content of about 60%, while the central area has lower moisture content of about 30%.

[0069] Based on the above data, the comprehensive weight (W(x,y)) of each location can be calculated by the given dynamic weight environmental factor modeling formula. The formula is as follows:

[0070]

[0071] Among them, W(x,y) is the comprehensive weight of the position (x,y); L(x,y) is the light intensity at the position (x,y), in lux; S(x,y) is the soil fertility at the position (x,y); S0 is the target soil fertility value; W(x,y) is also the soil moisture content at the position, in percentage (%); γ1, γ2, γ3 are the weight coefficients of light, soil fertility and moisture; α1 is the nonlinear influence range for controlling light intensity, usually set to a value between 0 and 1; β1 is the coefficient for regulating the influence of periodic changes in moisture on weight fluctuations, usually a value less than 1.

[0072] Set and select the following specific values ​​for calculation:

[0073] (γ1=0.4,γ2=0.3,γ3=0.3), which means that the influence of light, soil fertility and water on the weights in the calculation is relatively balanced.

[0074] α1=0.5, controls the nonlinear range of light intensity, moderate.

[0075] β1=0.8, adjusts the effect of moisture on weight fluctuation.

[0076] Next, the comprehensive weight of a specific area (such as the central area of ​​the park) can be calculated. In this area, the light intensity L(x,y) = 1000 lux , soil fertility S(x,y)=8, moisture content W(x,y)=30%), target soil fertility S0=10.

[0077] 1. Calculation of illumination:

[0078]

[0079] 2. Calculation of soil fertility:

[0080] |S(x,y)-S_0|=|8-10|=2

[0081] 3. Calculation of moisture content:

[0082] sin(β1W(x,y))=sin(0.8×30)=sin(24)≈0.4067

[0083] Substituting these values ​​into the composite weight formula:

[0084] W(x,y)=0.4×1+0.3×2+0.3×0.4067=0.4+0.6+0.122=1.122

[0085] Therefore, the comprehensive weight of the central area of ​​the park is 1.122, indicating that the ecological suitability of this area is high and it is suitable for growing plants with higher light requirements.

[0086] Next, assume that the soil fertility on the east side of the park is lower, the soil fertility is S(x,y)=4, and the light intensity and water conditions are similar to those in the central area. Calculate the comprehensive weight of the east side area.

[0087] 1. Calculation of illumination:

[0088]

[0089] 2. Calculation of soil fertility:

[0090] |S(x,y)-S0|=|4-10|=6

[0091] 3. Calculation of moisture content:

[0092] sin(β1W(x,y))=sin(0.8×30)=sin(24)≈0.4067

[0093] Substituting these values ​​into the composite weight formula:

[0094] W(x,y)=0.4×0.9999+0.3×6+0.3×0.4067=0.39996+1.8+0.122=2.32196

[0095] Therefore, the comprehensive weight of the eastern area is 2.32, indicating that the ecological suitability of this area is low and it is not suitable for growing plants that require high soil fertility.

[0096] In this way, the plant planting layout of different areas can be dynamically adjusted to generate appropriate planting plans based on environmental factors (such as light, soil fertility and water) in different areas. Areas with higher comprehensive weights can be planted with plants that require higher light and soil fertility, while areas with lower comprehensive weights can be planted with plants that are more shade-tolerant or adapted to poor soils, thereby achieving optimal resource allocation among plants and optimizing the growth environment.

[0097] In the optimization method of spatial distribution of garden plants, in addition to the dynamic weight environmental factor modeling algorithm, another core technology is the construction of variable Voronoi diagram. This algorithm ensures that the distribution of plants is highly matched with the suitability of the surrounding environment by dynamically adjusting the influence range of each planting point. The core idea is to calculate the weighted influence range of each planting point and use the geometric properties of the Voronoi diagram to generate an adaptive plant layout.

[0098] Light intensity: The light intensity in the central area of ​​the park is higher, reaching 1000 lux, while the light intensity in the surrounding shaded areas is weaker, only 200 lux.

[0099] Soil fertility: The soil fertility on the east side of the park is higher, close to the standard fertility value of 10, about 9; while the soil on the west side is relatively poor, with a fertility value of only 5.

[0100] Moisture content: There is a noticeable difference in moisture content in the park, with the west side, which has poor drainage, having a moisture content of 60%, while the east side has a moisture content of only 40%.

[0101] Construction of Voronoi diagram: Based on the above environmental data, the following formula is used to calculate the weighted distance from each point to the planting point to determine the boundary of the Voronoi unit. The formula is:

[0102]

[0103] Among them, d i Represents the weighted distance from the position (x, y) to the i-th planting point; (x i ,y i) represents the coordinates of the ith planting point; W(x,y) is the dynamic weight of the position. The key to this formula is that the distance between the area with high environmental suitability (i.e., high-weight area) and the planting point will be shortened, while the distance between the area with low suitability will be lengthened, thereby ensuring that the planting points are more likely to be distributed in suitable areas.

[0104] Set the park to be divided into several areas, and select some plant planting points (such as the southeast corner, middle and northwest corner of the park) for calculation. The following are the specific steps:

[0105] The three planting points selected in the park are set as follows:

[0106] Planting point 1 (located in the center of the park): (x1, y1) = (50, 50)

[0107] Planting point 2 (located on the east side of the park): (x2, y2) = (20, 50)

[0108] Planting point 3 (located on the west side of the park): (x3, y3) = (80, 50)

[0109] In order to calculate the weighted distances from these locations to each planting point, the Euclidean distance is first adjusted according to the above formula. The setting has been modeled by the dynamic weighted environmental factor, and the environmental suitability weights of the following locations are calculated (the setting has been obtained by the formula):

[0110] Location A (coordinates (x, y) = (30, 50)): Environmental suitability weight W (30, 50) = 1.5

[0111] Location B (coordinates (x, y) = (60, 50)): Environmental suitability weight W (60, 50) = 0.8

[0112] Location C (coordinates (x, y) = (70, 50)): Environmental suitability weight W (70, 50) = 2.0

[0113] For position A ((x, y) = (30, 50)), we need to calculate the weighted distance from it to the three planting points. By using the formula, we can get:

[0114] 1. Distance to planting point 1:

[0115]

[0116] 2. Distance to planting point 2:

[0117]

[0118] 3. Distance to planting point 3:

[0119]

[0120] Therefore, the weighted distances of location A from the three planting points are 8, 4, and 20, respectively. According to the Voronoi diagram algorithm, location A will be classified into the area affected by the nearest planting point 2. Next, the weighted distances of location B and location C can be calculated similarly. The following results are obtained:

[0121] The weighted distances of position B ((x, y) = (60, 50)) are 8, 16, and 4 respectively, and it is finally classified into the influence area of ​​planting point 3. The weighted distances of position C ((x, y) = (70, 50)) are 12, 8, and 6 respectively, and it is finally classified into the influence area of ​​planting point 2.

[0122] In this way, the Voronoi diagram divides each location into the area of ​​influence of the appropriate planting point, thereby optimizing the planting space layout of the plants and ensuring that each plant can grow under its best environmental conditions.

[0123] In the optimization method of spatial distribution of garden plants, multi-objective optimization and layout iteration are important strategies to achieve the optimal plant layout. This method adjusts the location of plant planting points by designing a multi-objective optimization function and combining iterative adaptive algorithms to ensure that plant distribution not only takes into account ecological suitability, but also takes into account aesthetics and reasonable allocation of resources. Specifically, the optimization objective function is designed as:

[0124]

[0125] Among them, F represents the overall optimization target value, which measures the comprehensive benefits of the entire garden layout; φ1 represents the ecological benefit contribution of plant growth; W(x,y) is the dynamic weight of the position (x,y); R(x,y) is the resource competition function, which is used to quantify the degree of competition for resources at this position among neighboring plants; λ1 is the nonlinear contribution of adjusting the dynamic weight to the ecological benefit, controlling the impact of ecological factors on the overall layout; φ2 is used to balance the negative impact of resource competition on the overall layout; Ω represents the range of the planting area, which is used to integrate and calculate the comprehensive benefits of the entire space. This objective function evaluates the ecological benefits and resource competition of the entire garden by solving the integral, and optimizes the spatial distribution of plants through adaptive iteration.

[0126] The environmental data in the park are as follows:

[0127] Light intensity: The central area of ​​the park has a strong light intensity of 1000 lux, while the surrounding shadow area has a weak light intensity of only 200 lux. Soil fertility: The soil fertility on the east side of the park is higher, with a fertility value of 9, close to the standard fertility value of 10; the soil on the west side is poor, with a fertility value of 5. Moisture content: The west side of the park has poor drainage, with a moisture content of 60%, while the east side has a lower moisture content of 40%. These environmental factors are integrated into the multi-objective optimization function to iteratively optimize the plant layout.

[0128] First, based on the above environmental data, calculate the dynamic weight (W(x,y)) of each location. Assume that the weight values ​​of the following locations have been obtained through the previous algorithm:

[0129] The light intensity at position A ((x, y) = (30, 50))) is low, the soil fertility is moderate, and the water content is high, so the environmental suitability is good, (W(30, 50) = 1.5).

[0130] The light intensity at position B (coordinate (x, y) = (60, 50)) is high, but the water content is low, and the soil fertility is medium, W (60, 50) = 0.8.

[0131] The light at position C (coordinate (x, y) = (70, 50)) is strong, but the soil at this position is poor, the water content is low, and the environmental suitability is poor, W (70, 50) = 2.0.

[0132] The resource competition function R(x,y) quantifies the degree of resource competition between plants at position (x,y). It is assumed that the resource competition is stronger in areas with denser plant roots, and vice versa, the competition is weaker in areas with sparser root systems. For simplicity, the resource competition function value for each location in the park is set as:

[0133] There are a certain number of plants around position A (coordinate (30,50)), and the resource competition is medium, R(30,50) = 0.5.

[0134] Position B (coordinate (60,50)) is in a relatively sparse area, where resource competition is low, R(60,50) = 0.3.

[0135] There are dense plants near position C (coordinate (70,50)), and the resource competition is strong, R (70,50) = 0.7.

[0136] The following coefficient values ​​were selected for calculation:

[0137] φ1=0.8(ecological benefit contribution of plant growth)

[0138] λ1=0.5 (the impact of adjustment weight on ecological benefits)

[0139] φ2=0.4 (balance the negative impact of resource competition on layout)

[0140] The objective function value at each position can be calculated:

[0141] 1. Optimization objective function of position A (coordinate (30,50)):

[0142]

[0143] First, calculate the impact of weights on ecological benefits:

[0144]

[0145] Then, minus the resource contention term:

[0146] F A =0.5435-0.4×0.5=0.5435-0.2=0.3435

[0147] 2. Optimization objective function of position B (coordinate (60,50)):

[0148]

[0149] Calculate the weights affecting ecological benefits:

[0150]

[0151] Then, minus the resource contention term:

[0152] F B =0.4795-0.4×0.3=0.4795-0.12=0.3595

[0153] 3. Optimization objective function of position C (coordinate (70,50)):

[0154]

[0155] Calculate the weights affecting ecological benefits:

[0156]

[0157] Then, minus the resource contention term:

[0158] F C =0.5854-0.4×0.7=0.5854-0.28=0.3054

[0159] According to the calculation results, the objective function value F at position B B=0.3595 is the maximum, so the optimization algorithm will move the planting point closer to position B. Through multiple iterative adjustments, the plant layout is gradually optimized each time according to the dynamic weight and resource competition feedback information, and finally the optimal layout that meets the ecological benefits and balances resource competition is obtained.

[0160] Embodiment 2:

[0161] Combined with Figure 3 , the establishment of a root space coordination model is crucial to ensure the rational allocation of resources among plants and improve the utilization efficiency of underground resources. In order to accurately predict the root distribution, a root density modeling method based on deep learning is proposed by combining underground imaging technology, dynamic prediction model and resource optimization algorithm. This method can accurately evaluate the distribution of plant roots in space and depth, so as to better coordinate the sharing of root resources among plants when arranging garden plants. The root density distribution function is shown as follows:

[0162]

[0163] Among them, R(x,y,z) represents the root density at the position (x,y,z), in roots / cubic meter; R0 is the maximum root density; z is the underground depth, in meters; κ1 is the depth influencing factor, the larger the value, the shallower the root distribution; T(x,y) is the soil temperature distribution function, in degrees Celsius; T0 is the optimum temperature for plant root growth, in degrees Celsius; κ2 is the temperature difference adjustment coefficient, reflecting the sensitivity of temperature change to root distribution; exp(-κ2·|T(x,y)-T0|) is used to simulate the inhibitory effect of temperature difference on root distribution.

[0164] A garden planning project is set up. The project contains a variety of plant species. It is necessary to optimize the root distribution between plants to ensure the rational distribution of resources (such as water and nutrients). In the design, two types of plants are selected: one is a tall tree (such as an oak tree) and the other is a low shrub (such as a rose).

[0165] 1. Root density modeling and parameter setting

[0166] In order to model the spatial distribution of roots, some initial parameters need to be determined first, which will directly affect the accuracy of the root distribution model.

[0167] For oak trees, the maximum root density R0 is set to 500 roots / m3;

[0168] For roses, the maximum root density R0 is set to 200 roots / m3;

[0169] The underground depth z is set to range from 0 m (ground) to 2 m (underground), and plant roots are usually distributed in this area;

[0170] The depth impact factor κ1 is taken as 1.5 for oak and 0.8 for rose;

[0171] The soil temperature distribution function T(x,y) sets the soil temperature range at different locations in the park to 15°C to 30°C;

[0172] The optimum growth temperature T0 for oak trees is set at 22°C, and the optimum growth temperature T0 for roses is set at 18°C;

[0173] The temperature difference adjustment coefficient κ2 was set to 0.5 for oak and 0.3 for rose.

[0174] 2. Root density calculation

[0175] Three representative locations were selected in the park, namely: location A (coordinate (x, y) = (10, 10)), location B (coordinate (x, y) = (50, 50)), and location C (coordinate (x, y) = (80, 20)). Based on the given environmental parameters, the root density model was used to calculate the root density at each location.

[0176] Position A (oak tree, coordinates (x, y) = (10, 10))

[0177] Assuming the soil temperature at this location is 20°C and the underground depth is 1 meter, the root density is calculated as:

[0178]

[0179] Calculation steps:

[0180]

[0181] exp(-0.5·2)=exp(-1)≈0.3679

[0182] Therefore, the root density at location A is:

[0183] R A =500·0.6·0.3679=110.37 pieces / cubic meter

[0184] Position B (rose, coordinates (x, y) = (50, 50))

[0185] Assuming the soil temperature at this location is 25°C and the underground depth is 1 meter, the root density is calculated as:

[0186]

[0187] Calculation steps:

[0188]

[0189] exp(-0.3·7)=exp(-2.1)≈0.1225

[0190] Therefore, the root density at location B is:

[0191] R B =200·0.4444·0.1225=10.86 pieces / cubic meter

[0192] Position C (rose, coordinates (x, y) = (80, 20))

[0193] Assuming the soil temperature at this location is 30°C and the underground depth is 2 meters, the root density is calculated as:

[0194]

[0195] Calculation steps:

[0196]

[0197] exp(-0.3·12)=exp(-3.6)≈0.0273

[0198] Therefore, the root density at location C is:

[0199] R C =200·0.6154·0.0273=3.36 pieces / cubic meter

[0200] Through the above calculations, we can see that the root density at position A (oak) and position B (rose) is significantly higher, especially the root density of the oak is very dense at this position, while the root density of the rose is lower at position C due to the difference in soil temperature and the change in underground depth. In actual garden design, the planting position of plants can be adjusted according to these data to avoid excessive competition of roots and optimize the utilization efficiency of underground water and nutrients.

[0201] In the optimization method of spatial distribution of garden plants, the dynamic growth prediction model of the root system is a key tool to improve the efficiency of plant resource utilization. By introducing dynamic growth factors based on the existing static root distribution model, it is possible to more accurately simulate the changes of plant roots over time and adjust the growth layout of plants in combination with seasonal fluctuations. This model predicts the growth of plant roots by introducing the time dimension, thereby providing garden designers with a more accurate plant layout plan. The following are specific application examples based on the dynamic root growth prediction model:

[0202] The model formula is as follows:

[0203] R t (x,y,z,t)=R(x,y,z)·(1-e-μ·t )·(1+δ·sin(ω·t+φ))

[0204] Among them, R t (x, y, z, t) is the root density at position (x, y, z) at time t, in roots / cubic meter; R(x, y, z) is the root density calculated based on the static model; μ is the root growth rate factor, which determines the root expansion speed; t is time, indicating the growth stage of the plant; δ is the amplitude factor of the root density fluctuation, indicating the impact of seasonal fluctuations on the root density; ω is the seasonal change frequency factor, in rad / year, which determines the periodicity of the root density fluctuation; φ is the phase offset, which is used to adjust the starting time of seasonal changes.

[0205] Let's continue designing the garden planning project mentioned above and use the dynamic root growth prediction model to further optimize the root distribution of plants. For ease of understanding, let's select two plant types: oak (tall tree) and rose (low shrub), and the static calculation of root density has been completed. Next, we will perform dynamic prediction based on the model to understand the changes of plant roots over time.

[0206] First, you need to set the model parameters according to the plant species and environmental conditions:

[0207] For oak trees, the maximum root density R0 is set to 500 roots / m3;

[0208] For roses, the maximum root density R0 is set to 200 roots / m3;

[0209] Root growth rate factor μ: The growth rate of the oak is set to be slower, set to 0.1 / year, and the growth rate of the rose is set to be faster, set to 0.3 / year; Amplitude factor δ: Taking into account seasonal changes, the oak is set to 0.1 and the rose is set to 0.2, indicating that the root system of the rose fluctuates greatly during seasonal changes; Seasonal change frequency factor ω: It is set that the root system will go through a complete seasonal cycle every year, set to ω=2πrad / year; Phase offset φ: It is set that the seasonal change starts from spring, set φ=0rad.

[0210] Select 5 years as the time span and calculate the distribution changes of plant roots in each year. In the selected park, the static root density of oak and rose is 110.37 roots / m3 (location A) and 10.86 roots / m3 (location B), respectively. On this basis, the change of root density each year will be calculated.

[0211] Location A (Oak Tree):

[0212] Assuming the root growth rate of the oak tree is 0.1 / year and the initial root density at the location is 110.37 roots / m3, calculate the change in root density each year over 5 years.

[0213] Year 1:

[0214] R t (10,10,1,1)=110.37·(1-e -0.1·1 )·(1+0.1·sin(2π·1+0))

[0215] Calculation steps:

[0216] 1-e -0.1 ≈0.0952

[0217] sin(2π·1)=0 (root density fluctuation is 0 in spring)

[0218] so,

[0219] R t (10,10,1,1)≈110.37·0.0952·1=10.50 pieces / cubic meter

[0220] Year 2:

[0221] R t (10,10,1,2)=110.37·(1-e -0.1·2 )·(1+0.1·sin(2π·2+0))

[0222] Calculation steps:

[0223] 1-e -0.2 ≈0.1813

[0224] sin(4π)=0

[0225] so,

[0226] R t (10,10,1,2)≈110.37·0.1813·1=20.00 roots / cubic meter

[0227] Year 3:

[0228] R t (10,10,1,3)=110.37·(1-e -0.1·3 )·(1+0.1·sin(2π·3+0))

[0229] Calculation steps:

[0230] 1-e -0.3 ≈0.2592

[0231] sin(6π)=0

[0232] so,

[0233] R t (10,10,1,3)≈110.37·0.2592·1=28.58 roots / cubic meter

[0234] In this way, the root density can be calculated for each time point, thereby predicting the growth trend of plant roots in the next few years and optimizing the garden design.

[0235] Position B (Rose):

[0236] Similarly, the root density of roses is dynamically predicted and the corresponding parameters are set. The root density of roses is set to have a larger fluctuation, an amplitude factor of 0.2, and a growth rate of 0.3 / year.

[0237] For example, the initial static root density of roses is set to 10.86 roots / cubic meter, and a 5-year dynamic prediction is also performed.

[0238] Through the calculation results of the dynamic root growth prediction model, garden designers can adjust the plant layout more accurately. For example, in the first year, the root density of the oak tree changes relatively little, while the root density of the rose fluctuates greatly. Therefore, according to the changes in the dynamic root density, more reasonable space can be left for the planting positions of oak trees and roses in the garden to avoid excessive competition between their roots at different growth stages.

[0239] Underground resource competition optimization model:

[0240] The objective function has the following form:

[0241]

[0242] Objective Function Explanation

[0243] C(x,y,z): represents the superposition of root densities of different plants, usually obtained by calculating the root density of each plant. For example, at a certain location (x,y,z), there are two plants A and B, whose root densities are R A (x,y,z) and R B (x,y,z), then C(x,y,z)=R A (x,y,z)+R B (x,y,z). A greater root density means more intense competition for resources.

[0244] D(x,y,z): represents the spatial distance between plant roots. The closer the distance, the more intense the resource competition. Therefore, increasing the distance between plants can effectively reduce the intensity of competition.

[0245] λ: Distance adjustment coefficient. This coefficient controls the effect of distance on resource competition. The larger the value, the stronger the effect of distance on alleviating resource competition. By adjusting this coefficient, the competition behavior between different plants can be simulated.

[0246] Parameter settings

[0247] To make it easier to understand, let's optimize the planting layout of two plants in a garden design project: oak (tall trees) and roses (low shrubs). Assume that the static calculation of the plant root density has been completed and the spatial distance between the roots is known. Next, we will set specific parameters for calculation.

[0248] For oak trees, the maximum root density R0 is set to 500 roots / m3;

[0249] For roses, the maximum root density R0 is set to 200 roots / m3;

[0250] For the distance D(x,y,z) between two plants, the initial distance is set to 3 meters;

[0251] The distance adjustment coefficient λ = 1.5, indicating that the distance has a stronger effect in alleviating competition;

[0252] Select a garden with a planting area Ω of 10m x 10m;

[0253] The root density of the plant is set to change over time according to the above dynamic growth model.

[0254] 1. Calculation of root resource competition intensity:

[0255] First, calculate the root resource competition intensity C(x,y,z) at each location (x,y,z). Set the static root density in the park as follows:

[0256] Location A (oak): static root density is 110.37 roots / m3;

[0257] Location B (rose): static root density is 10.86 roots / m3;

[0258] The root density of these locations in the first year (i.e., t = 1) is calculated as follows:

[0259] Root density calculation for location A (oak tree):

[0260] According to the dynamic root growth prediction model, the change of oak root density in the first year was calculated:

[0261] R t (x,y,z,1)=110.37·(1-e -0.1·1 )·(1+0.1·sin(2π·1+0))

[0262] R t (x,y,z,1)=110.37·0.0952·1=10.50 roots / cubic meter

[0263] Root density calculation for location B (rose):

[0264] According to the dynamic growth prediction model of roses, the root density of roses is assumed to fluctuate greatly, and the amplitude factor is 0.2:

[0265] R t (x,y,z,1)=10.86·(1-e -0.3·1 )·(1+0.2·sin(2π·1+0))

[0266] R t (x,y,z,1)=10.86·0.2592·1=2.82 roots / cubic meter

[0267] Therefore, the root density at location A is 10.50 roots / m3 and the root density at location B is 2.82 roots / m3.

[0268] 2. Calculate the root resource competition intensity C(x,y,z):

[0269] At the same location (x, y, z), the root density of different plants added together is the resource competition intensity C(x, y, z). Assuming that location A and location B are adjacent, and the root density is 10.50 and 2.82 roots / m3 respectively, then:

[0270] C(x,y,z)=R A (x,y,z)+R B (x,y,z)=10.50+2.82=13.32 roots / cubic meter

[0271] 3. The mitigating effect of computing resource competition on distance:

[0272] According to the formula, the mitigation effect of resource competition is controlled by the distance D(x,y,z). If the initial distance between the two plants is set to 3 meters, then:

[0273]

[0274] 4. Calculate the objective function F:

[0275] The objective function F is calculated by integrating the resource competition intensity in the entire planting area Ω. The result of the calculation reflects the resource competition intensity among plant roots in the entire area. For this calculation process, the entire park is set to an area of ​​10 meters x 10 meters, so:

[0276]

[0277] 5. Optimization suggestions

[0278] Based on the calculation results, we can get the intensity of root resource competition in the park. By adjusting the layout of plant planting points, such as increasing the distance between plants or changing the planting density, we can further reduce this competition intensity and improve the growth efficiency of plants. For example, increasing the distance between oak trees and roses can alleviate resource competition and increase the growth space of each plant.

[0279] Embodiment 3:

[0280] Combined with Figure 4 , set to plan an ecological park in the south of the city, the goal is to improve ecological benefits and achieve intelligent management and monitoring by scientifically optimizing the spatial layout of plants. The park covers an area of ​​20 hectares, with undulating terrain, changeable climatic conditions, and a wide variety of vegetation. The digital twin system is used to optimize the design and management of the garden.

[0281] 1. Data collection and integration: The core of the digital twin system is the integration and dynamic update of multi-source data. It is necessary to comprehensively collect the topography, climate and plant growth data of the park and build a digital twin model based on this.

[0282] Topographic data: The park's terrain was scanned using drones equipped with LiDAR technology to generate a 3D terrain model with an accuracy of 1 meter. The data showed that the highest point in the park is in the northern highlands (50 meters above sea level), the lowest point is in the southern lake area (5 meters above sea level), and the slope is between 5% and 20%.

[0283] Climate data: Real-time climate data is collected through a network of meteorological sensors, including temperature (15°C to 30°C), humidity (40% to 80%), light intensity (200lux to 1200lux) and precipitation (average annual precipitation is 1200mm).

[0284] Plant growth data: Geological radar and underground electrical resistance imaging technology are used to obtain plant root distribution data, and NDVI (normalized difference vegetation index) technology is used to monitor the health of the aboveground parts of plants. It is assumed that there are two types of vegetation in the park, trees (such as ginkgo) and shrubs (such as roses), and the static parameters of their growth status are known: the growth rate of ginkgo is μ = 0.1, and the growth rate of rose is μ = 0.3.

[0285] Real-time IoT data: Soil moisture sensors, carbon dioxide concentration sensors, temperature and humidity sensors and other devices upload data to the cloud platform in real time to update the digital twin model.

[0286] 2. Digital twin modeling: Based on the data collected above, a virtual model of the ecological park was constructed in the digital twin system.

[0287] The model includes:

[0288] 1. Three-dimensional terrain model: Based on LiDAR data, a terrain undulation map is constructed to divide highlands, slopes and low-lying areas.

[0289] 2. Ecological factor distribution model: Map climate data and soil data to the spatial grid of the park to generate a dynamic weight map W(x,y).

[0290] 3. Plant growth dynamic model: Using the above plant dynamic growth formula, simulate the root growth trend of ginkgo and rose t (x,y,z,t).

[0291] 3. Real-time data synchronization: In order to ensure that the digital twin model can be synchronized with the actual park status in real time, a data transmission network is built through the Internet of Things technology. The following are the specific steps:

[0292] 1. Sensor network deployment: 50 sensor nodes are deployed in the park to collect real-time data on climate, soil, plant growth, etc. These nodes upload data to the cloud platform through the LoRa network.

[0293] 2. Real-time data fusion: Using spatiotemporal interpolation algorithms, discrete sensor data are integrated into a continuous spatial distribution map. For example, soil moisture H(x,y) varies greatly on slopes, ranging from 20% to 60%, and is more uniform in low-lying areas, stable at 70%.

[0294] 4. Virtual simulation and optimized decision-making: The digital twin model supports simulation and optimization in a virtual environment, including the following functions:

[0295] 1. Ecological benefit simulation: By simulating plant growth and climate change, the ecological benefits (such as carbon absorption and cooling effect) in the next five years are predicted. The annual carbon absorption of each ginkgo tree is set at 22kg and that of the rose is 8kg. It is estimated that the carbon absorption of the entire garden can reach 12.5 tons in five years.

[0296] 2. Resource competition optimization: According to the underground resource competition optimization model, the layout of plant planting points is adjusted. The initial planting density of ginkgo and rose in a certain area (area of ​​100 square meters) is set to 4 plants / 10 square meters and 10 plants / 10 square meters. After optimization calculation, the density is adjusted to 3 plants / 10 square meters and 12 plants / 10 square meters to reduce the root competition of ginkgo.

[0297] 3. Disaster risk prediction: Use meteorological simulation to predict the impact of extreme weather (such as heavy rain or drought) on plant growth and optimize irrigation plans. If the soil moisture content in low-lying areas exceeds 80% during heavy rain, the system will generate drainage optimization suggestions.

[0298] 5. Virtual-Real Interaction and User Optimization

[0299] The digital twin system uses VR and AR technologies to achieve virtual-real interaction in garden design and management. The following are specific application scenarios:

[0300] 1. VR immersive experience: Landscape designers can enter the virtual park model by wearing VR equipment to directly observe the plant layout and ecological benefits. Designers can adjust the planting positions of ginkgo and roses in real time and observe the changes in carbon absorption after the adjustments.

[0301] 2. AR real-life overlay: On-site managers can use AR glasses to view the comparison between the actual park and the virtual model in real time. For example, if the soil moisture in a certain area is too low, the AR system will highlight the area and recommend additional irrigation.

[0302] In the digital twin system, the soil moisture in a low-lying area (50 square meters) is as high as 80%, and the resource competition between the ginkgo and rose roots is strong. According to the resource competition optimization formula:

[0303]

[0304] Set the root density of ginkgo and rose to R A =200 and R B =50, the distance between them is D = 3 meters, and the distance adjustment coefficient λ = 1.2. The competition intensity is:

[0305] C(x,y,z)=R A ·R B =200·50=10,000

[0306] The resource contention mitigation coefficient is:

[0307]

[0308] The objective function value is:

[0309]

[0310] By optimizing and adjusting the distance between the two to 5 meters, the resource competition mitigation coefficient is recalculated as follows:

[0311]

[0312] The objective function value is reduced to:

[0313] F=10,000·0.143=1,430

[0314] After optimization, the intensity of resource competition decreased by 34%, significantly alleviating the competition for underground resources.

[0315] Real-time data synchronization and dynamic adjustment

[0316] The key to the digital twin system is to ensure that the virtual model V(x, y, z, t) is synchronized with the actual environment. The formula is as follows:

[0317] V(x,y,z,t)=R(x,y,z,t)+Δ(t)

[0318] in:

[0319] V(x,y,z,t): represents the dynamic state of the virtual garden model at position (x,y,z) and time t.

[0320] R(x,y,z,t): A preliminary model of the garden generated from initial terrain and climate data.

[0321] Δ(t): Deviation corrected by real-time data, indicating the impact of actual environmental changes on the model.

[0322] The original soil moisture value of the northern slope of the park is set at 30%. According to real-time data, sensors show that the humidity in this area has increased to 50%. The virtual model is updated in the following ways:

[0323] 1. Calculate the deviation: Δ 湿度 =50%-30%=20%.

[0324] 2. Update model: Correct the humidity value of the northern slope in the virtual model from 30% to 50% to match the actual data.

[0325] In addition, it is assumed that due to the change in light intensity (from 1000 lux to 800 lux), the plant growth state in the virtual model needs to be updated. t (x,y,z,t), recalculate the growth rate μ and root density changes.

[0326] Ecological Benefit Forecast The ecological benefits of the park in the next five years are predicted through a virtual model. For example, each ginkgo tree absorbs 22kg of carbon per year, and each rose plant absorbs 8kg. The total carbon absorption of the park in five years is expected to be:

[0327] Total carbon absorption = (22 × 100) + (8 × 300) × 5 = 22,000 kg = 22 tons

[0328] Resource competition optimization

[0329] Combined with the underground resource competition optimization model, the planting density is adjusted to reduce resource competition. For example, the initial planting density of ginkgo and rose in a certain area (area of ​​100 square meters) is 4 plants / 10 square meters and 10 plants / 10 square meters respectively. By adjusting the spacing, the density is optimized to 3 plants / 10 square meters and 12 plants / 10 square meters. According to the objective function:

[0330]

[0331] When λ=1.2, the competition intensity F=2,170 before optimization and F=1,430 after optimization, a 34% reduction in competition intensity.

[0332] Disaster risk simulation

[0333] Using meteorological simulation, it is predicted that the heavy rain will cause the soil moisture content in the southern lake area to exceed 80%. The system generates drainage recommendations and simulates the impact of drainage on plant growth. By adjusting the irrigation system in real time, it ensures that the soil moisture returns to 70%.

[0334] Virtual reality (VR) experience: Landscape designers use VR devices to enter the virtual garden and view the plant layout and ecological status in real time. Designers can simulate the changes in carbon absorption under different layouts. For example, adjust the density of ginkgo planting on the northern slope to observe its impact on the overall ecological benefits. On-site managers use AR glasses to view the comparison between the actual garden and the virtual model. For example, if the actual soil moisture in a certain area is too low, the AR system will highlight the area and provide irrigation suggestions.

[0335] The project is set to be responsible for the design and management of an ecological park covering an area of ​​20 hectares, including trees (such as ginkgo), shrubs (such as roses) and low lawns. The project goal is to optimize plant distribution and improve carbon absorption efficiency while rationally utilizing water resources and other ecological services. A highly interactive visualization platform is used to assist in design and optimization, and the ecological benefits of different layouts can be observed by adjusting planting plans and irrigation systems in real time.

[0336] The interactive platform provides designers with optimization decision support through real-time simulation and feedback. The interactive optimization objective function is defined as:

[0337] Finteract =∫ Ω [C(G(t),E(t))-L(D)]d A

[0338] Among them: F interact represents the target value after interactive optimization, which is used to measure the comprehensive benefits after the user adjusts the plan; Ω is the spatial scope of the garden (such as the entire park); C(G(t), E(t)) is the value function of the garden ecological service, which depends on the plant growth state G(t) and environmental data E(t); L(D) is the resource consumption function, which represents the resource use cost caused by planting layout and irrigation design; d A is the area element of each spatial unit used for integral calculation.

[0339] 1. Topography and climate data: A three-dimensional terrain model was obtained using a drone equipped with LiDAR. The highest point in the park is 50 meters above sea level, the lowest point is 5 meters, and the slope is 5% to 15%. Climate sensors collect real-time data such as temperature (15℃ to 30℃), humidity (40% to 80%), and light (200lux to 1200lux).

[0340] 2. Plant data: The park is set to plant 100 ginkgo trees and 300 roses. The annual carbon absorption of ginkgo is 22kg / tree and that of rose is 8kg / tree. The health status of vegetation is monitored by NDVI, and the root distribution is obtained by combining underground electrical resistance imaging technology.

[0341] 3. Irrigation and resource usage data: The park is equipped with an irrigation system to update soil moisture data in real time. Current data shows that the humidity of the northern slopes is low (30%) and the humidity of the southern low-lying areas is high (80%).

[0342] After the initial model is generated, a virtual garden is created through the VR / AR system, and users can enter the virtual environment to view the distribution of plants, root expansion, and soil moisture distribution in real time.

[0343] 1. Users adjust planting plans: Through the VR platform, users found that the planting density of ginkgo trees on the northern slopes was low (3 trees / 10 square meters), resulting in carbon absorption efficiency not meeting expectations. Users tried to increase the planting density to 5 trees / 10 square meters, while reducing the planting density in the southern low-lying areas to avoid excessive resource competition.

[0344] Optimize calculation:

[0345] Carbon absorption of the original solution:

[0346] C 原 =(3·22+10·8)·10=1,060kg / year

[0347] Carbon absorption of the new solution (after increasing density):

[0348] C 新 =(5·22+8·8)·10=1,320kg / year

[0349] Carbon absorption increased by approximately 24.5%.

[0350] 2. Users adjust the irrigation system: Using AR to view the soil moisture distribution in the irrigation area in real time, users found that the humidity in the northern slope was too low (only 30%), while the humidity in the southern low-lying area was too high (80%). Users adjusted the irrigation range and flow through the interactive platform and redistributed the irrigation water, increasing the humidity in the northern slope to 50% and reducing it to 70% in the southern low-lying area.

[0351] Resource consumption optimization: The total irrigation cost of the original scheme (the unit water cost is set to 1.2 yuan / cubic meter):

[0352] L 原 =(50·0.3+50·0.8)·1.2=78 yuan

[0353] Total irrigation cost of the new solution:

[0354] L 新 =(50·0.5+50·0.7)·1.2=72 yuan

[0355] Irrigation costs were reduced by approximately 7.7%.

[0356] 3. User simulation of ecological benefits: Through the virtual engine, users observe the impact of adjustment plans on ecological benefits. After adjusting the planting and irrigation plans, the total carbon absorption and resource consumption of the park in the next five years are:

[0357] Total carbon absorption:

[0358] C 总 =1,320·5=6,600kg=6.6 tons

[0359] Total resource consumption:

[0360] L 总 =72·5=360 yuan

[0361] Users can adjust different parameters through the interactive platform and view the effect of the optimization scheme in real time. The system will update the virtual model F interact , providing instant feedback to ensure the optimal solution.

[0362] Through the highly interactive visualization platform, garden designers and managers can intuitively evaluate the impact of adjustment plans, quickly locate problems and optimize layouts. The following is the final optimization effect:

[0363] 1. The total amount of carbon absorbed increased from 5.3 tons to 6.6 tons, an increase of 24.5%;

[0364] 2. Irrigation costs were reduced from 390 yuan to 360 yuan, saving about 7.7%;

[0365] 3. Maximize ecological benefits while ensuring resource utilization.

[0366] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for optimizing the spatial distribution of garden plants, characterized in that The following steps are involved: S1, plant space adaptive layout algorithm; S1.1, Terrain modeling: Use LiDAR or UAV technology to generate a high-precision 3D model of the garden terrain; S1.2, Dynamic weight calculation: Generate a weight matrix based on soil type, light intensity and drainage characteristics; S1.3, variable Voronoi diagram construction: Apply the dynamic weight Voronoi algorithm in the terrain model to generate the planting range of each plant; S2, spatial coordination model of roots among plants; S2.

1. Root system simulation modeling: Use underground imaging techniques including electrical resistance imaging and geological radar to establish root distribution models; S2.2, Water and nutrient flow analysis: Combined with groundwater hydrological data, simulate the dynamic distribution of water and nutrients among the root system; S3, intelligent irrigation and monitoring system embedding; S3.1, collect sensor data in real time using wireless networks; S3.

2. Design dynamic irrigation plans based on plant water requirements and real-time data; S3.

3. Optimize plant planting density and irrigation range through data analysis; S4, digital twin garden system; S4.

1. Digital twin model generation: integrating terrain, climate, plant growth and other data into the virtual garden; S4.

2. Ensure that the virtual garden is consistent with the actual state by interacting with the digital twin model through sensors; S4.

3. Use digital models to predict plant growth, ecological benefit changes and disaster risks; S4.

4. Through the visual interface, users are allowed to modify and optimize the garden layout in real time.

2. The method for optimizing the spatial distribution of garden plants according to claim 1, characterized in that The plant space adaptive layout algorithm includes dynamic weight environmental factor modeling, expressing the influence of different ecological factors on plant distribution with nonlinear functions, dynamically responding to changes in factor values ​​in space, and generating appropriate weight distribution; The weight function is defined as: Among them, W(x,y) represents the comprehensive weight of the position (x,y), which is used to reflect the ecological suitability of the area; L(x,y) represents the light intensity at the position (x,y); S(x,y) represents the soil fertility at the position (x,y); S0 represents the target soil fertility value, which is used to evaluate the difference between the actual fertility and the target value; W(x,y) also represents the soil moisture content at the position; γ1, γ2, and γ3 are the influence weights of light, soil fertility, and moisture, respectively; α1 is the nonlinear influence range of controlling light intensity; β1 is the fluctuation effect of regulating the periodic change of moisture on the weight.

3. The method for optimizing the spatial distribution of garden plants according to claim 2, characterized in that The plant space adaptive layout algorithm includes the construction of a variable Voronoi diagram, and the influence range of each planting point is adjusted by dynamic weight to highly match the environmental suitability; the boundary of the Voronoi unit is determined by the following formula: where d i Represents the weighted distance from the position (x, y) to the i-th planting point; (x i ,y i ) represents the coordinate position of the ith planting point; (x, y) represents the spatial position of the current evaluation; W(x, y) is the dynamic weight, which represents the impact of environmental suitability on the current position; 1+W(x, y) is used to adjust the Euclidean distance so that the distance of high-weight (strong environmental suitability) areas is shortened and the distance of low-weight areas is lengthened.

4. The method for optimizing the spatial distribution of garden plants according to claim 3, characterized in that The plant space adaptive layout algorithm includes multi-objective optimization and layout iteration. A multi-objective optimization function is designed to adjust the distribution position of the planting points through an adaptive iterative algorithm. The optimization objective function is defined as: Where F represents the overall optimization target value, which measures the comprehensive benefit of the entire layout; φ1: represents the ecological benefit contribution of plant growth, such as carbon absorption, soil conservation, etc.; W(x, y) is the dynamic weight, which represents the ecological suitability of the position (x, y); R(x, y) is the resource competition function, which is used to quantify the degree of competition for resources at this location among neighboring plants; λ1 is the nonlinear contribution of adjusting the dynamic weight to the ecological benefit; φ2 is used to balance the negative impact of resource competition on the overall layout; Ω represents the range of the planting area, which is used to integrate and calculate the comprehensive benefit of the entire space.

5. The method for optimizing the spatial distribution of garden plants according to claim 4, characterized in that The establishment of the spatial coordination model of plant roots includes obtaining underground root spatial distribution data through underground imaging technology, combining a dynamic prediction model and a resource optimization algorithm to control the efficient coordination between plant roots and the optimal utilization of underground resources.

6. The method for optimizing the spatial distribution of garden plants according to claim 5, characterized in that The establishment of the spatial coordination model of root systems among plants includes a dynamic root growth prediction model: R t (x,y,z,t)=R(x,y,z)·(1-e -μ·t )·(1+δ·sin(ω·t+φ)) Among them, R t (x,y,z,t) is the root density at position (x,y,z) at time t; R(x,y,z)) is the basic value of the static root density model; μ is the root growth rate factor; t is time, in years; δ is the amplitude factor of root density fluctuation, describing the regulatory effect of seasonal changes on root density; ω is the seasonal change frequency factor, in rad / year, reflecting the periodicity of root fluctuations; φ represents the phase offset; (1-e -μ·t ) is used to simulate the cumulative effect of root growth over time; and (1+δ·sin(ω·t+φ)) simulates the cyclical effect of seasonal fluctuations on root density.

7. The method for optimizing the spatial distribution of garden plants according to claim 6, characterized in that The establishment of the spatial coordination model of the root systems among plants includes an underground resource competition optimization model, which controls the optimization of the layout and density of planting sites through resource competition.

8. The method for optimizing the spatial distribution of garden plants according to claim 1, characterized in that The digital twin system integrates multi-source data and real-time Internet of Things technology to build a highly dynamic and intelligent garden management platform; through virtual reality VR and augmented reality AR technology, garden design and management plans can be synchronized in real time in virtual and real environments.

9. The method for optimizing the spatial distribution of garden plants according to claim 8, characterized in that The digital twin system includes virtual and real-world synchronous modeling: through three-dimensional modeling and real-time data update, a virtual garden model V (x, y, z, t) is established, and the model is dynamically adjusted according to actual environmental changes; through virtual engines and physical simulation technology, the actual changes of the garden are fed back to the virtual model in real time; V(x,y,z,t)=R(x,y,z,t)+Δ(t) Among them, V(x,y,z,t) represents the spatial coordinates (x,y,z) and the state of the virtual garden model at time t; R(x,y,z,t) represents the preliminary garden model generated based on the initial terrain and climate data, the spatial distribution and attributes of the model; Δ(t) represents the deviation corrected by real-time data.

Images