Garden landscape full-life-cycle collaborative management method based on BIM and GIS

Abstract

The invention discloses a BIM (Building Information Modeling) and GIS (Geographic Information System)-based garden landscape full-life-cycle collaborative management method. According to the method, the plant design model of the BIM and the environmental data monitored by the GIS in real time are integrated, the constructed digital twinborn system can dynamically simulate the plant growth state, the adaptive irrigation amount is calculated according to the real-time soil humidity and the plant growth stage, and the problem of excessive irrigation or insufficient irrigation in traditional maintenance is avoided. And in combination with long-term accumulated growth data and environmental parameters, the system can pre-judge plant growth requirements and potential risks in advance, and generates maintenance suggestions fitting the actual site, so that the maintenance measures are always adaptive to the current state of the plant. Meanwhile, multi-dimensional factors are integrated to give an objective updating sequence, resources are inclined to the most urgently needed link, the resource utilization efficiency is improved, the operation and maintenance cost is reduced, the site initial ecological background is protected, and therefore the system can better maintain the stability of plant communities.

Description

Technical Field

[0001] This invention belongs to the field of landscape management technology, specifically a collaborative management method for the entire life cycle of landscape architecture based on BIM and GIS. Background Technology

[0002] Landscape architecture lifecycle management is a systematic management concept and practice that views a landscape project from conceptualization to its eventual demise as a complete and continuous process, requiring comprehensive planning and meticulous control. This management covers the initial site survey, conceptual design, and scheme demonstration; the mid-term construction, seedling procurement, and project supervision; and the later stages of maintenance, plant growth regulation, facility repair, and landscape renewal. Its core lies in breaking down the traditional fragmented model of each stage by establishing an integrated information platform and management standards to achieve data sharing and collaborative decision-making. This maximizes resource utilization efficiency, extends the landscape's value cycle, and reduces overall costs while ensuring both artistic and ecological benefits. It is a crucial guarantee for achieving the sustainability of landscape projects and improving the quality of the living environment.

[0003] However, in landscape management, existing technologies rely heavily on experience-based judgment for maintenance decisions, making it difficult to achieve precise control. Data from different lifecycle stages is fragmented, and decisions in the design, operation, and renewal stages lack unified data support, which can easily lead to unreasonable resource allocation, resulting in low overall management efficiency and difficulty in ensuring long-term ecological value. Summary of the Invention

[0004] The purpose of this invention is to provide a collaborative management method for the entire life cycle of garden landscape based on BIM and GIS in order to solve the problems mentioned above.

[0005] The technical solution adopted in this invention is as follows: a collaborative management method for the entire life cycle of landscape architecture based on BIM and GIS, the method comprising the following steps: S1: Integrate site ecological data collected by GIS with the initial BIM 3D model to conduct assessments such as ecological sensitivity classification, visual corridor simulation, and terrain slope optimization, and output clear design constraints.

[0006] S2: Organize multidisciplinary teams including landscape and architecture to conduct BIM collaborative design based on S1 data, import GIS regional climate and soil type data to verify the adaptability of plant configuration, simulate rainstorm runoff to optimize the drainage system, and generate an integrated BIM model with construction parameters.

[0007] S3: Combine BIM plant models with real-time GIS environmental data to build a digital twin system, simulate plant growth cycles, morphological changes and pest and disease risks, dynamically generate maintenance suggestions, and synchronize the data to the operation and maintenance platform.

[0008] S4: Connect BIM models with IoT devices, visualize facility status and environmental parameters through GIS, develop personalized maintenance plans by combining twin data, optimize visitor routes using heat maps of pedestrian traffic, and archive operation and maintenance data to support updates.

[0009] S5: Integrate full lifecycle data, analyze issues such as facility aging and plant health from multiple dimensions, generate update priorities and solutions, and implement them after BIM pre-simulation verification.

[0010] S6: After the scenic area is decommissioned, assess the recyclable resources and draw a distribution map using GIS. Combine the initial ecological data to formulate a restoration plan, and archive the data for industry reference.

[0011] In a preferred embodiment, step S1 is advanced using a workflow that integrates ArcGIS Pro and Revit. First, site topographic data is collected using oblique photogrammetry conducted by a drone, with a resolution of 5 cm and a collection frequency of once per hour. Hydrological data is collected using automatic rain gauges and water level sensors, with a sampling interval of 15 minutes. Vegetation status data is obtained through a combination of satellite remote sensing imagery and ground surveys, with an image resolution of 1 meter and a survey coverage of 100%. Data for surrounding ecologically sensitive areas is synchronized from the local ecological and environmental bureau's database, with an update cycle of once per day. When establishing the site geographic information database, the coordinate system uses the 2000 National Geodetic Coordinate System, and the elevation datum is the 1985 National Elevation Datum. Then, a three-dimensional initial model of the site is constructed using Revit, including elements such as topography, existing buildings, and vegetation distribution. The topographic model imports DEM data collected by GIS. The vegetation model is modeled according to actual species and locations. The model's coordinate system is kept consistent with the GIS database. Next, FME software is used for GIS and BIM data conversion and integration. Spatial overlay is completed using coordinate matching, corresponding the three-dimensional coordinates of the BIM model to the spatial coordinates of the GIS database. After overlay, an integrated digital twin model of the site is generated.

[0012] In a preferred embodiment, in step S2, all professional models are uniformly imported into the Navisworks platform to set collaborative permissions, and modifications from each profession are synchronized in real time. The model accuracy is maintained at LOD350, conflict detection is performed once per hour, and conflict points are automatically marked when the spacing between components is less than 10 cm. Subsequently, the average temperature, annual precipitation, and soil pH data of the region over the past 10 years provided by the GIS system are imported. The Ecotect Analysis model is used to input parameters such as plant growth temperature range, water requirement, and suitable soil pH to simulate the growth cycle and survival rate of plants on the site. A survival rate threshold of 90% is set, and plant species below the threshold are automatically marked and replaced with native plants that meet the regional climate conditions. Next, the SWMM model is used to import the rainstorm intensity formula, site impervious area ratio, and soil permeability coefficient data from the GIS to simulate the runoff depth, peak flow, and drainage time of a 50-year return period rainstorm. The diameter and slope of the drainage network are optimized to control the runoff coefficient below 0.4 and shorten the drainage time to within 30 minutes, ensuring that there are no waterlogged areas on the site. During the design process, the ecological sensitivity analysis results of S1 are called in real time. If the building model intrudes into a high-sensitivity area, the system will automatically prompt to adjust the position. If the plant configuration area belongs to a high-sensitivity area, it will be replaced with ecologically friendly vegetation. After adjustment, the collaborative conflict detection and simulation verification will be re-executed until all indicators meet the requirements. Finally, an integrated BIM model containing information on various professional components will be formed. The data format is IFC4.0, which will serve as the digital basis for the implementation of S3 construction.

[0013] In a preferred embodiment, step S3 employs a plant growth digital twin algorithm driven by a physical-data hybrid approach. Specifically, the initial parameters of the target plant are first extracted from the BIM design model in S2—taking ginkgo trees in the scenic area as an example, basic information such as seedling age (5 years), initial height (3 meters), crown width (1.5 meters), and genetic growth potential (0.5% per week) can be obtained. Next, real-time site environmental data collected by the GIS system is accessed, including daily average light intensity, daily precipitation, soil organic matter content, and soil pH. Then, a physical sub-model of plant growth is constructed. For example, the amount of organic matter produced by photosynthesis is calculated based on light intensity, the balance between water absorption and transpiration is simulated by combining precipitation, nutrient conversion efficiency is analyzed based on soil organic matter content, and a random forest model is trained using historical ginkgo growth data from the past three years in the scenic area to fit the prediction results of the physical model. The deviation from the actual growth status is used to create a physical-data hybrid-driven digital twin model. Then, IoT sensors installed near the ginkgo tree collect real-time data on its current height, crown width, number of leaves, and chlorophyll content. This data is compared with the twin model's predictions, and the deviation is calculated. A feedback mechanism dynamically adjusts parameters such as environmental factor weights and genetic growth potential coefficients in the model, ensuring the twin model accurately reflects the actual plant's growth trend. Finally, based on real-time updated environmental data and adjusted parameters, the twin model predicts changes in the ginkgo's growth status over the next three months, including a height increase to 3.6 meters, crown width expansion to 1.8 meters, and the timing of autumn leaf fall. It also analyzes the risk of leaf spot disease that may occur under recent continuous high humidity conditions, generating targeted intelligent control suggestions, such as reducing irrigation by one time per week and applying protective fungicides in advance. The formula for predicting the relative growth rate of plants is as follows: In the formula: RGR represents the relative growth rate of a plant, expressed as a percentage increase in height per week. α represents the comprehensive weighting coefficient of environmental factors, with a value ranging from 0 to 1, and is obtained by fitting historical growth data; I represents the site illumination intensity monitored in real time by GIS, in μmol·m⁻²·s⁻¹; I0 represents the optimal light intensity for this plant species, derived from a plant species database; P represents the daily precipitation monitored in real time by GIS, in mm; P0 represents the optimal daily precipitation for this plant species, derived from a plant species database; OM represents the soil organic matter content monitored in real time by GIS, in g / kg; OM0 indicates the optimal soil organic matter content for this plant species, derived from a plant species database; β represents the genetic factor weighting coefficient, which ranges from 0 to 1 and is determined by the genetic characteristics of the plant species. G represents the plant's genetic growth potential, expressed as a percentage increase in height per week; ε represents the random error term, ranging from -0.02 to 0.02, which represents the impact of unpredictable environmental fluctuations. This formula organically integrates real-time GIS environmental data with plant genetic parameters in the BIM model. It quantifies the suitability of environmental factors for growth through relative values, while introducing a random error term to adapt to the uncertainty of actual growth. This enables it to accurately predict plant growth rates and provides core calculation basis for twin models.

[0014] In a preferred embodiment, step S2, the specific process of using the random forest algorithm to simulate plant growth and generate maintenance suggestions, includes: extracting basic parameters such as the species, initial seedling age, current height, and crown width of the target plant from the BIM model in S2; accessing real-time data collected from the GIS system, such as cumulative solar radiation, weekly precipitation, soil organic matter content, surface soil moisture, and daily average temperature; and then importing the scenic area's plant maintenance history records for the past three years. Next, the data is processed by feature analysis, classifying continuous environmental data into categories such as "sufficient / moderate / insufficient," and converting plant state parameters into relative growth parameters. The model is trained using the growth rate, forming an input set containing 12 key information items. A random forest model is trained using the input set from historical data and corresponding result labels, with adjustments made to the number and depth of decision trees to improve accuracy. Real-time data is input into the model to predict the plant's growth stage, crown width growth, and the probability of pests and diseases in the next 15 days. Based on the predictions, maintenance suggestions are generated: if the probability of pests and diseases exceeds 65%, a corresponding fungicide spraying plan and optimal time are recommended; if the plant's daily water requirement exceeds the current soil moisture content, the amount of water required for a single irrigation is calculated; if the crown width growth is too rapid and affects the landscape, pruning within 10 days is recommended, with a height range provided. The model is iteratively updated monthly using maintenance effect data from the S4 maintenance platform, and the results are synchronized to the maintenance system for reference. The formula for calculating dynamic irrigation volume is: ; In the formula: I indicates the recommended irrigation volume per irrigation, in liters per square meter; D pred This represents the daily water requirement of plants predicted by the random forest model, expressed in liters per square meter. S current This indicates the current moisture content of the topsoil as monitored in real time by GIS, expressed in liters per square meter. K represents the soil permeability coefficient, which is dimensionless and ranges from 0.7 to 1.0. It is determined by GIS soil type data. P forecastThis indicates the GIS-predicted precipitation for the next 24 hours, expressed in liters per square meter. L factor This represents the illumination correction factor, which is dimensionless and takes values ​​of 1.2, 1.0, or 0.8, determined by the GIS illumination data. This formula deeply integrates machine learning predictions with GIS environmental data, which not only meets the water needs of plant growth but also avoids water waste. It breaks through the blindness of traditional fixed irrigation amounts and embodies the concept of data-driven precision maintenance.

[0015] In a preferred embodiment, in step S2, the Kalman filter algorithm is used to compare the twin model with the actual plant growth status and optimize the control strategy. The specific operation is as follows: For cherry trees in the scenic area, the initial seedling age, basal diameter, and other parameters are first extracted from the BIM design model in S2. The plant growth twin model constructed in S3 is used to predict the height growth, leaf sprouting number, and nutrient requirements in the next 15 days. At the same time, the actual growth data is collected in real time using a laser rangefinder and soil nutrient sensor installed next to the cherry trees. The predicted values ​​of the twin model and the actual monitoring values ​​are input into the Kalman filter model to calculate the state deviation between the two. The key parameters in the twin model are adjusted based on the filter gain. If the deviation exceeds a preset threshold of 5%, the growth simulation equation of the twin model is re-optimized, and new maintenance and control suggestions are generated. The corrected model parameters are synchronized to the intelligent operation and maintenance platform in S4 for maintenance personnel to execute. The above comparison and optimization process is repeated every 7 days using the latest actual monitoring data to ensure that the twin model always accurately reflects the real growth status of the plant and the control strategy is more targeted. The dynamic correction formula for the growth rate coefficient of the twin model is as follows: ; In the formula: β t+1 This represents the corrected plant growth rate coefficient for the next cycle. β t This represents the growth rate coefficient in the current cycle twin model; λ represents the learning rate, which is adjusted by parameters. H act This indicates the actual monitored plant height; H pre This represents the plant height predicted by the twin model; W t Indicates the time decay weight; This formula organically combines the actual and predicted growth deviations, learning rate, and time weights to dynamically correct the core parameters of the twin model, avoiding the rigidity problem of traditional fixed-parameter models and making the regulation strategy more in line with the actual growth needs of plants.

[0016] In a preferred embodiment, step S4 involves using a weighted comprehensive scoring algorithm to conduct a priority analysis for scenic area updates. The specific process is as follows: First, basic data such as the design life, years of use, and maintenance costs of the facilities to be evaluated are extracted from the BIM model. Simultaneously, spatial data such as the average annual visitor flow, the percentage of degraded plant communities, and the distance to ecologically sensitive areas are obtained through a GIS platform. Then, four core evaluation dimensions are determined: facility aging degree, visitor flow pressure, ecological impact degree, and maintenance cost-benefit ratio. Weights are allocated based on the scenic area management objectives: facility aging 30%, visitor flow 25%, ecological impact 25%, and cost-benefit ratio 20%.

[0017] In a preferred embodiment, in step S4, the data for each dimension are then standardized: the facility aging score is calculated by dividing the years of use by the design life; the visitor flow pressure score is obtained by dividing the average annual visitor flow by the facility's design carrying capacity; the ecological impact score is the proportion of the plant community degradation area to the total design area; and the maintenance cost-benefit ratio is obtained by dividing the expected increase in visitor satisfaction after maintenance by the total maintenance cost. Then, the standardized scores for each dimension are multiplied by their corresponding weights and summed to obtain the priority comprehensive score for each item to be updated. Finally, the scores are sorted from highest to lowest. The highest-scoring walkways are recommended to be replaced with non-slip, wear-resistant permeable concrete; the second-highest-scoring degraded plant communities are recommended to be replanted with native shade-tolerant species such as *Liriope muscari* and *Iris*; and the lower-scoring seats are recommended for partial repair rather than complete replacement. This algorithm adjusts the weight coefficients monthly based on the update implementation effect data from S5, and the results are synchronized to the scenic area management system for decision-making reference. The formula for the comprehensive score of scenic area update priority is: ; In the formula: S represents the overall priority score of the item to be updated; W1 represents the weight of the facility aging dimension; Y indicates the number of years the facilities extracted from BIM have been in use; Y0 represents the facility design life extracted from BIM; W2 represents the weight of the pedestrian flow pressure dimension; F represents the average annual pedestrian flow of the facility extracted by GIS; F0 indicates the facility's designed capacity to accommodate a certain number of people. W3 represents the weight of the ecological impact dimension; A represents the area of ​​degraded plant community extracted by GIS; A0 represents the total designed area of ​​the plant community; W4 represents the weight of the maintenance cost-benefit ratio dimension; B represents the expected benefit value after maintenance; C represents the total cost required for maintenance.

[0018] In a preferred embodiment, in step S5, the ecological sensitivity analysis results of S1, the parameters of various professional design models of S2, the digital twin records of plant growth of S3, and the operation and maintenance feedback data of S4 are first extracted through the BIM-GIS platform interface and uniformly stored in a relational database. The data format is converted to JSON to ensure field alignment, and the update frequency is set to once a week. Next, the evaluation dimensions are determined, including the degree of facility aging, the scope of ecological impact, the cost-effectiveness of operation and maintenance, and visitor experience feedback. Specific indicators are set for each dimension: the degree of facility aging is the ratio of the years of use to the design life; the scope of ecological impact is the proportion of degraded areas; the cost-effectiveness of operation and maintenance is the ratio of the improvement in satisfaction after maintenance to the cost; and the visitor experience feedback is the monthly complaint rate. The weights of each dimension are adjusted according to the annual goals of the scenic area: facility aging weight 0.3, ecological impact weight 0.25, cost-effectiveness weight 0.2, and visitor experience weight 0.25. Then, the data of each indicator are standardized and input into the fuzzy comprehensive evaluation model to calculate the membership value of each item to be updated. The priority list is obtained by sorting the membership values ​​from high to low. The system matches the highest-priority item to a pre-defined solution library, which contains solutions for different problems. The selection of solutions is based on existing site resources and budget constraints. The generated solutions are imported into a BIM model to simulate changes in facility status, plant growth trends, and adjustments to visitor flow after implementation. This verifies whether the solution meets the requirements of ecologically sensitive areas, satisfies visitor capacity, and remains within budget. Details of the solution are adjusted in real-time during the simulation until all verification indicators are met. Finally, the final solution data is synchronized to the S6 resource cycle database, including a list of materials required for implementation, labor costs, and expected results, for subsequent resource allocation.

[0019] In a preferred embodiment, in step S6, the following data is first extracted through a full lifecycle data interface: the wear and tear of hard materials (segments S1-S5), the survival rate and growth potential of healthy plants, and the priority of resource utilization in soil degradation recycling. The criteria layer includes resource availability, ecological value, and recycling cost, with weights of 0.4, 0.35, and 0.25, respectively. The scheme layer covers different types of hard materials and healthy plants. Each resource item is scored according to the criteria layer indicators: hard material availability is scored 9 points based on wear and tear less than 30%; ecological value is scored 8 points based on whether it conforms to the local ecosystem; and recycling cost is scored 7 points based on transportation distance less than 2 kilometers. Healthy plant availability is scored 9 points based on survival rate greater than 90%; ecological value is scored 8 points based on whether it is a native species; and recycling cost is scored 7 points based on low transplantation difficulty. The comprehensive score of each resource item is calculated and sorted to determine the list of recyclable resources. Next, a resource distribution map was created using GIS spatial analysis functions. The location coordinates of the resource list were imported, and different colors were used to mark resources according to type and priority. The map resolution was set to 5 meters, and information such as resource quantity and availability level was included to visually present the resource distribution. Subsequently, an ecological restoration plan was developed based on the initial ecological baseline data of S1. For degraded soil improvement, the dosage of amendment was determined based on the difference between the initial and current soil pH values. For native plant community reconstruction, suitable species were selected and planting density was determined based on the initial vegetation distribution to ensure the restoration plan was consistent with the initial ecological baseline. Finally, all data was organized into a standardized format, including resource assessment results, restoration plan details, and implementation effect records, and archived in an industry database with tiered access permissions. This provides experience and reference for the construction of other gardens and scenic areas, and is simultaneously updated to the resource recycling database.

[0020] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. In this invention, by integrating BIM plant design models with GIS real-time environmental monitoring data, a digital twin system is constructed that can dynamically simulate plant growth status. Based on real-time soil moisture and plant growth stages, it calculates appropriate irrigation amounts, avoiding the problems of over-irrigation or under-irrigation in traditional maintenance. Combining long-term accumulated growth data and environmental parameters, the system can predict plant growth needs and potential risks in advance, generating maintenance suggestions tailored to the actual site conditions, ensuring that maintenance measures are always adapted to the current state of the plants. Simultaneously, by integrating multi-dimensional factors, it provides an objective update ranking, directing resources towards the most urgently needed aspects, improving resource utilization efficiency while reducing operation and maintenance costs.

[0021] 2. In this invention, the dynamic simulation results of plant growth status and precise irrigation data are synchronized to subsequent operation and maintenance and update stages, providing accurate growth records and environmental parameter support for update decisions, ensuring seamless decision-making across all stages. The precise data generated from the prediction of relative plant growth rates and the calculation of dynamic irrigation volumes are also synchronized to subsequent operation and maintenance and update stages, providing accurate growth records and environmental parameter support for the comprehensive scoring formula for scenic area update priorities, ensuring seamless decision-making across all stages. Through computational linkage, the stability of plant communities can be better maintained, ecological degradation issues can be prioritized, and the initial ecological baseline of the site can be protected, thus enabling the system to better maintain the stability of plant communities. Attached Figure Description

[0022] Figure 1 This is a schematic diagram illustrating the process principle of the present invention. Detailed Implementation

[0023] 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.

[0024] Example: Refer to Figure 1 A collaborative management method for the entire lifecycle of landscape architecture based on BIM and GIS, characterized by the following steps: S1: Site Ecology and Spatial Foundation Assessment Based on BIM-GIS Integration: Data on site topography, hydrology, vegetation status, and surrounding ecologically sensitive areas were collected using GIS to establish a site geographic information database. Simultaneously, a three-dimensional initial site model was constructed using BIM. The GIS data was then spatially overlaid with the BIM model to conduct assessments including ecological sensitivity analysis, sight corridor simulation, and terrain slope optimization. The results of this assessment will directly serve as the core constraints for the S2 scheme design, ensuring that the design scheme meets the site's ecological baseline requirements.

[0025] S2: Cross-disciplinary collaborative solution design and dynamic verification: A team of professionals specializing in landscape architecture, architecture, water supply and drainage, and vegetation was organized to conduct collaborative BIM design based on the site data from S1, constructing detailed models for each discipline. Regional climate, soil type, and hydrological cycle data from GIS were imported to verify the climate adaptability of the plant configuration scheme and to optimize the drainage system of the hardscape through storm runoff simulation. During the design process, the ecological assessment results from S1 could be used in real time to adjust the scheme. The final integrated BIM model will serve as the digital basis for the construction implementation of S3.

[0026] S3: Digital Twin and Intelligent Regulation of the Entire Plant Growth Life Cycle: Based on the S2 plant design model, and combined with real-time GIS monitoring of site light, precipitation, and soil fertility data, a digital twin system for plant growth is constructed. Machine learning algorithms simulate the plant's growth cycle, morphological changes, and pest and disease risks under different environmental conditions, dynamically generating maintenance recommendations (including pruning time and irrigation amount). Simultaneously, the twin model is compared with the actual plant growth status to continuously optimize control strategies. The digital twin data from this step is synchronously pushed to the S4 operation and maintenance management platform, providing precise guidance for daily maintenance.

[0027] S4: BIM-GIS Integration for Smart Operation and Asset Management: The S2 BIM model is integrated with the scenic area's IoT devices (including cameras and environmental sensors) to visualize facility operation status (including street light malfunctions and damaged seats) and environmental parameters (including PM2.5, temperature, and humidity) through GIS maps. Combined with the plant growth twin data from S3, personalized maintenance plans are developed. Simultaneously, GIS-based visitor flow heatmap analysis optimizes scenic area routes and facility layout. The maintenance data from this step will be archived regularly to provide data support for the S5 scenic area updates.

[0028] S5: Data-driven scenic area renewal decisions throughout the entire lifecycle: Integrate all data from S1 to S4 (including site assessment, design schemes, plant growth records, and operation and maintenance feedback) and conduct multi-dimensional analysis through the BIM-GIS data platform; generate update priority rankings and proposed solutions for existing problems in the scenic area; the update schemes can be pre-simulated using the BIM model to verify their feasibility before implementation. The update schemes in this step will be synchronously updated to the resource recycling database in S6.

[0029] S6: Ecological Restoration and Resource Recycling Planning after the Retirement of Scenic Areas When a scenic area enters the decommissioning phase, the recyclable resources within the site are assessed based on full life-cycle data, and a resource distribution map is drawn using GIS. Combined with the initial ecological baseline data of S1, an ecological restoration plan for the site is formulated, such as the improvement of degraded soil and the reconstruction of native plant communities. At the same time, all data is archived into an industry database to provide experience and reference for the construction of other garden scenic areas.

[0030] In step S1, a workflow combining ArcGIS Pro and Revit is used to advance step S1. First, oblique photogrammetry of the site is used to collect site topographic data, with a resolution of 5 cm and a collection frequency of once per hour. Hydrological data is collected using automatic rain gauges and water level sensors, with a sampling interval of 15 minutes. Vegetation status data is obtained by combining satellite remote sensing imagery with ground surveys, with an image resolution of 1 meter and a survey coverage of 100%. Data on surrounding ecologically sensitive areas is synchronized from the local ecological and environmental bureau database, with an update cycle of once per day. When establishing the site geographic information database, the coordinate system adopts the 2000 National Geodetic Coordinate System, and the elevation datum is the 1985 National Elevation Datum. Then, a 3D initial model of the site is constructed using Revit, including elements such as topography, existing buildings, structures, and vegetation distribution. The topographic model imports DEM data collected by GIS. The vegetation model is modeled according to actual species and locations. The model coordinate system is kept consistent with the GIS database. Then, FME software is used to convert and integrate GIS and BIM data. Spatial overlay is completed using the coordinate matching method to correspond one-to-one between the 3D coordinates of the BIM model and the spatial coordinates of the GIS database. After overlay, an integrated digital twin model of the site is generated.

[0031] In step S2, all professional models are uniformly imported into the Navisworks platform, collaborative permissions are set, and modifications from each profession are synchronized in real time. The model accuracy is maintained at LOD350, conflict detection is performed hourly, and conflict points are automatically marked when the component spacing is less than 10 cm. Subsequently, the average temperature, annual precipitation, and soil pH data of the region over the past 10 years provided by the GIS system are imported. The Ecotect Analysis model is used to input parameters such as plant growth temperature range, water requirement, and suitable soil pH to simulate the growth cycle and survival rate of plants on the site. A survival rate threshold of 90% is set, and plant species below the threshold are automatically marked and replaced with native plants that meet the regional climate conditions. Next, the SWMM model is used to import the rainstorm intensity formula, site impervious area ratio, and soil permeability coefficient data from the GIS to simulate the runoff depth, peak flow, and drainage time of a 50-year return period rainstorm. The diameter and slope of the drainage network are optimized to control the runoff coefficient below 0.4 and shorten the drainage time to within 30 minutes, ensuring that there are no waterlogged areas on the site. During the design process, the ecological sensitivity analysis results of S1 are called in real time. If the building model intrudes into a high-sensitivity area, the system will automatically prompt to adjust the position. If the plant configuration area belongs to a high-sensitivity area, it will be replaced with ecologically friendly vegetation. After adjustment, the collaborative conflict detection and simulation verification will be re-executed until all indicators meet the requirements. Finally, an integrated BIM model containing information on various professional components will be formed. The data format is IFC4.0, which will serve as the digital basis for the implementation of S3 construction.

[0032] In step S3, a digital twin algorithm for plant growth driven by a physics-data hybrid approach is adopted. Specifically, the initial parameters of the target plant are first extracted from the BIM design model in S2—taking ginkgo trees in the scenic area as an example, basic information such as seedling age (5 years), initial height (3 meters), crown width (1.5 meters), and genetic growth potential (0.5% per week) can be obtained. Next, real-time site environmental data collected by the GIS system is accessed, including daily average light intensity, daily precipitation, soil organic matter content, and soil pH. Then, a physical sub-model of plant growth is constructed. For example, the amount of organic matter produced by photosynthesis is calculated based on light intensity, the balance between water absorption and transpiration is simulated by combining precipitation, nutrient conversion efficiency is analyzed based on soil organic matter content, and a random forest model is trained using the past three years' historical ginkgo growth data of the scenic area to fit the physical model's predictions with the actual growth status. To address the discrepancies, a physical-data hybrid-driven digital twin model is created. Then, IoT sensors installed near the ginkgo tree collect real-time data on its actual growth, including height, crown width, number of leaves, and chlorophyll content. This data is compared with the twin model's predictions, and the deviation is calculated. A feedback mechanism dynamically adjusts parameters such as environmental factor weights and genetic growth potential coefficients in the model, ensuring the twin model accurately reflects the actual plant's growth trend. Finally, based on real-time updated environmental data and adjusted parameters, the twin model predicts changes in the ginkgo's growth status over the next three months, including a height increase to 3.6 meters, crown width expansion to 1.8 meters, and the timing of autumn leaf fall. It also analyzes the risk of leaf spot disease that may occur under recent high humidity conditions, generating targeted intelligent control suggestions, such as reducing irrigation by one time per week and applying protective fungicides in advance. The formula for predicting plant relative growth rate (RGR) is as follows: In the formula: RGR represents the relative growth rate of a plant, expressed as a percentage increase in height per week. α represents the comprehensive weighting coefficient of environmental factors, with a value ranging from 0 to 1, and is obtained by fitting historical growth data; I represents the site illumination intensity monitored in real time by GIS, in μmol·m⁻²·s⁻¹; I0 represents the optimal light intensity for this plant species, derived from a plant species database (determined in conjunction with plant species in the BIM model). P represents the daily precipitation monitored in real time by GIS, in mm; P0 represents the optimal daily precipitation for this plant species, derived from a plant species database; OM represents the soil organic matter content monitored in real time by GIS, in g / kg; OM0 indicates the optimal soil organic matter content for this plant species, derived from a plant species database; β represents the genetic factor weighting coefficient, which ranges from 0 to 1 and is determined by the genetic characteristics of the plant species (extracted from the BIM model). G represents the genetic growth potential of the plant, expressed as a percentage increase in height per week (determined based on seedling age and species in the BIM model). ε represents the random error term, ranging from -0.02 to 0.02, which represents the impact of unpredictable environmental fluctuations. This formula organically integrates real-time GIS environmental data with plant genetic parameters in the BIM model. It quantifies the suitability of environmental factors for growth through relative values, while introducing a random error term to adapt to the uncertainty of actual growth. This enables it to accurately predict plant growth rates and provides core calculation basis for twin models.

[0033] In step S2, the specific process of using the random forest algorithm to simulate plant growth and generate maintenance suggestions includes: extracting basic parameters such as the type, initial seedling age, current height, and crown width of the target plants from the BIM model in S2; accessing real-time data collected from the GIS system, such as cumulative solar radiation, weekly precipitation, soil organic matter content, surface soil moisture, and daily average temperature; importing the scenic area's plant maintenance history records for the past three years (including the occurrence time, type, control measures and effects of pests and diseases, and the specific time and amount of pruning and irrigation); then, performing feature processing on the data, classifying continuous environmental data into categories such as "sufficient / moderate / insufficient," and further... Plant state parameters are converted into relative growth rates, forming an input set containing 12 key information items. A random forest model is trained using the input set from historical data and corresponding result labels, with the number and depth of decision trees optimized to improve accuracy. Real-time data is input into the model to predict the plant's growth stage, crown growth value, and the probability of pests and diseases in the next 15 days. Maintenance suggestions are generated based on the predictions: if the probability of pests and diseases exceeds 65%, a corresponding fungicide spraying plan and optimal time are recommended; if the plant's daily water requirement exceeds the current soil moisture content, the amount of irrigation required per application is calculated; if the crown growth is too rapid and affects the landscape, pruning within 10 days is recommended, with a height range provided. The model is iteratively updated monthly using maintenance effect data from the S4 maintenance platform, and the results are synchronized to the maintenance system for reference. The formula for calculating dynamic irrigation volume is: ; In the formula: I indicates the recommended irrigation volume per irrigation, in liters per square meter; D pred This represents the daily water requirement of plants predicted by the random forest model, expressed in liters per square meter. S current This indicates the current moisture content of the topsoil as monitored in real time by GIS, expressed in liters per square meter. K represents the soil permeability coefficient, which is dimensionless and ranges from 0.7 to 1.0. It is determined by GIS soil type data. P forecast This indicates the GIS-predicted precipitation for the next 24 hours, expressed in liters per square meter. L factor This represents the illumination correction factor, which is dimensionless and takes values ​​of 1.2 (sufficient illumination), 1.0 (moderate illumination), or 0.8 (insufficient illumination), determined by the GIS illumination data. This formula deeply integrates machine learning predictions with GIS environmental data, which not only meets the water needs of plant growth but also avoids water waste. It breaks through the blindness of traditional fixed irrigation amounts and embodies the concept of data-driven precision maintenance.

[0034] In step S2, the Kalman filter algorithm is used to compare the twin model with the actual plant growth status and optimize the control strategy. The specific operation is as follows: For the cherry trees in the scenic area, the initial seedling age, basal diameter and other parameters are extracted from the BIM design model in S2. The plant growth twin model constructed in S3 is used to predict the height growth, number of leaves sprouting and nutrient requirements in the next 15 days. At the same time, the actual growth data is collected in real time using a laser rangefinder and soil nutrient sensor installed next to the cherry trees. The predicted value of the twin model and the actual monitoring value are input into the Kalman filter model to calculate the state deviation between the two. The key parameters in the twin model are adjusted based on the filter gain. If the deviation exceeds the preset threshold of 5%, the growth simulation equation of the twin model is re-optimized and new maintenance control suggestions are generated. The corrected model parameters are synchronized to the smart operation and maintenance platform in S4 for maintenance personnel to execute. The above comparison and optimization process is repeated every 7 days using the latest actual monitoring data to ensure that the twin model always accurately reflects the real growth status of the plants and the control strategy is more targeted. The dynamic correction formula for the growth rate coefficient of the twin model is as follows: ; In the formula: β t+1 This represents the corrected plant growth rate coefficient for the next cycle (unit: cm / day). β t This represents the growth rate coefficient in the current cycle twin model; λ represents the parameter adjustment learning rate (dimensionless, ranging from 0.02 to 0.08, determined iteratively based on historical optimization results). H act Indicates the actual monitored plant height (unit: cm); H pre Indicates the plant height predicted by the twin model (unit: cm); W tThis represents the time decay weight (dimensionless, ranging from 0.9 to 1.1, with more recent data having a higher weight). This formula organically combines the actual and predicted growth deviations, learning rate, and time weights to dynamically correct the core parameters of the twin model, avoiding the rigidity problem of traditional fixed-parameter models and making the regulation strategy more in line with the actual growth needs of plants.

[0035] In step S4, a weighted comprehensive scoring algorithm is used to conduct a priority analysis of scenic area updates. The specific process is as follows: First, basic data such as the design life, years of use, and maintenance costs of the facilities to be evaluated (such as walkways, plant communities, and leisure seats) are extracted from the BIM model. At the same time, spatial data such as the average annual traffic flow, the percentage of plant community degradation area, and the distance to ecologically sensitive areas in the areas where each facility is located are obtained through the GIS platform. Then, four core evaluation dimensions are determined: facility aging degree, traffic flow pressure, ecological impact degree, and maintenance cost-benefit ratio. Weights are allocated based on the scenic area management objectives: facility aging 30%, traffic flow 25%, ecological impact 25%, and cost-benefit ratio 20%.

[0036] In step S4, the data for each dimension is then standardized: the facility aging score is calculated by dividing the years of use by the design life; the visitor flow pressure score is obtained by dividing the average annual visitor flow by the facility's design capacity; the ecological impact score is the proportion of degraded plant community area to the total design area; and the maintenance cost-benefit ratio is obtained by dividing the expected increase in visitor satisfaction after maintenance by the total maintenance cost. The standardized scores for each dimension are then multiplied by their corresponding weights and summed to obtain the priority comprehensive score for each item to be updated. Finally, the scores are sorted from highest to lowest. The highest-scoring walkways are prioritized for replacement with non-slip, wear-resistant permeable concrete; the second-highest-scoring degraded plant communities are recommended for replanting native shade-tolerant species such as *Liriope muscari* and *Iris*; and the lowest-scoring benches are recommended for partial repair rather than complete replacement. This algorithm adjusts the weight coefficients monthly based on the update implementation data from S5, and the results are synchronized to the scenic area management system for decision-making reference. The formula for the comprehensive score of scenic area update priority is: ; In the formula: S represents the priority score of the item to be updated (dimensionless, range 0~1). W1 represents the weight of the facility aging dimension (dimensionless, value 0.3). Y represents the number of years the facilities extracted from BIM have been in use (in years); Y0 represents the facility design life (in years) extracted from BIM; W2 represents the weight of the pedestrian flow pressure dimension (dimensionless, value 0.25). F represents the average annual passenger flow of the facility extracted by GIS (unit: person-times / year). F0 indicates the facility's designed capacity for passenger flow (unit: person-times / year). W3 represents the weight of the ecological impact dimension (dimensionless, value 0.25). A represents the area of ​​degraded plant community extracted by GIS (unit: square meters); A0 represents the total designed area of ​​the plant community (unit: square meters); W4 represents the maintenance cost-benefit ratio dimension weight (dimensionless, value 0.2). B represents the expected benefits after maintenance (unit: percentage increase in visitor satisfaction). C represents the total cost of maintenance (unit: 10,000 yuan). In step S5, the ecological sensitivity analysis results from S1, the parameters of various professional design models from S2, the digital twin records of plant growth from S3, and the operation and maintenance feedback data from S4 are first extracted through the BIM-GIS platform interface and stored uniformly in a relational database. The data format is converted to JSON to ensure field alignment, and the update frequency is set to once a week. Next, the evaluation dimensions are determined, including facility aging degree, ecological impact range, operation and maintenance cost-effectiveness, and visitor experience feedback. Specific indicators are set for each dimension: facility aging degree is the ratio of the years of use to the design life; ecological impact range is the proportion of degraded areas; operation and maintenance cost-effectiveness is the ratio of the improvement in satisfaction after maintenance to the cost; and visitor experience feedback is the monthly complaint rate. The weights of each dimension are adjusted according to the scenic area's annual goals: facility aging weight 0.3, ecological impact weight 0.25, cost-effectiveness weight 0.2, and visitor experience weight 0.25. Then, the data of each indicator is standardized and input into the fuzzy comprehensive evaluation model to calculate the membership value of each item to be updated. The items are then sorted from high to low membership degree to obtain a priority list. The system matches the highest-priority item to a pre-defined solution library, which contains solutions for different problems. The selection of solutions is based on existing site resources and budget constraints. The generated solutions are imported into a BIM model to simulate changes in facility status, plant growth trends, and adjustments to visitor flow after implementation. This verifies whether the solution meets the requirements of ecologically sensitive areas, satisfies visitor capacity, and remains within budget. Details of the solution are adjusted in real-time during the simulation until all verification indicators are met. Finally, the final solution data is synchronized to the S6 resource cycle database, including a list of materials required for implementation, labor costs, and expected results, for subsequent resource allocation.

[0037] In step S6, data on the wear and tear of hard materials (service lifespan), the survival rate and growth potential of healthy plants, and the degree of soil degradation (S1-S5) are first extracted through the full life-cycle data interface and uniformly converted into a structured format for storage. The analytic hierarchy process (AHP) is used, with the target layer prioritizing recyclable resources, the criteria layer including resource availability, ecological value, and recycling cost (weights of 0.4, 0.35, and 0.25 respectively), and the solution layer covering different types of hard materials and healthy plants. Each resource item is scored according to the criteria layer indicators: hard material availability is scored 9 points based on wear less than 30%, ecological value is scored 8 points based on compatibility with the local ecosystem, and recycling cost is scored 7 points based on transportation distance less than 2 kilometers; healthy plant availability is scored 9 points based on survival rate greater than 90%, ecological value is scored 8 points based on whether it is a native species, and recycling cost is scored 7 points based on ease of transplantation. The comprehensive score for each resource item is calculated and ranked to determine the recyclable resource list. Next, a resource distribution map was created using GIS spatial analysis functions. The location coordinates of the resource list were imported, and different colors were used to mark resources according to type and priority. The map resolution was set to 5 meters, and information such as resource quantity and availability level was included to visually present the resource distribution. Subsequently, an ecological restoration plan was developed based on the initial ecological baseline data of S1. For degraded soil improvement, the dosage of amendment was determined based on the difference between the initial and current soil pH values. For native plant community reconstruction, suitable species were selected and planting density was determined based on the initial vegetation distribution to ensure the restoration plan was consistent with the initial ecological baseline. Finally, all data was organized into a standardized format, including resource assessment results, restoration plan details, and implementation effect records, and archived in an industry database with tiered access permissions. This provides experience and reference for the construction of other gardens and scenic areas, and is simultaneously updated to the resource recycling database.

[0038] From the above, we can conclude that: In this invention, a digital twin system is constructed by integrating BIM plant design models with real-time GIS environmental monitoring data. This system can dynamically simulate plant growth status and calculate appropriate irrigation amounts based on real-time soil moisture and plant growth stages, avoiding the problems of over-irrigation or under-irrigation in traditional maintenance. Combining long-term accumulated growth data and environmental parameters, the system can predict plant growth needs and potential risks in advance, generating maintenance recommendations tailored to the actual site conditions, ensuring that maintenance measures are always adapted to the current state of the plants. Simultaneously, by integrating multi-dimensional factors, it provides an objective update ranking, directing resources towards the most urgently needed aspects, improving resource utilization efficiency while reducing operation and maintenance costs.

[0039] In this invention, the dynamic simulation results of plant growth status and precise irrigation data are synchronized to subsequent operation and maintenance and update stages, providing accurate growth records and environmental parameter support for update decisions, ensuring seamless decision-making across all stages. Precise data generated from plant relative growth rate prediction and dynamic irrigation volume calculation are also synchronized to subsequent operation and maintenance and update stages, providing accurate growth records and environmental parameter support for the comprehensive scoring formula for scenic area update priorities, ensuring seamless decision-making across all stages. Through computational linkage, the stability of plant communities can be better maintained, ecological degradation issues can be prioritized, and the initial ecological baseline of the site can be protected, thus enabling the system to better maintain the stability of plant communities.

[0040] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the term "comprising" or any other variations thereof is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0041] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A collaborative management method for the entire lifecycle of a landscape architecture based on BIM and GIS, characterized by: The method includes the following steps: S1: Integrate site ecological data collected by GIS with the initial BIM 3D model to carry out ecological sensitivity classification, visual corridor simulation, and terrain slope optimization, and output clear design constraints; S2: Organize the landscape and architecture teams to conduct BIM collaborative design based on S1 data, import GIS regional climate and soil type data to verify the adaptability of plant configuration, simulate rainstorm runoff to optimize the drainage system, and generate an integrated BIM model with construction parameters. S3: Combine BIM plant models with real-time GIS environmental data to build a digital twin system, simulate plant growth cycles, morphological changes and pest and disease risks, dynamically generate maintenance suggestions, and synchronize the data to the operation and maintenance platform; S4: Connect the BIM model with IoT devices, visualize facility status and environmental parameters through GIS, develop personalized maintenance plans by combining twin data, optimize visitor routes using heat maps of pedestrian flow, and archive operation and maintenance data to support updates; S5: Integrate full life cycle data, analyze facility aging and plant health issues from multiple dimensions, generate update priorities and solutions, and implement them after BIM pre-simulation verification; S6: After the scenic area is decommissioned, assess the recyclable resources and draw a distribution map using GIS. Combine the initial ecological data to formulate a restoration plan, and archive the data for industry reference.

2. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S1, the workflow of ArcGIS Pro and Revit is used to advance step S1. First, the site topographic data is collected by using UAV oblique photogrammetry with a resolution of 5 cm and a collection frequency of once per hour. Hydrological data is collected by using automatic rain gauges and water level sensors with a sampling interval of 15 minutes. Vegetation status data is obtained by combining satellite remote sensing images with ground surveys with an image resolution of 1 meter and a survey coverage of 100%.

3. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S2, all professional models are uniformly imported into the Navisworks platform to set collaborative permissions, and the modifications of each profession are synchronized in real time. The model accuracy is maintained at LOD350, and conflict detection is performed once per hour. When the spacing between components is less than 10 cm, conflict points are automatically marked. Then, the average annual temperature, precipitation and soil pH data of the region over the past 10 years provided by the GIS system are imported. The parameters of plant growth temperature range, water requirement and suitable soil pH value are input through the Ecotect Analysis model to simulate the growth cycle and survival rate of plants on the site. A survival rate threshold of 90% is set. Plant species below the threshold are automatically marked and replaced with native plants that meet the regional climate conditions. Then, the SWMM model is used to import the rainstorm intensity formula, site impervious area ratio and soil permeability coefficient data from the GIS to simulate the runoff depth, peak flow and drainage time of a 50-year return period rainstorm. The diameter and slope of the drainage network are optimized to control the runoff coefficient below 0.4 and shorten the drainage time to within 30 minutes to ensure that there are no water accumulation areas on the site.

4. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S3, a physical-data hybrid-driven digital twin algorithm for plant growth is adopted. First, the initial parameters of the target plant are extracted from the BIM design model in S2. Then, the site environmental data collected in real time by the GIS system is accessed, including daily average light intensity, daily precipitation, soil organic matter content, and soil pH. Next, a physical sub-model of plant growth is constructed, and a random forest model is trained using the historical growth data of ginkgo trees in the scenic area over the past three years. The deviation between the prediction results of the physical model and the actual growth status is fitted to form a physical-data hybrid-driven digital twin model. Then, the actual growth data of the plant, such as current height, crown width, number of leaves, and chlorophyll content, are collected in real time by IoT sensors installed nearby. These data are compared with the prediction results of the twin model, the deviation value is calculated, and the environmental factor weight coefficient and genetic growth potential coefficient parameters in the model are dynamically adjusted through a feedback mechanism so that the twin model can accurately reflect the actual growth trend of the plant. Finally, based on the real-time updated environmental data and the adjusted parameters, the twin model predicts future changes in growth status and analyzes the risk of leaf spot disease that may occur under the recent continuous high humidity environment, generating targeted intelligent control suggestions. The formula for predicting the relative growth rate of plants is as follows: ； In the formula: RGR represents the relative growth rate of a plant, expressed as a percentage increase in height per week. α represents the comprehensive weighting coefficient of environmental factors, with a value ranging from 0 to 1, and is obtained by fitting historical growth data; I represents the site illumination intensity monitored in real time by GIS, in μmol·m⁻²·s⁻¹; I0 represents the optimal light intensity for this plant species, derived from a plant species database; P represents the daily precipitation monitored in real time by GIS, in mm; P0 represents the optimal daily precipitation for this plant species, derived from a plant species database; OM represents the soil organic matter content monitored in real time by GIS, in g / kg; OM0 indicates the optimal soil organic matter content for this plant species, derived from a plant species database; β represents the genetic factor weighting coefficient, which ranges from 0 to 1 and is determined by the genetic characteristics of the plant species. G represents the plant's genetic growth potential, expressed as a percentage increase in height per week; ε represents the random error term, ranging from -0.02 to 0.02, which represents the impact of unpredictable environmental fluctuations.

5. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S2, the specific process of using the random forest algorithm to simulate plant growth and generate maintenance suggestions includes: extracting the target plant species, initial seedling age, current height, and crown width basic parameters from the BIM model in S2; accessing real-time data collected from the GIS system, such as cumulative solar radiation, weekly precipitation, soil organic matter content, surface soil moisture, and daily average temperature; and then importing the scenic area's plant maintenance history records for the past three years. Next, the data is processed, and continuous environmental data is categorized into "sufficient / moderate / insufficient" types. Plant state parameters are converted into relative growth rates, forming an input set containing 12 key information items. Historical data is then used to... The model is trained using the input set and corresponding result labels. The number and depth of decision trees are optimized to improve accuracy. After real-time data is input into the model, it can predict the plant's growth stage, crown growth value, and the probability of pests and diseases in the next 15 days. Based on the prediction results, maintenance suggestions are generated: if the probability of pests and diseases exceeds 65%, a corresponding fungicide spraying plan and optimal time are pushed; if the plant's daily water requirement is greater than the current soil moisture content, the amount of irrigation for a single irrigation is calculated; if the crown growth is too fast and affects the landscape, pruning is recommended within 10 days and a height range is given. The model is iteratively updated monthly using maintenance effect data from the S4 maintenance platform, and the results are synchronized to the maintenance system for reference. The formula for calculating dynamic irrigation volume is: ； In the formula: I indicates the recommended irrigation volume per irrigation, in liters per square meter; D pred This represents the daily water requirement of plants predicted by the random forest model, expressed in liters per square meter. S current This indicates the current moisture content of the topsoil as monitored in real time by GIS, expressed in liters per square meter. K represents the soil permeability coefficient, which is dimensionless and ranges from 0.7 to 1.

0. It is determined by GIS soil type data. P forecast This indicates the GIS-predicted precipitation for the next 24 hours, expressed in liters per square meter. L factor This represents the illumination correction factor, which is dimensionless and takes values ​​of 1.2, 1.0, or 0.8, determined by the GIS illumination data.

6. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S2, the Kalman filter algorithm is used to compare the twin model with the actual plant growth status and optimize the control strategy. First, the initial seedling age and basal diameter parameters are extracted from the BIM design model in S2. The plant growth twin model constructed in S3 is used to predict the height growth, number of leaves sprouting and nutrient requirements in the next 15 days. At the same time, the actual growth data is collected in real time using a laser rangefinder and soil nutrient sensor installed next to the cherry tree. The predicted values ​​of the twin model and the actual monitoring values ​​are input into the Kalman filter model to calculate the state deviation between the two. The key parameters in the twin model are adjusted based on the filter gain. If the deviation exceeds the preset threshold of 5%, the growth simulation equation of the twin model will be re-optimized, and new maintenance and control suggestions will be generated. The revised model parameters will be synchronized to the S4 smart operation and maintenance platform for maintenance personnel to execute, and the above comparison and optimization process will be repeated every 7 days using the latest actual monitoring data. The dynamic correction formula for the growth rate coefficient of the twin model is as follows: ； In the formula: β t+1 This represents the corrected plant growth rate coefficient for the next cycle. β t This represents the growth rate coefficient in the current cycle twin model; λ represents the learning rate, which is adjusted by parameters. H act This indicates the actual monitored plant height; H pre This represents the plant height predicted by the twin model; W t This represents the time decay weight.

7. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S4, a weighted comprehensive scoring algorithm is used to conduct a priority analysis of scenic area updates. The specific process is as follows: First, the design life, years of use, and maintenance cost data of the facilities to be evaluated are extracted from the BIM model. At the same time, the average annual traffic flow, the proportion of plant community degradation area, and the spatial distance data of ecologically sensitive areas in the area where each facility is located are obtained through the GIS platform. Then, four core evaluation dimensions are determined: the degree of facility aging, traffic pressure, degree of ecological impact, and maintenance cost-benefit ratio.

8. The method for collaborative management of the entire life cycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S4, the data for each dimension are then standardized: the facility aging score is calculated by dividing the years of use by the design life; the visitor flow pressure score is obtained by dividing the average annual visitor flow by the facility's design carrying capacity; the ecological impact score is the proportion of the plant community degradation area to the total design area; and the maintenance cost-benefit ratio is obtained by dividing the expected increase in visitor satisfaction after maintenance by the total maintenance cost. Then, the standardized scores for each dimension are multiplied by their corresponding weights and summed to obtain the priority comprehensive score for each item to be updated. Finally, the scores are sorted from highest to lowest. The highest-scoring walkways are recommended to be replaced with non-slip and wear-resistant permeable concrete; the second-highest-scoring degraded plant communities are recommended to be replanted with native shade-tolerant species such as Liriope and Iris; and the lower-scoring benches are recommended for partial repair rather than complete replacement. This algorithm adjusts the weight coefficients monthly based on the update implementation effect data from S5, and the results are synchronized to the scenic area management system for decision-making reference. The formula for the comprehensive score of scenic area update priority is: ； In the formula: S represents the overall priority score of the item to be updated; W1 represents the weight of the facility aging dimension; Y indicates the number of years the facilities extracted from BIM have been in use; Y0 represents the facility design life extracted from BIM; W2 represents the weight of the pedestrian flow pressure dimension; F represents the average annual pedestrian flow of the facility extracted by GIS; F0 indicates the facility's designed capacity to accommodate a certain number of people. W3 represents the weight of the ecological impact dimension; A represents the area of ​​degraded plant community extracted by GIS; A0 represents the total designed area of ​​the plant community; W4 represents the weight of the maintenance cost-benefit ratio dimension; B represents the expected benefit value after maintenance; C represents the total cost required for maintenance.

9. A collaborative management method for the entire lifecycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S5, the ecological sensitivity analysis results from S1, the parameters of various professional design models from S2, the digital twin records of plant growth from S3, and the operation and maintenance feedback data from S4 are first extracted through the BIM-GIS platform interface and stored uniformly in a relational database. Next, the evaluation dimensions are determined, including facility aging degree, ecological impact range, operation and maintenance cost-effectiveness, and visitor experience feedback. Specific indicators are set for each dimension. The facility aging degree is taken as the ratio of the years of use to the design life, and the ecological impact range is taken as the proportion of degraded areas. Then, the data of each indicator are standardized and input into a fuzzy comprehensive evaluation model to calculate the membership degree value of each item to be updated. A priority list is obtained by sorting the items by membership degree from high to low. A preset solution library is matched according to the item with the highest priority. The solution library contains treatment measures corresponding to different problems, and the selection of measures is combined with the existing resources and budget constraints of the site. The generated solutions are imported into the BIM model to simulate changes in facility status, plant growth trends, and adjustments to visitor flow after implementation, verifying whether the solutions meet the requirements of ecologically sensitive areas, whether they meet the carrying capacity of visitor flow, and whether they are within the budget.

10. A collaborative management method for the entire lifecycle of a landscape architecture based on BIM and GIS as described in claim 1, characterized in that: In step S6, the data on the wear and tear of hard materials, the survival rate and growth potential of healthy plants, and the degree of soil degradation from S1 to S5 are first extracted through the full life cycle data interface and then converted into a structured format for storage.

Images