Garden engineering full life cycle dynamic simulation and collaborative design method and system based on digital twinning and BIM

Abstract

The application discloses a garden engineering full life cycle dynamic simulation and collaborative design method and system based on digital twinning and BIM. The system comprises a data acquisition layer, a unified operation model core engine, a dynamic simulation and safety prediction module, a twinning visualization platform and a real-time early warning execution module. The unified operation model integrates a plant dynamic growth sub-model, a park static environment characteristic sub-model and an external force environment dynamic load sub-model, realizing multi-time scale coupling simulation of plant long-term growth succession and instantaneous external force response. The dynamic simulation module simulates the elastic swing, plastic deformation, branch breakage and overall collapse process of plants under the action of external force (wind, snow), and accurately predicts the damage risk of collapsed plants to road personnel, power lines, buildings, waterways and other infrastructure based on spatial swept volume collision detection and damage criterion. The application realizes the intelligent leap of garden engineering from static display to dynamic prediction, from post-disposal to pre-warning.

Description

Technical Field

[0001] This invention relates to the fields of landscape engineering and digital twin technology, and more specifically, to a method and system for dynamic simulation and collaborative design of the entire life cycle of landscape engineering based on digital twins and BIM. Background Technology

[0002] Existing digital management methods for landscape engineering projects primarily focus on static asset visualization, basic environmental monitoring, or independent plant growth simulation. However, in actual projects, plants within the landscape are "living" elements. Their long-term growth (morphological changes, mechanical performance degradation) and instantaneous external force responses (wind-induced swaying, snow-induced fracture, overall lodging) are strongly coupled. Simultaneously, complex spatial and mechanical interactions exist between the dynamic changes of plants and infrastructure such as roads, power lines, buildings, and waterways within the landscape. Existing BIM and digital twin systems typically simplify plants into static placeholders, ignoring their dynamic behavior under external forces and their safety impact on surrounding facilities. This results in the inability to predict risks such as plant lodging causing injuries to pedestrians, short circuits in power lines, impacts on buildings, and damage to waterways under extreme weather conditions. These technical deficiencies reduce the safety and reliability of the entire lifecycle management of landscape engineering projects and prevent proactive risk warnings. Summary of the Invention

[0003] In view of the problems existing in the prior art, the present invention provides a method and system for dynamic simulation and collaborative design of the entire life cycle of landscape engineering based on digital twins and BIM, which is used to solve the above-mentioned technical problems.

[0004] To achieve the above and other objectives, the technical solution adopted by the present invention is as follows:

[0005] This invention provides a method for dynamic simulation and collaborative design of the entire lifecycle of landscape engineering based on digital twins and BIM, including the following steps:

[0006] Step 1: Construct a unified computational model, which includes a plant dynamic growth sub-model, a park static environmental characteristic sub-model, and an external force environment dynamic load sub-model.

[0007] Step 2: Collect real-time plant status parameters, park infrastructure spatial parameters, and meteorological environment data, and assign initial parameter values ​​to the unified calculation model;

[0008] Step 3: Iteratively execute multiphysics coupling simulation with discrete time steps, update plant morphology and mechanical parameters, calculate wind load and / or snow load, and substitute into the plant dynamics equation to solve the dynamic response of the plant under external forces.

[0009] Step 4: Determine whether local breakage or overall collapse has occurred based on the dynamic response of the plants. If so, generate a spatial sweep of the plant collapse process and perform collision detection and damage assessment with the static environmental feature sub-model of the park. Output the safety risk prediction results for personnel, power lines, buildings and / or water pipe networks on the road.

[0010] Step 5: Based on real-time data transmission from IoT sensors, use an extended Kalman filter to correct the core state vector of the unified computation model online. Repeat steps 3 to 4 with a rolling time window. When the predicted risk exceeds the preset threshold, trigger a graded warning and push disposal suggestions.

[0011] Preferably, the plant dynamic growth sub-model includes a time-dependent decay function for wood mechanical properties, specifically:

[0012] Flexural strength:

[0013]

[0014] Elastic modulus:

[0015]

[0016] Root holding torque:

[0017]

[0018] in For health index, , , DBH is the tree species-specific aging coefficient, where DBH is the diameter at breast height (DBH). , , is the regression coefficient.

[0019] Preferably, the external force environment dynamic load sub-model uses a quasi-steady wind load model to calculate wind pressure, and automatically switches to a CFD coupling mode based on large eddy simulation when the wind speed or turbulence intensity exceeds a set threshold.

[0020] Preferably, the plant dynamics equation is a multi-degree-of-freedom discrete model:

[0021]

[0022] in A bilinear degradation model was adopted; the overall collapse was judged using two criteria: trunk breakage criterion. Criteria for root uprooting The probability density function of the collapse direction is a normal distribution centered on the main wind direction, and the standard deviation is taken as 30° or 60° according to the stability of the wind field.

[0023] Preferably, the generation of the spatial swept body and collision detection specifically involves: simplifying the fallen plants into variable cross-section cones, and adopting a hierarchical collision detection strategy that combines AABB bounding boxes and the GJK algorithm; damage classification includes: determining personnel risk based on road intrusion and open time periods, power lines based on the comparison between the tension force at the collision point and the tensile strength of the conductor, and buildings based on the comparison between the impact kinetic energy and the impact resistance limit of the external wall.

[0024] Preferably, the state vector of the extended Kalman filter is:

[0025]

[0026] When the amount of historical data from the sensor is less than a set threshold, the stiffness reduction coefficient output by the pre-trained neural network is used to replace the EKF.

[0027] Preferably, the plant dynamic growth sub-model supports multi-resolution representation: the community level uses a regular parameterized vegetation model, the individual level uses a simplified functional-structural plant model, and the organ level uses a complete functional-structural plant model; the system dynamically switches the level of detail according to the line of sight and the precision specified by the user.

[0028] Preferably, the length of the rolling time window in step 5 is associated with the type of meteorological event: 0.5 to 2 hours for gale events and 2 to 6 hours for snow disaster events; the graded warning thresholds are: yellow warning is when the risk probability is ≥30% or the impact energy ratio is ≥0.6, orange warning is when it is ≥60% or ≥0.8, and red warning is when it is ≥90% or ≥1.0.

[0029] Preferably, the root uprooting criterion adopts a progressive model: defining the root rotation angle. Its dynamic equation is:

[0030]

[0031] in

[0032]

[0033] when ≥ It is determined that the tube has been completely pulled out.

[0034] The system for dynamic simulation and collaborative design of the entire life cycle of landscape engineering based on digital twins and BIM includes: a data acquisition unit, a unified computing model core engine unit, a dynamic exercise and safety prediction unit, a real-time model calibration and assimilation unit, a twin visualization and collaborative design platform, and a real-time early warning and decision execution module; wherein the data acquisition unit is connected to the unified computing model core engine unit, and the real-time model calibration and assimilation unit is fed back to the unified computing model core engine unit.

[0035] As described above, the method and system for dynamic simulation and collaborative design of the entire life cycle of landscape engineering based on digital twins and BIM provided by this invention have at least the following beneficial effects:

[0036] By constructing a unified computational model integrating a plant dynamic growth sub-model, a park static environmental characteristic sub-model, and an external force dynamic load sub-model, this invention achieves multi-timescale coupled simulation of long-term plant growth succession and instantaneous wind / snow force response. Combined with an extended Kalman filter online correction and rolling prediction mechanism, it accurately identifies the safety risks to personnel and facilities caused by plant lodging under extreme weather conditions, effectively solving the problem that existing digital twin systems for landscape engineering cannot dynamically predict the interactive risks of plant-environment-facilities. On one hand, the processing method of this invention, through solving plant dynamics equations and spatial swept-body collision detection, significantly reduces the risk of injury or damage to facilities caused by plant lodging due to natural disasters such as strong winds and heavy snow, improving the accuracy of safety early warning throughout the entire life cycle of the landscape, thereby avoiding potential personal injury and property loss, and reducing emergency response costs during operation. On the other hand, through real-time data assimilation from IoT sensors and machine learning calibration of historical anomaly patterns, the system can proactively prevent and respond instantly to extreme weather events, ensuring the efficient and safe operation of landscape engineering at all stages of design, construction, and maintenance. This intelligent dynamic simulation and early warning mechanism can not only effectively avoid economic losses and management trust crises caused by safety accidents, but also improve the overall operational stability of the park, ensure the safety and sustainability of the system, and greatly enhance business continuity and reliability in complex weather environments. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention. Detailed Implementation

[0039] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0040] Example: A method for dynamic simulation and collaborative design of the entire lifecycle of landscape engineering based on digital twins and BIM, including the following steps.

[0041] Step 1: Construct a unified computation model

[0042] The unified computational model is composed of a plant dynamic growth sub-model, a park static environmental characteristic sub-model, and an external force environment dynamic load sub-model coupled together.

[0043] The plant dynamic growth sub-model adopts the functional-structural plant model (FSPM) framework, discretizing the plant trunk and primary branches into rigid segments, with adjacent segments connected by torsional spring-damper hinges. This sub-model is based on tree age. Health Index Output morphological parameters (tree height, diameter at breast height, crown radius, leaf area index) and mechanical parameters (elastic modulus, bending strength, root holding moment).

[0044] The static environmental feature sub-model of the park is based on BIM (IFC format) and GIS data, and establishes a three-dimensional geometric model and physical attribute library of roads, power lines, buildings and water pipe networks, and uses R tree to build a spatial index.

[0045] The dynamic load sub-model of the external environment includes a wind load module and a snow load module. The quasi-steady wind load model is used to calculate wind pressure, and the snow load of evergreen tree species is handled by the snow pressure calculation formula.

[0046] Step 2: Collect data and initialize

[0047] Deploy sensors such as wireless inclinometers, surface strain gauges, triaxial accelerometers, and automatic weather stations. Initialization includes: assigning a unique ID to each plant, loading the variety parameter library, setting initial values ​​for the health index, loading the BIM model and building a spatial index, and importing weather forecast data.

[0048] Step 3: Multiphysics Coupled Simulation

[0049] The discrete time step is set to 0.01~0.05 seconds. Within each step, the following steps are executed sequentially: (1) Update plant morphology and mechanical parameters; (2) Calculate wind / snow load based on current meteorological data; (3) Solve the plant dynamics equations. .in For the quality matrix, Here is the Rayleigh damping matrix. The stiffness matrix is ​​nonlinear (using a bilinear degradation model: elastic segment, plastic segment, and fracture segment). Explicit central difference method is used for integration.

[0050] Step 4: Fracture and collapse assessment, swept body generation and collision detection

[0051] Localized fracture: When the bending stress at a branch node exceeds its bending strength, the node and downstream mass are removed.

[0052] Overall collapse: The trunk breakage criterion (bending stress at the base of the trunk exceeds the bending strength) and the root uprooting criterion (overturning moment exceeds the root holding moment) are applied. The probability density function of the collapse direction follows a normal distribution centered on the prevailing wind direction, with the standard deviation set at 30° (dominant wind) or 60° (variable gusts) to account for wind field stability.

[0053] Spatial sweep volume: The fallen plant is simplified into a variable cross-section cone, and it moves with uniform acceleration along the direction of collapse. The trajectory volume from upright to fallen is recorded.

[0054] Collision detection: The first level uses AABB bounding box for coarse screening, and the second level uses GJK algorithm to calculate the precise distance and compare it with the safety margin (0.5m for people / 1.0m for power / 0.2m for buildings) to determine the collision.

[0055] Damage assessment: For personnel risk, the following conditions must be met simultaneously: road intrusion, open hours, and swept object height less than 1.8m; for power lines, the assessment is based on the comparison between the calculated tension force at the impact point and the conductor tensile strength; for buildings, the assessment is based on the comparison between the calculated impact kinetic energy and the impact resistance limit of the exterior wall.

[0056] Step 5: EKF Online Correction and Rolling Warning

[0057] Extended Kalman Filter Setup: State Vector

[0058] (Trunk elastic stiffness, root holding moment, effective windward area), observation vector (Tilt angle, strain, peak acceleration). The state transition adopts a random walk model, and the process noise covariance matrix and observation noise covariance matrix are set according to the sensor calibration accuracy. When the wind speed at a height of 10 meters is ≥10 m / s, EKF updates are performed at a frequency of 1 Hz.

[0059] When historical sensor data is insufficient, a pre-trained two-layer fully connected neural network is used to output a stiffness reduction coefficient, temporarily replacing the EKF.

[0060] Scrolling window length: 0.5~2 hours for strong wind events, 2~6 hours for snow disaster events. Risk probability is calculated based on the combined probability of all plant collapses, and the impact energy ratio is the ratio of the maximum value to the limit value. Graded warning: Yellow (risk probability ≥30% or energy ratio ≥0.6), Orange (≥60% or ≥0.8), Red (≥90% or ≥1.0).

[0061] Specifically, the aging decay function of wood mechanical properties has the following form:

[0062] Flexural strength:

[0063]

[0064] Elastic modulus:

[0065]

[0066] Root holding torque:

[0067]

[0068] in , , The aging coefficient is a species-specific factor, determined through accelerated wood aging tests. , , The root holding moment regression coefficient was determined through field pull-out tests. Health Index The calculation is weighted by fusing information from three aspects: sensor anomaly, UAV multispectral NDVI, and historical attenuation trend.

[0069] Specifically: The external force environment dynamic load sub-model defaults to a quasi-steady wind load model. When the wind speed at a height of 10 meters exceeds 30 m / s or the turbulence intensity exceeds 0.25, the system automatically switches to a CFD coupling mode based on Large Eddy Simulation (LES): OpenFOAM is used to solve the incompressible Navier-Stokes equations, the plants are replaced by a porous medium model, the flow field and structure solver are weakly coupled, the flow field is solved until convergence in each coupling time step, the wind pressure is mapped to the structure, the deformation is updated, and the leaf area density distribution is recalculated.

[0070] Specifically: The plant dynamics equations adopt a multi-degree-of-freedom discrete model, with a stiffness matrix... A bilinear degradation model is used to describe the elastic, plastic, and fracture segments. Two parallel criteria for overall collapse are calculated: the trunk breakage criterion calculates the ratio of the maximum bending stress at the base to the current bending strength; the root uplift criterion calculates the ratio of the total overturning moment to the root holding moment. The probability density function for the collapse direction follows a normal distribution, with the mean taken from the prevailing wind direction and the standard deviation taken as 30° (for stable prevailing wind conditions) or 60° (for variable gust conditions). The three most probable collapse sectors and their probabilities are output.

[0071] Specifically: Spatial sweep volume generation simplifies the fallen plant into a variable cross-section cone, with the base radius taken as the radius of the plant canopy. In the collapse trajectory simulation, the cone accelerates down around the root fulcrum (the point where the trunk breaks or the center of the root ball), and the three-dimensional volume covered by its continuous motion is recorded.

[0072] Collision detection employs a hierarchical strategy: The first level establishes AABB bounding box trees for swept objects and static facilities respectively, quickly eliminating non-collision pairs; the second level uses the GJK algorithm to calculate the precise shortest distance for candidate pairs. After a collision is determined, power line damage is calculated using the ratio of tension force to rated tensile strength according to mechanical formulas, and building damage is calculated using the ratio of impact energy to the external wall limit according to the kinetic energy formula.

[0073] The specific parameters for online correction of the Extended Kalman Filter (EKF) are as follows: the state vector includes trunk elastic stiffness, root holding torque, and effective windward area; the observation vector includes tilt angle, strain, and peak acceleration. The process noise covariance matrix and observation noise covariance matrix are set according to the sensor accuracy. When the effective sensor data for a certain plant is less than 100 samples, a pre-trained neural network (2 fully connected layers, inputs are tree species code, tree height, diameter at breast height, current wind speed, historical average tilt angle, and historical average strain, output is stiffness reduction factor) is used to replace the EKF.

[0074] The plant dynamic growth sub-model supports three resolution levels: community level (line of view > 200m or global planning mode) uses an L-system regular parameterized vegetation model, with the population as the computational unit; individual level (line of view 50-200m) uses a simplified FSPM to simulate macroscopic carbon allocation and biomass accumulation; organ level (line of view < 50m or key monitoring status) uses a full FSPM to simulate organ-scale development and resource competition. The system automatically switches between these resolutions based on the camera distance in the 3D scene and user-generated markers.

[0075] The triggering conditions for rolling forecasts are: real-time wind speed ≥ 10 m / s for more than 5 minutes, or real-time snow depth ≥ 5 cm, or an extreme weather warning in the weather forecast. The window length is adaptively adjusted: for strong wind events, 0.5 hours are used when wind speed ≥ 15 m / s and changes drastically, otherwise 2 hours; for snow disaster events, 2 hours are used when snow depth increases at a rate ≥ 2 cm / h, otherwise 6 hours. Warning thresholds are divided into three levels, corresponding to different response levels and handling recommendations.

[0076] The root uprooting process employs a gradual model, distinct from instantaneous criteria. Root rotation angle is introduced. As a state variable, its dynamic equation is: moment of inertia multiplied by angular acceleration plus damping torque plus holding torque equals overturning moment. The holding torque decreases linearly with rotation angle. This represents the critical rotation angle at which the root system completely fails (calibrated through field pull-out tests, typical value 15°~20°). The root rotation process is solved by numerical integration when... achieve When it was determined that the plant had been completely uprooted, it began to tilt as a whole.

[0077] A dynamic simulation and collaborative design system for the entire lifecycle of landscape engineering based on digital twins and BIM includes:

[0078] Data acquisition unit: Composed of inclinometer, strain gauge, accelerometer, weather station, snow sensor, and multispectral camera, it transmits data via 4G / 5G / edge gateway.

[0079] The core engine unit of the unified computing model is deployed on an edge server or in the cloud. It integrates three sub-models: dynamic plant growth, static park environment, and dynamic external environment, and uses message queue communication.

[0080] Dynamic simulation and safety prediction unit: Executes multiphysics coupled simulation, collapse judgment, swept body generation, collision detection and damage rating algorithms, and supports GPU parallel acceleration.

[0081] The model real-time calibration and assimilation unit runs an extended Kalman filter, maintains the state vector of each plant, and updates it at a frequency of 1 Hz; it is supplemented by a pre-trained neural network for backup.

[0082] Twin visualization and collaborative design platform: developed based on Unity3D, it supports 3D scene rendering, risk layer overlay, design scheme import and simulation playback.

[0083] Real-time early warning and decision execution module: pushes early warning information to the management terminal via TCP / UDP, and coordinates with the park broadcasting system and remote power outage control system to execute intervention measures.

[0084] The data flow for each unit is as follows: data acquisition → unified computing model → dynamic drill → safety prediction → (triggering early warning); at the same time, sensor data is sent to EKF, and the updated state parameters are written back to the unified computing model, forming a closed loop.

[0085] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0086] It should be understood that determining B based on A does not mean determining B solely based on A; it also means determining B based on A and / or other information.

[0087] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0088] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamic simulation and collaborative design of the entire lifecycle of landscape engineering based on digital twins and BIM, characterized in that, Includes the following steps: Step 1: Construct a unified computational model, which includes a plant dynamic growth sub-model, a park static environmental characteristic sub-model, and an external force environment dynamic load sub-model. Step 2: Collect real-time plant status parameters, park infrastructure spatial parameters, and meteorological environment data, and assign initial parameter values ​​to the unified calculation model; Step 3: Iteratively execute multiphysics coupling simulation with discrete time steps, update plant morphology and mechanical parameters, calculate wind load and / or snow load, and substitute into the plant dynamics equation to solve the dynamic response of the plant under external forces. Step 4: Determine whether local breakage or overall collapse has occurred based on the dynamic response of the plants. If so, generate a spatial sweep of the plant collapse process and perform collision detection and damage assessment with the static environmental feature sub-model of the park. Output the safety risk prediction results for personnel, power lines, buildings and / or water pipe networks on the road. Step 5: Based on real-time data transmission from IoT sensors, use an extended Kalman filter to correct the core state vector of the unified computation model online. Repeat steps 3 to 4 with a rolling time window. When the predicted risk exceeds the preset threshold, trigger a graded warning and push disposal suggestions.

2. The method according to claim 1, characterized in that, The plant dynamic growth sub-model includes a time-dependent decay function for wood mechanical properties, specifically: Flexural strength: Elastic modulus: Root holding torque: in For health index, , , DBH is the tree species-specific aging coefficient, where DBH is the diameter at breast height (DBH). , , is the regression coefficient.

3. The method according to claim 1, characterized in that, The external force environment dynamic load sub-model uses a quasi-steady wind load model to calculate wind pressure. When the wind speed or turbulence intensity exceeds a set threshold, it automatically switches to a CFD coupling mode based on large eddy simulation.

4. The method according to claim 1, characterized in that, The plant dynamics equations are multi-degree-of-freedom discrete models. in A bilinear degradation model was adopted; the overall collapse was judged using two criteria: trunk breakage criterion. Criteria for root uprooting The probability density function of the collapse direction is a normal distribution centered on the main wind direction, and the standard deviation is taken as 30° or 60° according to the stability of the wind field.

5. The method according to claim 1, characterized in that, The generated space sweep body and collision detection are specifically as follows: the fallen plants are simplified into variable cross-section cones, and a hierarchical collision detection strategy combining AABB bounding boxes and GJK algorithm is adopted; the damage classification includes: personnel risk is determined according to road intrusion and open time period, power line is determined by comparing the tension at the collision point with the tensile strength of the conductor, and building is determined by comparing the impact kinetic energy with the impact resistance limit of the external wall.

6. The method according to claim 1, characterized in that, The state vector of the extended Kalman filter: When the amount of historical data from the sensor is less than a set threshold, the stiffness reduction coefficient output by the pre-trained neural network is used to replace the EKF.

7. The method according to claim 1, characterized in that, The plant dynamic growth sub-model supports multi-resolution representation: the community level uses a regular parameterized vegetation model, the individual level uses a simplified functional-structural plant model, and the organ level uses a complete functional-structural plant model; the system dynamically switches the level of detail according to the line of sight and the precision specified by the user.

8. The method according to claim 1, characterized in that, In step 5, the length of the rolling time window is related to the type of meteorological event: 0.5 to 2 hours for strong wind events and 2 to 6 hours for snow disaster events; the graded warning thresholds are: yellow warning is when the risk probability is ≥30% or the impact energy ratio is ≥0.6, orange warning is when it is ≥60% or ≥0.8, and red warning is when it is ≥90% or ≥1.

0.

9. The method according to claim 1, characterized in that, The root uprooting criterion adopts a progressive model: defining the root rotation angle. Its dynamic equation is: in when ≥ It is determined that the tube has been completely pulled out.

10. A dynamic simulation and collaborative design system for the entire lifecycle of landscape engineering based on digital twins and BIM, used to implement the method described in any one of claims 1 to 9, characterized in that, include: The system includes a data acquisition unit, a unified computing model core engine unit, a dynamic drill and security prediction unit, a real-time model calibration and assimilation unit, a twin visualization and collaborative design platform, and a real-time early warning and decision execution module. The data acquisition unit is connected to the core engine unit of the unified computing model, and the real-time model calibration and assimilation unit is connected to the core engine unit of the unified computing model.

Images