Digital twinning dynamic optimization system and method for resourceful treatment of construction waste

Abstract

The invention relates to the field of solid waste resourceful treatment, in particular to a digital twinning dynamic optimization system and method for construction waste resourceful treatment, and the method comprises the following steps: collecting related operation data of a construction waste resourceful treatment physical production line, and synchronizing the collected data to a digital twinning virtual production line; the virtual production line and the physical production line are kept consistent in state; monitoring the operation state and synchronous data of the physical production line, judging whether an optimization triggering condition is met or not, if so, entering the next step, and if not, continuously monitoring; based on the synchronization state of the virtual production line and the physical production line, performing analogue simulation on the multiple groups of candidate control parameter combinations, and predicting the production line operation effect corresponding to each group of parameters; according to a preset optimization target, a multi-target optimization algorithm is adopted to evaluate the analogue simulation result, and a global optimal control parameter set is screened out; and issuing the global optimal control parameter set to equipment corresponding to the physical production line, and driving the equipment to execute control operation.

Description

Technical Field

[0001] This invention relates to the field of solid waste resource utilization, and in particular to a digital twin dynamic optimization system and method for the resource utilization of construction waste. Background Technology

[0002] With the continued acceleration of urbanization in my country, the annual amount of construction waste has exceeded 1.5 billion tons, making its resource utilization a key measure to alleviate resource shortages and improve the ecological environment. Currently, construction waste resource utilization production lines generally consist of multiple devices connected in series, such as feeding, crushing, screening, air separation, and magnetic separation. The core objective is to achieve efficient sorting and recycling of materials, but existing technologies still have many bottlenecks that urgently need to be addressed.

[0003] First, there is insufficient data acquisition and synchronization efficiency: existing production lines mostly rely on single-type sensors to collect local data, lacking the ability to integrate and collect data from all dimensions such as incoming material characteristics (such as humidity and composition), equipment operating parameters (such as speed and current), and product quality (such as purity). In addition, the data suffers from problems such as incomplete cleaning and low time alignment accuracy, which makes it difficult for the physical production line and the virtual model to achieve accurate state synchronization, making it difficult to support subsequent simulation and optimization.

[0004] Secondly, there is a weakness in coordinated control and response capabilities: most equipment adopts independent control modes, forming "information silos" and lacking cross-equipment data communication and coordination mechanisms, making it impossible to coordinate and control from a holistic production line perspective. When incoming material characteristics change abruptly or equipment operating status fluctuates, the production line can only respond passively, and sometimes intervention only occurs after product quality fails to meet standards, resulting in serious waste of materials and energy.

[0005] Secondly, the scientific nature and efficiency of optimization decisions are lacking: the adjustment of existing production line parameters relies heavily on the manual experience of operators and lacks quantitative decision-making basis, resulting in large differences in production stability between different shifts; at the same time, a high-precision simulation model that integrates geometry, physics and data-driven approaches has not been built, making it impossible to perform rapid parallel simulation of multiple sets of control parameter combinations, and making it difficult to find the global optimal solution among multiple objectives such as throughput, energy consumption and product purity.

[0006] Finally, there is a lack of closed-loop control mechanisms: existing technologies are mostly one-way processes of "optimization-execution" without establishing a real-time monitoring and feedback calibration mechanism for the execution effect. This leads to a gradual decline in the adaptability of simulation models to physical production lines, making it difficult to guarantee the long-term effectiveness of optimization strategies. Summary of the Invention

[0007] Based on this, it is necessary to provide a digital twin dynamic optimization system and method for the resource utilization of construction waste to address the aforementioned technical problems. By constructing a fully intelligent closed-loop control system that integrates data acquisition synchronization, status monitoring triggering, simulation prediction, optimization decision-making, execution feedback, and model calibration, the system achieves precise status synchronization between the physical and virtual production lines. It collects data on incoming material characteristics, equipment operation, and product quality from multiple types of devices in all dimensions, and synchronizes this data to the virtual production line after preprocessing such as cleaning, deduplication, and time alignment. This solves the problems of partial data acquisition and low synchronization accuracy in existing technologies, and provides reliable data support for subsequent simulation and optimization.

[0008] This invention provides a digital twin dynamic optimization method for the resource recovery of construction waste, the method comprising: Collect relevant operational data from the physical production line for the resource recovery of construction waste, and synchronize the collected data to the digital twin virtual production line to ensure that the virtual production line and the physical production line are in the same state. Monitor the physical production line's operating status and synchronized data, and determine whether the optimization trigger conditions are met. If they are met, proceed to the next step; otherwise, continue monitoring. Based on the synchronization status of the virtual production line and the physical production line, simulation is performed on multiple sets of candidate control parameter combinations to predict the production line operation effect corresponding to each set of parameters. Based on the preset optimization objectives, a multi-objective optimization algorithm is used to evaluate the simulation results and select the globally optimal set of control parameters. The global optimal control parameter set is sent to the corresponding equipment on the physical production line to drive the equipment to perform control operations; The system monitors the operating status and data of the physical production line after it performs control operations, feeds them back to the digital twin virtual production line, completes closed-loop control, and enters the next round of optimization cycle.

[0009] In one embodiment, the step of collecting relevant operational data from the physical production line for the resource recovery of construction waste and synchronizing the collected data to the digital twin virtual production line includes: Data related to incoming material characteristics, equipment operation, and product quality are collected through various devices, including sensors and visual inspection equipment. The collected data is cleaned, deduplicated, and time-aligned before being synchronized to the virtual production line via industrial communication.

[0010] In one embodiment, monitoring the physical production line operating status and synchronization data, and determining whether the optimization triggering conditions are met, includes: Real-time monitoring of synchronized data and equipment operating status, and analysis of data change trends and anomalies; The monitoring results are compared with the preset triggering criteria. If any criterion is met, it is determined that the optimized triggering conditions are met.

[0011] In one embodiment, the simulation of multiple sets of candidate control parameter combinations based on the synchronization status of the virtual production line and the physical production line, and the prediction of the production line operation effect corresponding to each set of parameters, includes: Call upon the geometric, physical, and data-driven fusion model of the digital twin and load synchronized state data; Parallel simulation of multiple candidate control parameter combinations is performed, and the predicted results of the running effect are output and sampled for verification.

[0012] In one embodiment, the step of evaluating the simulation results using a multi-objective optimization algorithm based on a preset optimization objective and selecting the globally optimal set of control parameters includes: Set the weight and priority of each optimization objective according to production needs, and determine the core constraints; The comprehensive benefits of each group of prediction results are calculated using a multi-objective optimization algorithm, and the globally optimal parameter set that meets the constraints is selected.

[0013] In one embodiment, the step of distributing the globally optimal control parameter set to the equipment corresponding to the physical production line and driving the equipment to perform control operations includes: The optimal parameter set is transmitted to the corresponding device controller via the industrial control bus using encrypted transmission. The equipment updates parameters according to the step-by-step adjustment strategy and transmits execution progress and status data back in real time.

[0014] In one embodiment, the monitoring of the physical production line's operating status and data after performing control operations, and the feedback to the digital twin virtual production line to complete closed-loop control and enter the next optimization cycle, includes: Collect and analyze the operating parameters, product quality, and energy consumption data after the equipment has been executed. The analyzed data is fed back to the virtual production line to calibrate the simulation model parameters and update the training sample library.

[0015] This invention also provides a digital twin dynamic optimization system for the resource recovery of construction waste, applied to the digital twin dynamic optimization method for the resource recovery of construction waste described in any of the above embodiments, the system comprising: The data synchronization module is used to collect relevant operational data from the physical production line of construction waste recycling and synchronize the collected data to the digital twin virtual production line, so that the virtual production line and the physical production line keep their status consistent. The trigger judgment module is used to monitor the physical production line operation status and synchronize data, and determine whether the optimization trigger conditions are met. If they are met, the next step is taken; otherwise, monitoring continues. The simulation prediction module is used to simulate multiple sets of candidate control parameter combinations based on the synchronization status of the virtual production line and the physical production line, and predict the production line operation effect corresponding to each set of parameters. The optimization decision module is used to evaluate the simulation results based on the preset optimization objectives and to select the globally optimal set of control parameters using a multi-objective optimization algorithm. The execution module is used to distribute the global optimal control parameter set to the corresponding equipment on the physical production line, driving the equipment to perform control operations; The closed-loop feedback module is used to monitor the operating status and data of the physical production line after it performs control operations, and feeds it back to the digital twin virtual production line to complete closed-loop control and enter the next round of optimization cycle.

[0016] The present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the digital twin dynamic optimization method for the resource recovery of construction waste as described above.

[0017] The present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, implements the digital twin dynamic optimization method for the resource recovery of construction waste as described above.

[0018] The present invention also provides a computer program product, comprising a computer program that, when executed by a processor, implements the digital twin dynamic optimization method for the resource recovery of construction waste as described above. The aforementioned digital twin dynamic optimization system and method for the resource utilization of construction waste achieves precise state synchronization between the physical and virtual production lines by constructing a fully intelligent closed-loop control system encompassing data acquisition synchronization, state monitoring triggering, simulation prediction, optimization decision-making, execution feedback, and model calibration. It collects data on incoming material characteristics, equipment operation, and product quality from multiple types of devices, and synchronizes this data to the virtual production line after preprocessing (cleaning, deduplication, and time alignment). This solves the problems of incomplete data acquisition and low synchronization accuracy in existing technologies, providing reliable data support for subsequent simulation and optimization. Furthermore, it leverages the geometric, physical, and data-driven fusion model of the digital twin to conduct parallel simulations of multiple candidate control parameters, significantly improving the efficiency and accuracy of operational effect prediction. This overcomes the shortcomings of traditional single-model simulations, such as slow speed and poor adaptability, laying the foundation for rapid optimization decisions. Finally, it utilizes a multi-objective optimization algorithm... By considering core objectives such as processing capacity, energy consumption, and product purity, and setting weights, priorities, and constraints based on production needs, a globally optimal parameter set is selected. This breaks down the "information silos" of independent equipment control, achieving global collaborative optimization of the production line and avoiding the problems of poor production stability and target imbalance caused by relying on human experience. A closed-loop feedback control mechanism is constructed, feeding back the operational data after equipment execution to the virtual production line for model parameter calibration and training sample library updates. This ensures that the optimization strategy continuously adapts to changes in production line conditions, solving the problems of insufficient continuous optimization capabilities and long-term effectiveness of existing technologies. Encrypted transmission ensures the security of parameter distribution, and a step-by-step adjustment strategy ensures stable equipment operation, reducing material loss and energy waste during production line adjustments. Ultimately, this achieves improved quality, increased efficiency, and reduced costs in the resource-based treatment of construction waste, significantly enhancing the production line's intelligence level and market competitiveness. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 The flowchart of the digital twin dynamic optimization method provided by the present invention is shown below; Figure 2 This is a block diagram of the digital twin dynamic optimization system provided by the present invention; Figure 3 An internal structural diagram of the computer device provided by the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The following is combined Figures 1 to 3 This invention describes a digital twin dynamic optimization system and method for the resource recovery and treatment of construction waste.

[0023] In one embodiment, a digital twin dynamic optimization method for the resource recovery of construction waste includes the following steps: Step S100: Collect relevant operational data of the physical production line for the resource recovery of construction waste, and synchronize the collected data to the digital twin virtual production line to ensure that the virtual production line and the physical production line are in the same state.

[0024] Step S200: Monitor the physical production line operation status and synchronization data, and determine whether the optimization trigger conditions are met. If they are met, proceed to the next step; otherwise, continue monitoring.

[0025] Step S300: Based on the synchronization status of the virtual production line and the physical production line, simulate multiple sets of candidate control parameter combinations to predict the production line operation effect corresponding to each set of parameters.

[0026] Step S400: Based on the preset optimization objective, a multi-objective optimization algorithm is used to evaluate the simulation results and select the globally optimal set of control parameters.

[0027] In step S500, the global optimal control parameter set is sent to the corresponding equipment on the physical production line to drive the equipment to perform control operations. Step S600: Monitor the operating status and data of the physical production line after it performs control operations, feed them back to the digital twin virtual production line, complete closed-loop control, and enter the next round of optimization cycle.

[0028] The aforementioned digital twin dynamic optimization method for the resource utilization of construction waste achieves precise state synchronization between the physical and virtual production lines by constructing a fully intelligent closed-loop control system encompassing data acquisition synchronization, state monitoring triggering, simulation prediction, optimization decision-making, execution feedback, and model calibration. This system collects data on incoming material characteristics, equipment operation, and product quality from multiple types of devices, pre-processing this data through cleaning, deduplication, and time alignment before synchronizing it to the virtual production line. This solves the problems of incomplete data acquisition and low synchronization accuracy in existing technologies, providing reliable data support for subsequent simulation and optimization. Furthermore, by relying on the geometric, physical, and data-driven fusion model of the digital twin to conduct parallel simulations of multiple candidate control parameters, the efficiency and accuracy of operational effect prediction are significantly improved. This overcomes the shortcomings of traditional single-model simulations, such as slow speed and poor adaptability, laying the foundation for rapid optimization decisions. Finally, through multi-objective optimization algorithms, the system comprehensively processes data. Core objectives such as throughput, energy consumption, and product purity are weighted, prioritized, and constrained based on production needs. A globally optimal parameter set is selected, breaking down the "information silos" of independent equipment control and achieving global collaborative optimization of the production line. This avoids the problems of poor production stability and target imbalance caused by reliance on manual experience. A closed-loop feedback control mechanism is constructed, feeding back operational data from equipment execution to the virtual production line for model parameter calibration and training sample library updates. This ensures that the optimization strategy continuously adapts to changes in production line conditions, solving the problems of insufficient continuous optimization capabilities and long-term effectiveness in existing technologies. Encrypted transmission ensures the security of parameter distribution, and a step-by-step adjustment strategy ensures stable equipment operation, reducing material loss and energy waste during production line adjustments. Ultimately, this achieves improved quality, efficiency, and cost reduction in the resource-based treatment of construction waste, significantly enhancing the production line's intelligence level and market competitiveness.

[0029] In one embodiment, collecting relevant operational data from the physical production line for the resource recovery of construction waste and synchronizing the collected data to the digital twin virtual production line includes the following steps: Step S110 involves collecting data related to incoming material characteristics, equipment operation, and product quality through various devices, including sensors and visual inspection equipment.

[0030] Specifically, the data acquisition device includes: Near-infrared moisture analyzer: Deployed at the discharge end of the plate chain feeder, it is used to detect the original moisture content of the incoming material. The equipment type is an online near-infrared moisture analyzer.

[0031] Laser scanning volume meter: Deployed at the discharge end of the plate chain feeder, it calculates the volumetric flow rate of materials through laser three-dimensional scanning.

[0032] Current transformer: Installed at the wiring terminals of the drive motors of various main equipment (feeders, crushers, air separators, etc.) to collect the working current.

[0033] High-speed industrial cameras: deployed at the heavy material outlet of the air classifier and the finished aggregate conveyor belt (key quality control points), with a built-in AI vision recognition module (based on YOLOv8 training).

[0034] Inverter data interface: Directly reads the control parameters (speed, rotation speed, frequency) of the feeder, crusher, and air separator fan. The communication protocol is Modbus RTU.

[0035] The collected data includes: Incoming material moisture content ( (Unit: %): The data is from real-time detection by a near-infrared moisture meter, reflecting the moisture content of the material (range 10%-30%, covering the common humidity range of construction and decoration waste).

[0036] Incoming volumetric flow rate ( (Unit: m³ / h): The source is the detection value of the laser scanning volumetric gauge, which directly outputs the material volume conveying rate.

[0037] feeder speed ( (Unit: m / s): The source is the converted value of the output frequency of the feeder inverter (frequency and speed are linearly related: speed = 0.01 × frequency, frequency range 0-50Hz).

[0038] Crusher speed ( (Unit: rpm): The source is the conversion value of the main motor inverter of the crusher (speed = 60 × frequency / number of motor pole pairs, number of motor pole pairs 2, frequency range 0-50Hz).

[0039] Air classifier fan frequency ( , (Unit: Hz): The source is the direct reading value from the frequency converters of the two high-pressure density air separators, ranging from 0 to 50 Hz.

[0040] Host operating current ( (Unit: A): The source is the real-time current value detected by the current transformer, reflecting the load status of the equipment.

[0041] Visual recognition purity ( (Unit: %): The source is the calculation result of the effective aggregate (brick, concrete) ratio identified by the AI ​​module after the material image is captured by a high-speed industrial camera.

[0042] All data is collected at a frequency of 1 second per collection to ensure data timeliness and alignment with the timeline.

[0043] Step S120: The collected data is cleaned, deduplicated, and time-aligned preprocessed, and then synchronized to the virtual production line via industrial communication.

[0044] Specifically, invalid data (such as...) >100% or <0% For outliers >500A (out of the transformer's range), linear interpolation was used to fill in the occasional missing data (missing rate ≤0.5%).

[0045] Delete redundant data collected repeatedly (multiple sets of identical data with the same timestamp due to communication delays), and retain the initial collection value.

[0046] Based on the unified system clock, the timestamps of all data are corrected to multiples of 1 second (such as 16:00:00, 16:00:01) to ensure the consistency of time of data collected by different devices.

[0047] The preprocessed data is uploaded to the digital twin platform server using the Profinet industrial Ethernet communication protocol (transmission rate 100Mbps).

[0048] Synchronization delay is controlled within 50ms to ensure that the difference between the virtual production line and the physical production line status is ≤1 second.

[0049] In one embodiment, monitoring the physical production line operating status and synchronization data, and determining whether the optimization triggering conditions are met, includes the following steps: Step S210: Monitor the synchronized data and equipment operating status in real time, and analyze the data change trends and abnormal situations.

[0050] Specifically, key monitoring (Incoming material moisture content) (Visual purity) T (System processing capacity) P (Energy consumption per ton of product), of which: T = × Average bulk density of materials (the average bulk density of construction and decoration waste is taken as 1.8t / m³, based on the "Technical Standard for Construction Waste Treatment" GB / T 51377-2019). P = (Power of each device × Running time) / Processing capacity (Device power = √3 × Voltage × Current × Power factor, voltage is 380V, power factor is 0.85). The PLC reads device fault codes (such as motor overload, sensor disconnection) and operating modes (automatic / manual) to determine if the device is working properly.

[0051] right The trend value is calculated using a 5-minute moving average (to avoid accidental triggering due to instantaneous fluctuations). The formula is: Moving average = (first 299 seconds) / (last 299 seconds) +Current ) / 300.

[0052] right The difference between the real-time value and the average value 10 minutes ago is calculated using the 10-minute average comparison method to determine the sudden change in humidity.

[0053] Step S220: Compare the monitoring results with the preset trigger criteria. If any criterion is met, it is determined that the optimized triggering conditions are met.

[0054] Specifically, the preset triggering criteria and basis are as follows: Standard 1: Sudden change in incoming material moisture content (| ( t ) - ( t -10min)|>5%.

[0055] Basis: When the moisture content of construction and renovation waste fluctuates by more than 5%, it will cause changes in the material's viscosity, significantly affecting crushing efficiency (excessive moisture content easily leads to material blockage) and air separation accuracy (wet material is prone to clumping), requiring timely adjustment of parameters.

[0056] Standard 2: Product purity trend deviation (| Moving average - |>2%).

[0057] in accordance with: (Product purity target value) = 98% (industry standard for high-quality recycled aggregate purity). A deviation of more than 2% will result in the product failing to meet the standard. Triggering optimization can prevent the generation of batches of unqualified products.

[0058] Standard 3: Periodic triggering (60-minute intervals).

[0059] Basis: Production line operating conditions may experience slow drift (such as load changes caused by equipment wear). Regular optimization can maintain long-term stable operation, and 60 minutes is the optimal cycle that balances optimization efficiency and equipment stability.

[0060] Standard 4: Triggered by manual command.

[0061] Basis: When operators discover abnormal material characteristics (such as excessive large impurities) or receive a production plan adjustment notice, they can manually trigger optimization through the industrial control platform.

[0062] If any of the above criteria are met, it is determined that the trigger condition is met and proceeds to the next step; otherwise, return to step S100 for continuous monitoring.

[0063] In one embodiment, the step of simulating multiple combinations of candidate control parameters based on the synchronization status of the virtual production line and the physical production line, and predicting the production line operation effect corresponding to each set of parameters, includes the following steps: Step S310: Invoke the geometric, physical, and data-driven fusion model of the digital twin and load the synchronization state data.

[0064] Specifically, the geometric model can be built using SolidWorks 2023, with data sources including production line layout diagrams and equipment mechanical drawings (including precise three-dimensional dimensions such as the 12m length of the plate chain feeder and the 800×600mm size of the crusher feed inlet). Its core function is to restore the spatial position and connection relationship of the equipment, ensuring that the simulation scene is consistent with the physical production line.

[0065] In the physical model, the air classifier can be modeled using ANSYS Fluent 2023 to construct a gas-solid two-phase flow model. Input parameters include duct dimensions (width 1.2m × height 0.8m), fan performance curves (frequency versus airflow / pressure relationship), and material density (2.4t / m³ for brick and concrete aggregates, 1.2t / m³ for impurities), used to simulate the separation process of materials by airflow. The screening machine (stepped screen + composite tension screen) can be modeled using EDEM 2023 to construct a particle dynamics model. Input parameters include amplitude (5mm), vibration frequency (18Hz), screen aperture size (10mm / 20mm), and material friction coefficient (0.35 for aggregates and steel), used to simulate particle screening efficiency.

[0066] The algorithm chosen is the random forest regression model (based on the Scikit-learn 1.2.2 library) because this algorithm has excellent fitting effect on multi-feature nonlinear mapping and strong noise resistance (adapting to production line data fluctuations).

[0067] Collect historical operating data from the production line for three consecutive months (a total of 2160 hours), covering incoming material moisture content of 10%-30% and three production shifts (early shift 8:00-16:00, middle shift 16:00-24:00, and night shift 0:00-8:00) to ensure the model's generalization ability.

[0068] Input feature vector X :[ , , , , , (6 dimensions, all data after standardization).

[0069] Output target vector Y :[ T , P , Q (System processing capacity, energy consumption per ton of product, product purity).

[0070] Model validation can employ 5-fold cross-validation, validation set R² ≥0.92 T ), ≥0.88 ( P ), ≥0.93 ( Q This meets the simulation accuracy requirements.

[0071] The real-time data after preprocessing in step S120 ( , The simulation starts by loading current control parameters into the fusion model and updating the current state of the virtual production line (such as material inventory and equipment operating parameters) to ensure that the simulation start point is consistent with the physical production line.

[0072] Step S320: Simulate multiple sets of candidate control parameter combinations in parallel, output the predicted results of the running effect, and perform sampling verification.

[0073] Specifically, candidate control parameter combinations include: Parameter range: Based on the equipment's safe operation threshold and the historical optimal parameter range, the candidate parameter range is set as follows: : 0.3-0.5m / s (corresponding to a frequency of 30-50Hz for the strain gauge, to avoid material accumulation due to excessively low speed and equipment overload due to excessively high speed).

[0074] 800-1200rpm (corresponding to a frequency converter frequency of 26.7-40Hz, matching the design speed range of the equipment).

[0075] , 35-45Hz (effective frequency range determined based on air classifier sorting efficiency test).

[0076] Number of combinations: 300 candidate parameters are generated using the Latin hypercube sampling method (balancing coverage and simulation efficiency, 300 combinations can cover more than 95% of the effective parameter combinations).

[0077] Simulation tool: Based on the Python multi-process parallel framework (ProcessPoolExecutor), it calls the data-driven model for simulation. The simulation time for a single set of parameters is ≤10ms, and the total time for 300 sets of parameters is ≤3 seconds (meeting the requirements of real-time optimization).

[0078] Output: Predicted values ​​for each set of parameters [ (Predicting system processing capacity) (Energy consumption per ton of product) (Predicted product purity)], where = ×1.8 (average bulk density of material). = Predicted total energy consumption / , Use the model output directly.

[0079] The process of calculating the prediction deviation between the physical model and the data-driven model is as follows: Offline phase: High-fidelity simulations are performed on 50% of the historical parameter combinations using a physical model to calibrate the data-driven model parameters and ensure that the prediction bias of the data-driven model meets the requirements. - ≤3%, - ≤5%, - |≤2%, solidify the model after completion.

[0080] In the online phase: the step of "randomly selecting 10% of candidate parameters to call the physical model for verification" has been removed. The model is driven by calibrated data to perform parallel simulation of 300 sets of parameters, and the results are output within 3 seconds to meet the requirements of real-time optimization.

[0081] in, This represents the actual processing volume obtained from the physical model simulation; This represents the actual energy consumption per ton of product obtained from the physical model simulation. This indicates the purity of the actual product obtained from the physical model simulation.

[0082] In one embodiment, the step of evaluating the simulation results using a multi-objective optimization algorithm based on a preset optimization objective and selecting the globally optimal set of control parameters includes the following steps: Step S410: Set the weight and priority of each optimization objective according to production needs, and determine the core constraints.

[0083] Specifically, the optimization target weights and their basis are as follows: Overall fitness function: Constraints must be met in, This represents the weighting coefficient for processing volume; This represents the energy consumption weighting coefficient; This represents the purity deviation weighting coefficient; Indicates the maximum processing capacity of the design; This indicates the baseline energy consumption.

[0084] Some examples are as follows: =50t / h (maximum designed capacity of the production line, sourced from the equipment technical manual), weight =0.5 (Prioritize processing efficiency to match the annual capacity demand of 300,000 tons).

[0085] When the actual processing volume hour, Takes the value 1, when When the ratio is calculated, it should be based on the actual ratio.

[0086] =4.5kWh / t (average energy consumption per ton of product over the past 3 months, sourced from historical data statistics), weight =0.3 (Secondary objective is cost reduction, energy consumption is the core cost item).

[0087] When actual energy consumption hour, , Equivalent to F Bonus points.

[0088] =98% (Industry standard for purity of high-quality recycled aggregate, sourced from "Technical Specification for Application of Recycled Aggregates" JGJ / T 240-2011), weight =0.2 (Purity must be consistently up to standard to avoid excessive weighting that could lead to a significant decrease in throughput).

[0089] Core constraints: Q≥ (Minimum allowable purity of product) = 97% (National minimum purity requirement for recycled aggregate, sourced from GB / T 14684-2022 "Construction Sand"), to ensure product qualification.

[0090] Equipment operation safety constraints: ≤400A (80% of the rated current of each main motor to avoid overload and burn out the motor). ≤0.5m / s ≤1200rpm (upper limit of equipment design).

[0091] Step S420: Calculate the comprehensive benefits of each group of prediction results using a multi-objective optimization algorithm, and select the globally optimal parameter set that meets the constraints.

[0092] Specifically, the optimization algorithm selection and parameter settings are as follows: Algorithm type: Genetic algorithm (because this algorithm is good at handling multi-objective nonlinear optimization problems and has strong global search capabilities), implemented based on the DEAP library.

[0093] Algorithm parameters: population size 300 (consistent with the number of candidate parameter combinations), number of iterations 50 generations, crossover probability 0.8, mutation probability 0.1 (optimal parameters determined through multiple experiments, balancing search efficiency and convergence).

[0094] The screening process is as follows: Step 1: Calculate the overall fitness of 300 candidate parameters FAt the same time, check whether the constraints are met (Q≥97%). ≤400A, etc.), remove parameter groups that do not meet the constraints (about 5%-10%).

[0095] Step 2: For parameter sets that meet the constraints, according to... F Sort the values ​​in descending order and select the top 5 groups as candidate optimal solutions.

[0096] Step 3: For the first 5 sets of parameters, run the data-driven model simulation again to verify and eliminate outliers caused by random errors in the algorithm, and finally select... F The set of parameters with the largest value and the most stable prediction results is selected as the global optimal control parameter set.

[0097] Example result: If the initial =18% suddenly jumped to 25% (triggered optimization). After algorithmic selection, the optimal parameter set is: =0.47m / s (5% lower than the current speed) =42Hz =40Hz, prediction Q =98.2% (constraints satisfied) P 8% lower than the current level T =48t / h (close to) ).

[0098] In one embodiment, the step of distributing the globally optimal control parameter set to the equipment corresponding to the physical production line and driving the equipment to perform control operations includes the following steps: Step S510: The optimal parameter set is transmitted to the corresponding device controller via the industrial control bus using encrypted transmission.

[0099] Specifically, the encryption method is as follows: the optimal parameter set is encrypted using the AES-256 symmetric encryption algorithm (the key is configured offline to the PLC and the server and updated regularly) to prevent the parameters from being tampered with during transmission.

[0100] Transmission channel: Data is sent to the corresponding PLC controller (model S7-1500, response time ≤10ms) of each device via industrial control bus (Modbus TCP protocol). The PLC address corresponds one-to-one with the device (e.g., the feeder PLC address is 192.168.1.101).

[0101] The instruction includes parameter name, target value, adjustment duration, and checksum. Example: {“Equipment Type”:“Feeder”,“Parameter Name”:: "Target value": 0.47 m / s, "Adjustment duration": 60 seconds, "Check code": "3F7D2A9B"}, the check code is used by the PLC to verify the integrity of the instruction.

[0102] In step S520, the device performs parameter updates according to the step-by-step adjustment strategy and transmits execution progress and status data back in real time.

[0103] Specifically, the step-by-step adjustment strategy is as follows: feeder speed ( Adjust the speed in increments of 0.01 m / s, with a 10-second interval between each step (a 5% reduction is achieved in 60 seconds) to avoid sudden speed drops that could lead to material accumulation.

[0104] Air classifier frequency ( , Adjustments are made in 1Hz increments, with a 5-second interval between each step (the adjustment from 38Hz to 42Hz / 40Hz is completed in 10 seconds) to avoid sudden changes in airflow affecting sorting accuracy.

[0105] Crusher speed ( If adjustments are needed, adjust in steps of 50 rpm each time, with a 15-second interval between each step, to match the material conveying rhythm (only start step-by-step adjustments when the difference between the optimal speed and the current speed is ≥100 rpm, to avoid frequent small adjustments that could cause equipment wear).

[0106] During equipment operation, the PLC transmits the current parameter values, operating current, and fault status back to the digital twin platform every 2 seconds. The platform displays the adjustment progress in real time (e.g., "The feeder speed has been adjusted to 0.47m / s, 100% complete").

[0107] If the current exceeds the limit during the adjustment process ( If the value is greater than 400A or the equipment malfunctions, immediately pause the adjustment and revert to the parameters before adjustment, while issuing an alarm signal.

[0108] In one embodiment, the monitoring of the physical production line's operating status and data after performing control operations, and the feedback to the digital twin virtual production line to complete closed-loop control and enter the next optimization cycle, includes the following steps: Step S610: Collect the operating parameters, product quality, and energy consumption data after the equipment is executed, and then organize and analyze them.

[0109] Specifically, after the parameters are adjusted, running data is continuously collected for 30 minutes. T, P, , (etc.), the sampling frequency remains at 1 second / time to ensure the stable operation cycle of the coverage equipment.

[0110] Calculate the actual values ​​of key indicators: such as actual processing volume. =Average ×1.8, actual energy consumption per ton = Total power consumption / Actual purity = The 30-minute average.

[0111] Deviation analysis: Calculate the relative deviation between the actual value and the predicted value. =| - | / ×100% =| - | / ×100% =| - | / ×100% The allowable deviation range is (Relative deviation in processing volume) ≤5%, (Relative deviation in energy consumption) ≤8%, (Relative deviation in purity) ≤2%.

[0112] Step S620: Feed the analyzed data back to the virtual production line, calibrate the simulation model parameters, and update the training sample library.

[0113] Specifically, if the deviation is within the allowable range: add new data from the last 30 minutes to the training sample library of the data-driven model, and retrain the model every 7 days according to the "rolling update" principle (retain data from the last 6 months, with a maximum capacity of 100,000 data entries in the training sample library, and delete the oldest data according to the "first-in, first-out" principle when adding new data to ensure the timeliness of the samples) to maintain the model's generalization ability.

[0114] If the deviation exceeds the allowable range (e.g.) =3%): Analyze the causes of deviations (such as changes in material properties or equipment wear), adjust the physical model parameters (such as updating the material friction coefficient), increase the sampling density of candidate parameters under this operating condition, and retrain the data-driven model.

[0115] After calibration, the system returns to step S100, continuously monitors the physical production line status, and waits for the next optimization trigger condition to form a closed-loop control cycle.

[0116] The following describes the digital twin dynamic optimization system for the resource recovery of construction waste provided by this invention. The digital twin dynamic optimization system for the resource recovery of construction waste described below can be referred to in correspondence with the digital twin dynamic optimization method for the resource recovery of construction waste described above.

[0117] In one embodiment, a digital twin dynamic optimization system for the resource recovery of construction waste, applied to the digital twin dynamic optimization method for the resource recovery of construction waste described in any of the above embodiments, includes a data synchronization module 100, a trigger judgment module 200, a simulation prediction module 300, an optimization decision module 400, a distribution execution module 500, and a closed-loop feedback module 600. The data synchronization module 100 is used to collect relevant operational data of the physical production line for the resource utilization of construction waste and synchronize the collected data to the digital twin virtual production line so that the virtual production line and the physical production line keep their status consistent. The trigger judgment module 200 is used to monitor the physical production line operation status and synchronize data, and determine whether the optimization trigger conditions are met. If they are met, the next step is taken; otherwise, monitoring continues. The simulation prediction module 300 is used to simulate multiple sets of candidate control parameter combinations based on the synchronization status of the virtual production line and the physical production line, and predict the production line operation effect corresponding to each set of parameters. The optimization decision module 400 is used to evaluate the simulation results based on the preset optimization objectives and to select the globally optimal set of control parameters using a multi-objective optimization algorithm. The execution module 500 is used to distribute the global optimal control parameter set to the corresponding equipment on the physical production line, driving the equipment to perform control operations. The closed-loop feedback module 600 is used to monitor the operating status and data of the physical production line after it performs control operations, and feeds it back to the digital twin virtual production line to complete closed-loop control and enter the next round of optimization cycle.

[0118] Figure 3 This example illustrates a schematic diagram of the physical structure of an electronic device, which can be a smart terminal. Its internal structure diagram can be as follows: Figure 3 As shown, the electronic device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a digital twin dynamic optimization method for the resource recovery of construction waste.

[0119] Those skilled in the art will understand that Figure 3The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the electronic device to which the present invention is applied. A specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0120] On the other hand, the present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, realizes a digital twin dynamic optimization method for the resource utilization of construction waste.

[0121] On another front, a computer program product or computer program is provided, comprising computer instructions stored in a computer storage medium. A processor of an electronic device reads the computer instructions from the computer storage medium, and when the processor executes the computer instructions, it implements a digital twin dynamic optimization method for the resource recovery of construction waste.

[0122] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory.

[0123] By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0124] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0125] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A digital twin dynamic optimization method for the resource utilization of construction waste, characterized in that, The method includes: Collect relevant operational data from the physical production line for the resource recovery of construction waste, and synchronize the collected data to the digital twin virtual production line to ensure that the virtual production line and the physical production line are in the same state. Monitor the physical production line's operating status and synchronized data, and determine whether the optimization trigger conditions are met. If they are met, proceed to the next step; otherwise, continue monitoring. Based on the synchronization status of the virtual production line and the physical production line, simulation is performed on multiple sets of candidate control parameter combinations to predict the production line operation effect corresponding to each set of parameters. Based on the preset optimization objectives, a multi-objective optimization algorithm is used to evaluate the simulation results and select the globally optimal set of control parameters. The global optimal control parameter set is sent to the corresponding equipment on the physical production line to drive the equipment to perform control operations; The system monitors the operating status and data of the physical production line after it performs control operations, feeds them back to the digital twin virtual production line, completes closed-loop control, and enters the next round of optimization cycle.

2. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 1, characterized in that, The process of collecting relevant operational data from the physical production line for the resource recovery of construction waste and synchronizing the collected data to the digital twin virtual production line includes: Data related to incoming material characteristics, equipment operation, and product quality are collected through various devices, including sensors and visual inspection equipment. The collected data is cleaned, deduplicated, and time-aligned before being synchronized to the virtual production line via industrial communication.

3. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 2, characterized in that, The monitoring of the physical production line's operating status and synchronized data, and the determination of whether the optimization trigger conditions are met, include: Real-time monitoring of synchronized data and equipment operating status, and analysis of data change trends and anomalies; The monitoring results are compared with the preset triggering criteria. If any criterion is met, it is determined that the optimized triggering conditions are met.

4. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 3, characterized in that, The simulation of multiple candidate control parameter combinations based on the synchronization status of the virtual production line and the physical production line is used to predict the production line operation effect corresponding to each set of parameters, including: Call upon the geometric, physical, and data-driven fusion model of the digital twin and load synchronized state data; Parallel simulation of multiple candidate control parameter combinations is performed, and the predicted results of the running effect are output and sampled for verification.

5. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 4, characterized in that, The process involves evaluating the simulation results using a multi-objective optimization algorithm based on a preset optimization objective, and selecting the globally optimal set of control parameters, including: Set the weight and priority of each optimization objective according to production needs, and determine the core constraints; The comprehensive benefits of each group of prediction results are calculated using a multi-objective optimization algorithm, and the globally optimal parameter set that meets the constraints is selected.

6. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 5, characterized in that, The step of distributing the globally optimal control parameter set to the corresponding equipment on the physical production line and driving the equipment to perform control operations includes: The optimal parameter set is transmitted to the corresponding device controller via the industrial control bus using encrypted transmission. The equipment updates parameters according to the step-by-step adjustment strategy and transmits execution progress and status data back in real time.

7. The digital twin dynamic optimization method for the resource utilization of construction waste according to claim 6, characterized in that, The monitoring of the physical production line's operational status and data after executing control operations is fed back to the digital twin virtual production line, completing closed-loop control and entering the next optimization cycle, including: Collect and analyze the operating parameters, product quality, and energy consumption data after the equipment has been executed. The analyzed data is fed back to the virtual production line to calibrate the simulation model parameters and update the training sample library.

8. A digital twin dynamic optimization system for the resource recovery of construction waste, applied to the digital twin dynamic optimization method for the resource recovery of construction waste as described in any one of claims 1 to 7, characterized in that, The system includes: The data synchronization module is used to collect relevant operational data from the physical production line of construction waste recycling and synchronize the collected data to the digital twin virtual production line, so that the virtual production line and the physical production line keep their status consistent. The trigger judgment module is used to monitor the physical production line operation status and synchronize data, and determine whether the optimization trigger conditions are met. If they are met, the next step is taken; otherwise, monitoring continues. The simulation prediction module is used to simulate multiple sets of candidate control parameter combinations based on the synchronization status of the virtual production line and the physical production line, and predict the production line operation effect corresponding to each set of parameters. The optimization decision module is used to evaluate the simulation results based on the preset optimization objectives and to select the globally optimal set of control parameters using a multi-objective optimization algorithm. The execution module is used to distribute the global optimal control parameter set to the corresponding equipment on the physical production line, driving the equipment to perform control operations; The closed-loop feedback module is used to monitor the operating status and data of the physical production line after it performs control operations, and feeds it back to the digital twin virtual production line to complete closed-loop control and enter the next round of optimization cycle.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images