An intelligent control system and method for the whole process of in-situ mining of oil sand mines

Abstract

The present application relates to the technical field of oil sand mining, and discloses an in-situ mining full-process collaborative intelligent control system and method for oil sand mines, which comprises a perception layer, an intelligent execution layer, an edge control layer, a digital twin core layer and a collaborative scheduling layer.The perception layer is used for collecting operation parameters in real time and performing preliminary mean filtering processing on the collected data.The intelligent execution layer is used for receiving collaborative control instructions and completing automatic execution of full-process operation actions.The edge control layer is used for receiving data from the perception layer and uploading the data.The digital twin core layer is used for constructing a full-factor digital twin model and performing real-time iterative updating.The collaborative scheduling layer is used for optimizing scheduling algorithms, outputting collaborative control instructions and realizing dynamic matching of each operation link.The system can realize collaborative linkage, adaptive regulation and control and precise operation of each link in the full process, improve oil sand recovery rate and backfill quality, reduce formation disturbance and operation risk, and promote the intelligent and unmanned development of in-situ mining technology for oil sand mines.

Description

Technical Field

[0001] This invention relates to the field of oil sands mining technology, and more specifically, to a collaborative intelligent control system and method for the entire process of in-situ oil sands mining. Background Technology

[0002] As an important unconventional oil and gas resource, in-situ extraction technology for oil sands has become the mainstream direction for oil sands development due to its advantages such as high resource utilization and minimal surface damage.

[0003] In existing in-situ oil sands mining technologies, each operational stage often employs independent control equipment and operating modes, lacking a collaborative management and control mechanism across the entire process. This leads to problems such as a mismatch between excavation volume and lifting capacity, a disconnect between lifting volume and separation and processing capacity, and an imbalance between tailings separation and backfilling requirements. This not only reduces mining efficiency and resource recovery rate but also easily causes pipeline blockages, equipment overloads, and other malfunctions. At the same time, traditional control methods rely heavily on manual operation and experience-based adjustments, making it difficult to adapt to complex and changing geological conditions and material characteristics in real time. This results in low control precision, and the lack of precise control over backfilling of mined-out areas after mining can easily cause geological disasters such as ground subsidence and collapse, resulting in irreversible damage to the ecological environment.

[0004] Furthermore, existing intelligent technologies for oil sands mining are mostly concentrated on single stages, such as the automation of underground excavation equipment or the intelligentization of surface separation equipment. An integrated intelligent control system covering the entire process of mining, separation, and backfilling has not yet been formed. While layered intelligent management and control architectures have emerged in the mining and oilfield sectors, these architectures can only achieve general status monitoring and routine scheduling. They cannot address the multi-stage coupling characteristics of oil sands mining, enabling dynamic flow matching for excavation, lifting, separation, and backfilling. Nor can they combine the geological and technological characteristics of oil sands to achieve multi-objective collaborative optimization. This results in existing technologies failing to solve the technical problems of full-process coordination in oil sands mining, hindering the intelligent and unmanned development of in-situ oil sands mining technology.

[0005] Therefore, developing an intelligent control system and control method that can achieve coordinated management, adaptive adjustment, and precise operation of the entire process of in-situ mining, separation, and backfilling of oil sands has become the key to solving existing technical problems. Summary of the Invention

[0006] The purpose of this invention is to provide a collaborative intelligent control system and method for the entire process of in-situ oil sands mining. This system can realize the collaborative linkage, adaptive control and precise operation of all links in the entire process, improve the oil sands recovery rate and backfill quality, reduce formation disturbance and operation risks, and promote the intelligent and unmanned development of in-situ oil sands mining technology.

[0007] The technical solution of this invention:

[0008] In a first aspect, the present invention provides a collaborative intelligent control system for the entire process of in-situ oil sands mining, comprising:

[0009] The sensing layer consists of sensor groups deployed in the in-situ excavation area, transportation pipeline, shaft lifting section, separation area and backfilling area of ​​the oil sands mine. It is used to collect operation parameters in real time and perform preliminary mean filtering on the collected data.

[0010] The intelligent execution layer is used to receive collaborative control instructions and automatically execute the entire process of operations.

[0011] An edge control layer, which is communicatively connected to the perception layer and the intelligent execution layer respectively, is used to receive data from the perception layer and perform local real-time control, perform Kalman filtering noise reduction and abnormal data removal preprocessing on the data, and upload the data.

[0012] The digital twin central layer, which is connected to the edge control layer, is used to construct a full-element digital twin model. Based on the data from the perception layer, the full-element digital twin model is iteratively updated in real time to realize real-time mapping, simulation, and state prediction between physical and virtual spaces.

[0013] The collaborative scheduling layer, which is connected to the central layer of the digital twin, is used to achieve dynamic matching of various operational links by using a multi-objective optimization scheduling algorithm to optimize the recovery rate, minimize energy consumption, achieve backfill compaction, and minimize formation disturbance.

[0014] Furthermore, the intelligent execution layer includes an excavation mechanism and a transport device located in the in-situ excavation area, a lifting system located at the wellhead, a multiphase separation unit located in the separation area, and a backfilling device located in the backfilling area; the material outlet of the excavation mechanism is connected to the inlet of the transport device, the outlet of the transport device is connected to the inlet of the lifting system, the outlet of the lifting system is connected to the inlet of the multiphase separation unit, the tailings outlet of the multiphase separation unit is connected to the inlet of the backfilling device, and the outlet of the backfilling device is connected to the grouting port of the goaf.

[0015] Furthermore, the sensor group includes a geological sensing sensor, an equipment status sensor, a material characteristic sensor, an environmental safety sensor, and a backfill monitoring sensor. There are multiple sensor groups, which are respectively installed in the excavation mechanism, the lifting system, the multiphase separation unit, and the backfilling equipment.

[0016] Furthermore, the edge control layer includes an excavation zone edge controller, a wellhead edge controller, a separation station edge controller, and a backfill zone edge controller. The excavation zone edge controller is communicatively connected to the sensor group in the excavation mechanism; the wellhead edge controller is communicatively connected to the sensor group in the lifting system; the separation station edge controller is communicatively connected to the sensor group in the multiphase separation unit; and the backfill zone edge controller is communicatively connected to the sensor group in the backfilling equipment.

[0017] Furthermore, the digital twin central layer includes a data fusion module, a model iteration module, a simulation and deduction module, and a visualization and monitoring module.

[0018] Furthermore, the full-element digital twin model includes a three-dimensional geomechanical model constructed based on the finite element method, an equipment operation model constructed based on equipment mechanism, and a process coupling model characterizing the process correlation of each link.

[0019] Furthermore, the multi-objective optimization scheduling algorithm of the collaborative scheduling layer includes a particle swarm optimization algorithm and a model predictive control algorithm. It constructs a predictive model for model predictive control based on the full-element digital twin model to predict the system state in the next 10 steps. Using the operating parameters of each link equipment as particle codes, the multi-objective particle swarm algorithm is used to optimize and search for control variables in the prediction time domain. The dynamic constraints of the system are processed through model predictive control, and finally the optimal collaborative control command is output.

[0020] Furthermore, the excavating mechanism, the conveying device, the lifting system, the multiphase separation unit, and the backfilling equipment are all equipped with adaptive control modules; the excavating mechanism adaptively adjusts the milling torque, speed, and excavation direction according to the formation hardness and oil content; the conveying device achieves variable frequency speed regulation and automatic unblocking according to the material flow rate and pipeline status; the lifting system automatically avoids gas lock and sand blockage problems according to the wellbore pressure and material phase fraction; the multiphase separation unit adjusts the separation parameters and chemical dosage according to the online component analysis results of the material; and the backfilling equipment achieves layered backfilling and closed-loop control of compaction based on a full-element digital twin model.

[0021] In a second aspect, the present invention provides a collaborative intelligent control method for the entire process of in-situ mining of oil sands, using the collaborative intelligent control system for the entire process of in-situ mining of oil sands as described in any one of the first aspects, comprising the following steps:

[0022] Based on geological exploration data and equipment parameters, a full-element digital twin initial model is constructed, and the process parameters, equipment operation parameters and multi-objective optimization thresholds for each link are preset.

[0023] The sensing layer collects geological, equipment, material, environmental and backfill data in real time and transmits the raw data to the edge controller in the corresponding area;

[0024] The edge controller preprocesses the received raw data, performs local closed-loop control for control requirements with high real-time requirements, and uploads the preprocessed data to the digital twin central layer.

[0025] The digital twin central layer integrates multi-source data and uses the extended Kalman filter algorithm to iteratively update the digital twin model in real time, realizing a virtual real-time mapping of the physical operation process. The simulation and deduction module completes the prediction of operation trends.

[0026] The collaborative scheduling layer is based on the updated digital twin model. It uses a multi-objective optimization algorithm to calculate the optimal operating parameters of each link's equipment, realizes dynamic flow matching of each link, generates collaborative control commands and issues them.

[0027] Each device in the intelligent execution layer receives control commands, automatically adjusts its operating status, and completes the entire process.

[0028] The perception layer collects the effect parameters after the operation and sends them back to the digital twin central layer to verify the effect and update the model.

[0029] Furthermore, in the entire process operation, the intelligent execution layer achieves closed-loop matching of flow at each stage. Specifically, the digging volume of the excavating mechanism matches the lifting capacity of the lifting system in real time, the material conveying volume of the lifting system matches the processing capacity of the multiphase separation unit in real time, and the tailings output of the multiphase separation unit matches the backfilling demand of the backfilling equipment in real time.

[0030] Compared with the prior art, the embodiments of the present invention have at least the following advantages or beneficial effects:

[0031] 1. This invention constructs a hierarchical collaborative control architecture adapted to the entire process of oil sands mining, integrating core dynamic flow matching and multi-objective optimization algorithms into the system. It breaks through the limitations of existing general architectures, solves the problems of independent operation and poor collaboration in each link of traditional technology, and realizes full-domain perception, centralized control and distributed execution of the entire process of in-situ mining-separation-backfilling of oil sands.

[0032] 2. This invention uses the digital twin central layer as the core to construct a full-element digital twin model including geology, equipment, and process. It adopts extended Kalman filtering to realize the real-time iteration of the model, ensuring the accuracy of the virtual-real mapping. It realizes the real-time mapping and simulation of the physical operation process and the virtual model, which can predict the operation trend in advance, optimize the operation parameters, and improve the scientific decision-making and predictive ability of the system.

[0033] 3. The collaborative scheduling layer of this invention adopts a multi-objective optimization algorithm that combines particle swarm optimization and model prediction. It optimizes the specific objectives of oil sands mining, realizes dynamic coordination and flow closed-loop matching of equipment in each link, ensures balanced matching of excavation, lifting, separation and backfilling, significantly improves mining efficiency and resource utilization, and reduces equipment energy consumption and failure probability.

[0034] 4. Each device in the intelligent execution layer of this invention is equipped with an adaptive control module, which can adaptively adjust the operating parameters according to the formation conditions, material characteristics and operation status, thereby achieving precise excavation, efficient separation and precise backfilling, minimizing formation disturbance, ensuring backfilling quality, and combining economy and environmental friendliness. Attached Figure Description

[0035] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0038] Example

[0039] Based on the basic theoretical framework of this invention, a collaborative intelligent control system for the entire process of in-situ oil sands mining is proposed, comprising:

[0040] The sensing layer consists of sensor groups deployed in the in-situ excavation area, transportation pipeline, shaft lifting section, separation area and backfilling area of ​​the oil sands mine. It is used to collect operation parameters in real time and perform preliminary mean filtering on the collected data.

[0041] The intelligent execution layer is used to receive collaborative control instructions and automatically execute the entire process of operations.

[0042] The edge control layer, which communicates with the perception layer and the intelligent execution layer respectively, is used to receive data from the perception layer and perform local real-time control, perform Kalman filtering noise reduction and abnormal data removal preprocessing on the data, and upload the data.

[0043] The digital twin hub layer, which communicates with the edge control layer, is used to construct a full-element digital twin model. Based on the data from the perception layer, the full-element digital twin model is iteratively updated in real time to realize real-time mapping, simulation, and state prediction between physical and virtual spaces.

[0044] The collaborative scheduling layer, which communicates with the central layer of the digital twin, is used to optimize the recovery rate, minimize energy consumption, achieve backfill compaction, and minimize formation disturbance. It employs a multi-objective optimization scheduling algorithm to output collaborative control commands and achieve dynamic matching of various operational links.

[0045] It should be noted that the complete data flow closed loop of the system in this invention is as follows: the edge controller uploads the preprocessed data to the digital twin central layer, the control commands generated by the collaborative scheduling layer are sent to each edge controller through the 5G industrial internet, and then the edge controller forwards them to the corresponding execution devices through the Profinet industrial bus to achieve complete closed-loop control; wherein, the edge controllers are deployed in a distributed manner, and each edge controller only manages the sensing and execution devices in the corresponding work area, ensuring the real-time performance of the control.

[0046] In some examples, the intelligent execution layer includes an excavation mechanism and a transport device located in the in-situ excavation area, a lifting system located at the wellhead, a multiphase separation unit located in the separation area, and a backfilling device located in the backfilling area; the material outlet of the excavation mechanism is connected to the inlet of the transport device, the outlet of the transport device is connected to the inlet of the lifting system, the outlet of the lifting system is connected to the inlet of the multiphase separation unit, the tailings outlet of the multiphase separation unit is connected to the inlet of the backfilling device, and the outlet of the backfilling device is connected to the grouting port of the goaf.

[0047] It should be noted that the above connections use pipeline connections. The transport pipeline is a wear-resistant material transport pipeline, and the backfill pipeline is a grouting pipeline. Both adopt armored insulation design to meet the requirements of long-distance underground transportation and ensure that the material will not be blocked during transportation.

[0048] In some examples, the sensor group includes geological sensing sensors, equipment status sensors, material property sensors, environmental safety sensors, and backfill monitoring sensors. There are multiple sensor groups, which are respectively installed in the excavation mechanism, lifting system, multiphase separation unit, and backfilling equipment.

[0049] It should be noted that the above-mentioned sensors are as follows: the geological sensing sensor is a logging-while-drilling sensor; the material characteristic sensor is a multiphase flow sensor; the equipment status sensor is a vibration and torque sensor; the environmental safety sensor is a combustible gas, temperature, and humidity sensor; and the backfill monitoring sensor is a compaction and settlement sensor. All sensors meet the explosion-proof requirements for mining and can work stably under complex underground working conditions.

[0050] The specific installation locations of each sensor are as follows:

[0051] The geological sensing sensor is installed at the front end of the milling head of the excavating mechanism, 5cm away from the milling teeth, and moves synchronously with the milling head; the material characteristic sensor is deployed every 100m along the transport pipeline, and one is also deployed at the shaft inlet, separation station inlet, and separation station outlet; the equipment status sensor is installed at the power end and transmission end of each executing device; the environmental safety sensor is deployed every 50m in the underground excavation area, inside the shaft, and in the separation station workshop; the backfill monitoring sensor is deployed around the goaf area, 1m away from the goaf boundary, and a grouting pressure sensor is deployed at the grouting port of the backfilling operation face.

[0052] The aforementioned sensor installation locations include environmental safety sensors deployed throughout the entire work area, ensuring safety monitoring throughout the entire process; and equipment status sensors, with corresponding sensors for each executing device, ensuring that the status of each device can be monitored in real time, eliminating any blind spots in monitoring.

[0053] Each sensor is connected to the edge controller of the corresponding work area via an intrinsically safe Profinet industrial bus for mining. The underground sensors and the edge controller are connected by armored shielded cables to achieve explosion-proof and anti-interference connection. The edge controller has a built-in 5G industrial gateway and communicates with the digital twin central layer through the 5G industrial Internet with a transmission latency of less than 10ms.

[0054] The aforementioned communication connections utilize armored shielded cables for underground connections, effectively resisting electromagnetic interference and ensuring stable data transmission. The 5G industrial gateway supports edge computing, enabling local data preprocessing while ensuring low-latency communication between the cloud and the edge, meeting real-time control requirements.

[0055] In some examples, the edge control layer includes a digging zone edge controller, a wellhead edge controller, a separation station edge controller, and a backfill zone edge controller. The digging zone edge controller is communicatively connected to a sensor group in the digging mechanism; the wellhead edge controller is communicatively connected to a sensor group in the lifting system; the separation station edge controller is communicatively connected to a sensor group in the multiphase separation unit; and the backfill zone edge controller is communicatively connected to a sensor group in the backfilling equipment.

[0056] It should be noted that the aforementioned edge controllers, including those for the excavation and backfilling areas underground, all adopt an intrinsically safe explosion-proof design for mining, meeting the requirements for explosion-proof operation underground. They can move synchronously with the operating equipment to achieve real-time control of mobile operations.

[0057] In some examples, the digital twin central layer includes a data fusion module, a model iteration module, a simulation and deduction module, and a visualization and monitoring module.

[0058] It should be noted that the modules of the above-mentioned digital twin central layer include the data fusion module, which can realize the standardized processing of multi-source heterogeneous data; the model iteration module, which can realize the real-time updating of the model; the simulation and inference module, which can realize the early prediction of operation trends; and the visualization and monitoring module, which can realize the three-dimensional visualization of the entire process. The four modules work together to realize the real-time mapping between the virtual and the real.

[0059] The specific modeling steps for the three-dimensional geomechanical model are as follows: First, data acquisition and preprocessing: collect borehole data, seismic exploration data, and well logging data from the target oil sands deposit. Well logging data includes resistivity logging, sonic logging, and density logging. Extract core geological parameters such as formation lithology, porosity, permeability, geostress, Poisson's ratio, and elastic modulus. After removing outlier data, establish a standardized geological parameter database. Second, mesh generation: use ANSYS finite element analysis software to divide the mesh into units based on the geological complexity of the reservoir. The mesh size for complex interlayer and fault areas is set to 0.5m, while the mesh size for homogeneous oil sands areas is set to 1-2m to ensure a stable mesh. The first step is to accurately characterize the spatial distribution characteristics of the strata. The second step is to construct the model by inputting parameters from the geological parameter database and defining the strata boundary conditions: the periphery of the goaf is set as a fixed boundary, and the periphery of the reservoir is set as a displacement boundary. The third step is to define the initial conditions: the initial geostress is set according to the field measured values, the initial temperature is set as the original temperature of the strata, and the elastoplastic constitutive equation is used to describe the mechanical properties of the strata. A three-dimensional geomechanical model is constructed and initialized. The fourth step is to calibrate and verify the model by collecting the field measured strata stress and displacement data and comparing them with the model calculation results. By adjusting the model parameters, the model calculation error is controlled within 1%, ensuring the consistency between the model and the physical strata.

[0060] In some examples, the full-element digital twin model includes a three-dimensional geomechanical model based on the finite element method, an equipment operation model based on equipment mechanism, and a process coupling model that characterizes the process correlation of each link.

[0061] In some examples, the multi-objective optimization scheduling algorithm of the collaborative scheduling layer includes particle swarm optimization algorithm and model predictive control algorithm. It constructs a predictive model of model predictive control based on a full-element digital twin model to predict the system state in the next 10 steps. Using the operating parameters of each link equipment as particle codes, the multi-objective particle swarm algorithm is used to optimize and search for control variables in the prediction time domain. The dynamic constraints of the system are processed through model predictive control, and finally the optimal collaborative control command is output.

[0062] The specific steps for combining the particle swarm optimization algorithm and the model predictive control algorithm, as well as the optimization target weights and core parameter settings, are as follows: Algorithm combination steps:

[0063] Predictive model construction: Based on the full-element digital twin model, a predictive model for model predictive control is constructed. The prediction time domain is set to 10 steps, with a time interval of 5 seconds between each step. The predictive model predicts the system state for the next 10 steps, including the amount of excavation, lifting, separation, backfilling, and operating parameters of each piece of equipment.

[0064] Optimization Objectives and Weights: The weights of the four optimization objectives were determined using the Analytic Hierarchy Process (AHP) to ensure that the priority of the objectives met the actual needs of oil sands mining. The specific weight allocation was as follows: maximizing recovery rate, minimizing energy consumption, achieving backfill compaction, and minimizing formation disturbance. A multi-objective optimization objective function was constructed: minF = -0.35×η+ 0.25×E + 0.25×(1-ρ) + 0.15×S, where η is the recovery rate, E is the energy consumption per unit production, ρ is the backfill compaction, and S is the formation subsidence. This objective function was used to achieve coordinated optimization of multiple objectives.

[0065] Particle coding rules: Real number coding is adopted. Each particle corresponds to a complete set of control variable schemes. The particle dimension is 5, corresponding to 5 core control variables. The specific coding correspondence is as follows: Particle 1st dimension = milling torque of excavating mechanism (unit: N·m), value range 500~1500; 2nd dimension = conveying device speed (unit: m / s), value range 0.3~1.0; 3rd dimension = lifting system flow rate (unit: m³ / s). 3 / h), with a value range of 10~30; the fourth dimension = centrifuge speed of the multiphase separation unit (unit: r / min), with a value range of 1500~2500; the fifth dimension = grouting pressure of the backfilling equipment (unit: MPa), with a value range of 2~5.

[0066] Particle swarm parameter settings: the number of particles is set to 50, the maximum number of iterations is set to 100, the inertia weight ω is linearly decreased from 0.9 to 0.4 (to ensure global search capability in the early stage and improve local search accuracy in the later stage), and the acceleration factors c1 (self-cognition factor) and c2 (social cognition factor) are both set to 2 to ensure the convergence speed and optimization accuracy of the algorithm.

[0067] Collaborative optimization and command output: A multi-objective particle swarm optimization algorithm is used to perform a global optimization search on the control variables in the prediction time domain to obtain a Pareto optimal solution set; the dynamic constraints of the model predictive control system are processed, including equipment operating parameter threshold constraints (such as the excavation torque not exceeding the equipment's rated torque) and material flow matching constraints (such as the excavation volume ≤ lifting capacity). The optimal control parameters are selected from the Pareto optimal solution set, and collaborative control commands are generated and sent to the edge control layer.

[0068] It should be noted that the above algorithms combine logic: Model predictive control is responsible for handling the dynamic constraints of the system, while particle swarm optimization is responsible for global multi-objective optimization. The combination of the two ensures both the global optimality of the optimization and the dynamic stability of the system, thus meeting the requirements of real-time control.

[0069] In some examples, the excavating mechanism, conveying device, lifting system, multiphase separation unit, and backfilling equipment are all equipped with adaptive control modules; the excavating mechanism adaptively adjusts the milling torque, speed, and excavation direction based on formation hardness and oil content; the conveying device achieves variable frequency speed regulation and automatic unblocking based on material flow rate and pipeline status; the lifting system automatically avoids air lock and sand blockage problems based on wellbore pressure and material phase fraction; the multiphase separation unit adjusts separation parameters and chemical dosage based on online material composition analysis results; and the backfilling equipment achieves layered backfilling and closed-loop control of compaction based on a full-element digital twin model.

[0070] It should be noted that the adaptive control modules of the above-mentioned devices can operate independently or receive global instructions from the collaborative scheduling layer, realizing a combination of local adaptation and global collaboration, thus ensuring the adaptability of the devices to operating conditions and the collaboration of the entire process.

[0071] like Figure 1 As shown, to better understand the present invention, the present invention also provides a collaborative intelligent control method for the entire process of in-situ oil sands mining. This method uses the aforementioned collaborative intelligent control system for the entire process of in-situ oil sands mining and includes the following steps:

[0072] Based on geological exploration data and equipment parameters, a full-element digital twin initial model is constructed, and the process parameters, equipment operation parameters and multi-objective optimization thresholds for each link are preset.

[0073] The sensing layer collects geological, equipment, material, environmental and backfill data in real time and transmits the raw data to the edge controller in the corresponding area;

[0074] The edge controller preprocesses the received raw data, performs local closed-loop control for control requirements with high real-time requirements, and uploads the preprocessed data to the digital twin central layer.

[0075] The digital twin central layer integrates multi-source data and uses the extended Kalman filter algorithm to iteratively update the digital twin model in real time, realizing a virtual real-time mapping of the physical operation process. The simulation and deduction module completes the prediction of operation trends.

[0076] The collaborative scheduling layer is based on the updated digital twin model. It uses a multi-objective optimization algorithm to calculate the optimal operating parameters of each link's equipment, realizes dynamic flow matching of each link, generates collaborative control commands and issues them.

[0077] Each device in the intelligent execution layer receives control commands, automatically adjusts its operating status, and completes the entire process.

[0078] The perception layer collects the effect parameters after the operation and sends them back to the digital twin central layer to verify the effect and update the model.

[0079] It should be noted that in the above methods, safety management and fault handling are carried out throughout the entire process, which can realize automatic fault handling and safety linkage control, ensuring the safety of unmanned operation; at the same time, effect feedback and model iteration can realize the self-learning of the model, ensuring the long-term adaptability of the system.

[0080] In some examples, the intelligent execution layer achieves closed-loop matching of flow at each stage of the entire process. Specifically, the digging volume of the excavating mechanism matches the lifting capacity of the lifting system in real time, the material conveying volume of the lifting system matches the processing capacity of the multiphase separation unit in real time, and the tailings output of the multiphase separation unit matches the backfilling demand of the backfilling equipment in real time.

[0081] To better understand the present invention, the following tests were conducted based on the system and method provided by the present invention.

[0082] The oil sands layer in this test project is located at a depth of 1000m underground. The in-situ excavation area is set within this oil sands layer, and the goaf formed after the excavation is completed will become the backfill area. A vertically arranged shaft extends from the surface to the underground excavation area. The shaft contains DN200 transport pipelines, DN150 backfill pipelines, and armored power and communication cables. In the surface area, a shaft control cabinet is located at the shaft head, and a surface separation station is located 50m away from the shaft head. The separation station has a processing capacity of 50m³. 3 / h.

[0083] In the sensing layer, a geological sensing sensor is installed at the front end of the milling head of the intelligent excavation mechanism, 5cm away from the milling teeth. It moves synchronously with the excavation mechanism to collect the formation stress, displacement, oil content, porosity, temperature and pressure parameters at the excavation location in real time.

[0084] A material characteristic sensor is deployed every 100m along the transport pipeline, and one is also deployed at the wellhead, the separation station inlet, and the separation station outlet to collect parameters such as phase fraction, pipeline pressure, and flow rate of the oil, gas, water, and rock mixture.

[0085] Equipment status sensors are deployed at the power and transmission ends of each actuator to collect parameters such as speed, torque, vibration, wear, power, and liquid level.

[0086] An environmental safety sensor is deployed every 50m in the underground excavation area, inside the shaft, and in the separation station workshop to collect safety parameters such as combustible gas concentration, leakage status, temperature, and humidity.

[0087] Around the goaf, 1m away from the goaf boundary, backfill monitoring sensors are deployed. At the same time, grouting pressure sensors are deployed at the grouting ports of the backfilling operation face to collect parameters such as goaf volume, backfill material density, ground settlement, and grouting pressure.

[0088] All sensors are connected to the corresponding edge controllers via an intrinsically safe Profinet industrial bus for mining. The underground connection uses armored shielded cables to meet the requirements for explosion protection and anti-interference in underground environments. After the raw data is collected, it is first subjected to preliminary mean filtering to remove obvious abnormal spikes, and then transmitted to the corresponding edge controller.

[0089] In the intelligent execution layer, intelligent excavation mechanisms and intelligent conveying devices are deployed in the excavation area, with the material outlet of the intelligent excavation mechanism connected to the inlet of the conveying pipeline; an intelligent lifting system is deployed at the wellhead, with the outlet of the conveying pipeline connected to the inlet of the intelligent lifting system; an intelligent multiphase separation unit is deployed in the separation station, with the outlet of the intelligent lifting system connected to the inlet of the intelligent multiphase separation unit; and intelligent backfilling equipment is deployed in the backfilling area, with the tailings outlet of the intelligent multiphase separation unit connected to the inlet of the backfilling pipeline, the outlet of the backfilling pipeline connected to the inlet of the intelligent backfilling equipment, and the outlet of the intelligent backfilling equipment connected to the grouting port of the goaf.

[0090] Each device communicates with the edge controller of its corresponding area, receives control commands forwarded by the edge controller, automatically adjusts operating parameters, and is equipped with an adaptive control module that can adaptively adjust operating parameters according to working conditions.

[0091] The edge control layer deploys four edge controllers, each corresponding to a sensing and execution device in each area:

[0092] The excavation area edge controller is deployed at the rear end of the underground excavation mechanism, moving synchronously with it. It adopts an intrinsically safe explosion-proof design for mining and is connected to all sensors, intelligent excavation mechanisms, and conveying devices in the excavation area. The wellhead edge controller is deployed in the wellhead control cabinet on the surface and is connected to the sensors inside the shaft, the valve equipment at the wellhead, and the intelligent lifting system. The separation station edge controller is deployed in the central control room of the surface separation station and is connected to all equipment and sensors in the separation station. The backfilling area edge controller is deployed in the control cabinet at the underground backfilling face, moving synchronously with the backfilling operation. It adopts an intrinsically safe explosion-proof design for mining and is connected to the sensors and intelligent backfilling equipment in the backfilling area.

[0093] The edge controller receives raw data from the perception layer and performs preprocessing such as Kalman filtering for noise reduction and abnormal data removal. For control requirements with high real-time requirements, such as pipeline blockage unblocking control and equipment overload protection, it directly executes millisecond-level local closed-loop control without waiting for instructions from the central layer, ensuring real-time control.

[0094] Meanwhile, the edge controller has a built-in 5G industrial gateway, which uploads the pre-processed data to the digital twin hub layer in the cloud via the 5G industrial internet with a transmission latency of less than 10ms. It also receives collaborative control commands from the hub layer and forwards them to the corresponding execution devices via the Profinet industrial bus.

[0095] The digital twin central layer is deployed on a cloud server and communicates with all edge controllers. First, based on borehole, seismic, and well logging data from geological exploration, a three-dimensional geomechanical model is constructed using the finite element method to characterize the spatial distribution of formation parameters. Then, based on the CAD drawings of the equipment, a mechanism model of each piece of equipment is constructed to characterize the operating characteristics of the equipment. At the same time, a process coupling model of excavation-transportation-lifting-separation-backfilling is constructed to characterize the process correlation of each link. Finally, it is integrated to form a full-element digital twin initial model of oil sands geological body-well network-equipment-process-backfilling area.

[0096] The core implementation logic of the extended Kalman filter algorithm is as follows:

[0097] The state equation is defined as: Xk = f(Xk-1, Uk) + Wk, where Xk is the model state vector at time k, containing 12 core state parameters such as formation stress, formation displacement, equipment speed, torque, material flow rate, and backfill compaction; Xk-1 is the model state vector at time k-1; Uk is the control input vector at time k, containing control parameters such as excavation torque, transport speed, and lifting flow rate; f(·) is the state transition function, which uses a nonlinear function to characterize the variation law of each state parameter; Wk is the process noise, which follows a Gaussian distribution N(0, Q), and Q is the process noise covariance matrix, which is calibrated based on field measured data, with the diagonal elements ranging from 1e-4 to 1e-2.

[0098] The observation equation is defined as: Zk = h(Xk) + Vk, where Zk is the observation vector collected by the sensing layer at time k, which corresponds one-to-one with the model state vector and is obtained in real time by each sensor; h(·) is the observation function, which is used to map the model state vector to the observation value; Vk is the observation noise, which follows a Gaussian distribution N(0, R), and R is the observation noise covariance matrix, which is set according to the sensor accuracy, and the diagonal elements take values ​​from 1e-5 to 1e-3.

[0099] Iterative update steps: Prediction stage: Calculate the predicted value of the model state at time k, Xk|k-1 = f(Xk-1|k-1,Uk), and calculate the prediction error covariance matrix Pk|k-1 = FkPk-1|k-1Fk^T + Q, where Fk is the state transition matrix, obtained by taking the partial derivative of the state transition function f(·) with respect to Xk-1, and Pk-1|k-1 is the model state estimation error covariance matrix at time k-1.

[0100] Update phase: Calculate the Kalman gain Kk = Pk|k-1Hk^T(HkPk-1|k-1Hk^T + R), where Hk is the observation matrix, obtained by taking the partial derivative of the observation function h(·) with respect to Xk; then calculate the model state estimate at time k, Xk|k = Xk|k-1+ Kk(Zk - h(Xk|k-1)), and calculate the model state estimate error covariance matrix at time k, Pk|k = (I - KkHk)Pk|k-1, where I is the identity matrix; complete an iterative update every 5 seconds to ensure real-time synchronization between the virtual model and the physical operation process.

[0101] The digital twin central layer comprises four modules: a data fusion module, which integrates multi-source heterogeneous data uploaded from various edge controllers to build a unified real-time data pool and achieve standardized data processing; a model iteration module, which receives real-time data and uses the extended Kalman filter algorithm to update the state parameters of the digital twin model in real time, using the perceived data as observations, completing an iteration every 5 seconds to ensure that the state error between the virtual model and the physical entity is less than 1%; a simulation and deduction module, which, based on the updated digital twin model, performs virtual simulations of mining parameter adjustments, equipment failure handling, and geological evolution trends, predicting the operational status within the next hour and outputting pre-decision solutions; and a visualization and monitoring module, which provides a 3D visualization of the entire operational process and supports real-time monitoring and operation by remote terminals.

[0102] The collaborative scheduling layer is deployed on a cloud server and communicates with the digital twin central layer to obtain the state data of the digital twin model. With the optimization objectives of maximizing recovery rate, minimizing energy consumption, achieving backfill compaction standards, and minimizing formation disturbance, a multi-objective optimization scheduling algorithm combining particle swarm optimization and model predictive control is employed.

[0103] First, a predictive model for model predictive control is constructed based on a digital twin model to predict the system state for the next 10 steps. Then, the operating parameters of each stage of the equipment—excavation torque, conveying speed, lifting flow rate, separation speed, and grouting pressure—are encoded as particles, with each particle representing a set of candidate control schemes. Next, a multi-objective particle swarm optimization algorithm is used to iteratively search for the optimal control parameters in the solution space. Simultaneously, model predictive control processes the dynamic constraints of the system, the operating parameter range of the equipment, and the matching constraints of material flow rate, ultimately outputting the optimal coordinated control command.

[0104] The generated control commands are sent to each edge controller via the 5G Industrial Internet. The edge controllers then forward the commands to the corresponding execution devices via the Profinet industrial bus. This algorithm enables dynamic matching of excavation volume with lifting capacity, lifting volume with separation and processing capacity, and separation tailings volume with backfilling requirements. This ensures real-time balance of flow in each stage and avoids disconnection or overload issues.

[0105] Based on the above system, the control method in this test specifically includes the following steps:

[0106] Based on geological exploration data and equipment parameters, a full-element digital twin initial model was constructed, with preset excavation torque of 1000 N·m, rotation speed of 50 r / min, and lifting flow rate of 20 m³ / min. 3 Initial parameters included: recovery rate ≥ 80%, energy consumption ≤ 20 kWh / ton of oil, and formation subsidence ≤ 5 mm.

[0107] The sensors in the sensing layer collect data in real time. The oil content of the formation in the excavation area is 15%, the stress is 20 MPa, and the flow rate of the transport pipeline is 20 m³ / h. 3 / h, pressure 5MPa, crude oil water content in the separation unit 0.4%, backfill density 95%, the raw data is transmitted to the corresponding edge controller via Profinet industrial bus.

[0108] The edge controller performs Kalman filtering on the data to reduce noise and eliminate abnormal data. It detects a slight increase in the pressure of the conveyor pipeline, which is judged to be a slight blockage trend. It directly executes local control commands to adjust the conveyor speed from 0.5m / s to 0.6m / s to achieve automatic unblocking. At the same time, the pre-processed data is uploaded to the digital twin central layer through the 5G industrial internet.

[0109] The digital twin central layer integrates multi-source data and uses extended Kalman filtering to update the digital twin model, mapping the operational status of each stage in real time. The simulation module predicts that if the current parameters are maintained, the ground stress in the excavation area will rise to 22 MPa after 1 hour, with no risk of collapse.

[0110] Based on the updated model, the collaborative scheduling layer runs a multi-objective optimization algorithm to calculate the optimal control command: the excavator torque is adjusted to 1100 N·m, and the lifting flow rate is maintained at 20 m³ / s. 3 / h, the centrifuge speed of the separation unit is adjusted to 2100r / min, and the grouting pressure of the backfilling equipment is adjusted to 3MPa, so as to achieve dynamic matching of excavation volume, lifting volume, separation volume and backfilling volume, and then the instructions are sent to each edge controller.

[0111] Each execution device receives instructions forwarded by the edge controller through the Profinet industrial bus, automatically adjusts operating parameters, and completes in-situ excavation, material transportation, wellbore lifting, oil, gas, water and rock separation and backfilling operations. The flow rate of each link is matched in real time, without overload or disconnection problems.

[0112] The data collected by the sensing layer after the operation showed that the crude oil water content was 0.35%, the backfill compaction was 96%, and the formation stress was 21 MPa, all of which met the preset thresholds. The digital twin central layer updated the model based on the feedback data, realizing the model's self-iteration.

[0113] The system monitors the equipment and environmental status throughout the process. If no fault is detected, normal operation continues. If the concentration of combustible gas exceeds the standard, the linkage control will be triggered immediately to suspend all operations, close relevant valves, start explosion-proof ventilation equipment, and adjust the backfilling equipment for emergency sealing.

[0114] Test Example 1

[0115] This test example 1 is a test of the synergistic effect of the entire process. In the in-situ mining project of continental sedimentary oil sands, the system of this invention and the traditional independent control system were compared and tested. The test period was 30 days, and the specific test process is as follows:

[0116] For traditional independent control systems: System composition: Traditional independent control systems do not adopt a hierarchical collaborative architecture, lack a digital twin central layer and a collaborative scheduling layer, and consist only of independent control units, sensors, and execution devices for each link. Specifically, the excavation area uses an independent excavation control box, equipped only with basic torque and speed control functions, and a single geological sensor that only collects formation hardness data; the lifting system uses an independent wellhead control cabinet, only enabling manual setting and adjustment of the lifting flow rate; the multiphase separation unit uses an independent separation station control system, which can only perform separation operations according to preset parameters and cannot be linked with the lifting system; the backfilling equipment uses an independent backfilling control device, which can only operate at a fixed grouting pressure and lacks closed-loop control for compaction; sensor data from each link is stored independently, lacking multi-source data fusion capabilities, and failing to achieve full-process data sharing and collaborative decision-making.

[0117] Control methods: Traditional independent control systems adopt manual operation and fixed parameter control mode, with each link operating independently and no dynamic flow matching mechanism; the excavation volume is set by the operator based on experience and cannot be adjusted in real time according to the lifting capacity; the lifting flow rate is fixed and requires manual intervention to adjust when it is out of sync with the processing capacity of the separation unit; the discharge of separation tailings is not linked to the backfilling requirements, and the backfilling equipment needs to be started manually after the tailings volume is counted; the backfill compaction is only tested by manual sampling and cannot be controlled in real time, and the formation disturbance is entirely based on experience prediction without an early warning mechanism.

[0118] Test Procedure: Two test wells with the same geological conditions and depth of 1000m were selected from the same oil sands block. The oil layer thickness and oil saturation were consistent. The system of this invention (test well 1) and the traditional independent control system (test well 2) were used for production operations. The test period was 30 days, with continuous operation for 24 hours a day. During the test, the core performance indicators of the two test wells were recorded, including recovery rate, energy consumption per unit production, backfill compaction, formation subsidence, number of equipment failures and number of pipeline blockages. During the test, the geological conditions, operation time and material characteristics were kept consistent.

[0119] After the testing period ended, the core performance indicators of the two sets of test wells were compiled, and the specific comparison data are shown in the table below:

[0120]

[0121] The 30-day comparative test shows that the system of this invention has achieved significant improvements in core performance indicators compared with traditional independent control systems: recovery rate increased by 13.2%, effectively improving the utilization rate of oil sands resources; energy consumption per unit output decreased by 29.2%, significantly reducing mining costs and improving operational economics; backfill compaction increased by 8.4%, and formation subsidence decreased by 59.0%, effectively reducing formation disturbance, avoiding geological disaster risks, and improving operational safety; the number of equipment failures and pipeline blockages decreased by 86.7% and 91.7% respectively, reducing equipment maintenance costs and operation downtime, and improving operational continuity; at the same time, the number of manual operators decreased by 75.0%, significantly reducing labor costs and promoting the unmanned development of in-situ oil sands mining.

[0122] The test results fully verified the effectiveness and superiority of the system of the present invention. Through full-process collaborative management and control, real-time digital twin mapping, multi-objective optimized scheduling and equipment adaptive control, it effectively solves the problems of independent operation, poor coordination and low control accuracy of each link in traditional technology. It can realize precise, efficient, safe and energy-saving operation of the entire process of in-situ mining of oil sands, and fully meet the intelligent development needs of in-situ mining of oil sands.

[0123] Test Example 2

[0124] To further verify the adaptability and stability of the system of the present invention under complex geological conditions, test example 2 was set up. An oil sands block with developed faults and heterogeneous strata was selected. The well depth was 1200m, containing complex interlayers and large fluctuations in formation stress. The system of the present invention was used for a 15-day continuous operation test, focusing on monitoring the system's collaborative control effect, model iteration accuracy and equipment self-adaptability under complex geological conditions.

[0125] The test block contains two small faults, the lithology of the formation is heterogeneous, the oil content fluctuates between 12% and 18%, and the formation stress fluctuates between 18 and 25 MPa. During the operation, the system simulates complex working conditions such as sudden changes in formation stress and fluctuations in material properties to test its emergency response capability and adaptive control capability.

[0126] The system is started according to the control method provided by this invention, and a full-element digital twin model including faults and complex interlayers is constructed with preset optimization thresholds. The perception layer collects parameters such as formation stress, oil content, and material phase fraction in real time. The edge control layer performs local preprocessing and real-time control. The digital twin central layer completes a model iteration every 5 seconds. The collaborative scheduling layer dynamically outputs optimized control commands. The intelligent execution layer adaptively adjusts the operating parameters of the equipment. During the test, the model iteration error, emergency response time, equipment operation stability, and core performance indicators are recorded daily.

[0127] During testing, the state error between the digital twin model and the physical entity was consistently kept within 0.8%. Even when the underlying stress suddenly increased from 20 MPa to 24 MPa, the model could complete the iterative update within 10 seconds, accurately mapping the physical operation state. The simulation module predicted the trend of formation stress change 15 minutes in advance, providing a reliable basis for collaborative scheduling.

[0128] When the simulated pipeline experienced a sudden pressure surge from 5MPa to 8MPa due to blockage, the edge control layer detected the anomaly within 50ms and immediately implemented local unblocking control. The speed of the conveying device was adjusted from 0.6m / s to 0.8m / s, while the coordination and scheduling layer adjusted the torque of the excavating mechanism from 1100N·m to 900N·m to reduce material input. The pipeline unblocking was completed within 120s without causing any operational interruption.

[0129] During the 15-day test, the system achieved an average recovery rate of 83.7%, an energy consumption of 19.5 kWh / ton of oil, a backfill compaction rate of 95.9%, and a maximum formation settlement of 4.1 mm. No equipment failures or safety accidents occurred. The equipment's adaptive module was able to adjust operating parameters in real time based on changes in formation oil content and stress, ensuring stable operation.

[0130] Test Example 2 shows that the system of the present invention still has high stability and adaptability under complex geological conditions such as fault development and heterogeneous strata. The iterative accuracy and emergency response capability of the digital twin model can meet the operational requirements of complex working conditions. The equipment adaptive control module and the whole-process collaborative scheduling mechanism can effectively cope with sudden situations such as sudden changes in stratum stress and fluctuations in material properties, ensuring the continuity and safety of operations, and further verifying the practicality and reliability of the system of the present invention.

[0131] The above are merely preferred embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A collaborative intelligent control system for the entire process of in-situ oil sands mining, characterized in that, include: The sensing layer consists of sensor groups deployed in the in-situ excavation area, transportation pipeline, shaft lifting section, separation area and backfilling area of ​​the oil sands mine. It is used to collect operation parameters in real time and perform preliminary mean filtering on the collected data. The intelligent execution layer is used to receive collaborative control instructions and automatically execute the entire process of operations. An edge control layer, which is communicatively connected to the perception layer and the intelligent execution layer respectively, is used to receive data from the perception layer and perform local real-time control, perform Kalman filtering noise reduction and abnormal data removal preprocessing on the data, and upload the data. The digital twin central layer, which is connected to the edge control layer, is used to construct a full-element digital twin model. Based on the data from the perception layer, the full-element digital twin model is iteratively updated in real time to realize real-time mapping, simulation, and state prediction between physical and virtual spaces. The collaborative scheduling layer, which is connected to the central layer of the digital twin, is used to achieve dynamic matching of various operational links by using a multi-objective optimization scheduling algorithm to optimize the recovery rate, minimize energy consumption, achieve backfill compaction, and minimize formation disturbance.

2. The system according to claim 1, characterized in that, The intelligent execution layer includes an excavation mechanism and a transport device located in the in-situ excavation area, a lifting system located at the wellhead, a multiphase separation unit located in the separation area, and a backfilling device located in the backfilling area. The material outlet of the excavation mechanism is connected to the inlet of the transport device, the outlet of the transport device is connected to the inlet of the lifting system, the outlet of the lifting system is connected to the inlet of the multiphase separation unit, the tailings outlet of the multiphase separation unit is connected to the inlet of the backfilling device, and the outlet of the backfilling device is connected to the grouting port of the goaf.

3. The system according to claim 2, characterized in that, The sensor group includes geological sensing sensors, equipment status sensors, material characteristic sensors, environmental safety sensors, and backfill monitoring sensors. There are multiple sensor groups, which are respectively installed in the excavation mechanism, the lifting system, the multiphase separation unit, and the backfilling equipment.

4. The system according to claim 3, characterized in that, The edge control layer includes an excavation zone edge controller, a wellhead edge controller, a separation station edge controller, and a backfill zone edge controller. The excavation zone edge controller is communicatively connected to the sensor group in the excavation mechanism; the wellhead edge controller is communicatively connected to the sensor group in the lifting system; the separation station edge controller is communicatively connected to the sensor group in the multiphase separation unit; and the backfill zone edge controller is communicatively connected to the sensor group in the backfilling equipment.

5. The system according to claim 1, characterized in that, The digital twin central layer includes a data fusion module, a model iteration module, a simulation and deduction module, and a visualization and monitoring module.

6. The system according to claim 5, characterized in that, The full-element digital twin model includes a three-dimensional geomechanical model based on the finite element method, an equipment operation model based on equipment mechanism, and a process coupling model that characterizes the process correlation of each link.

7. The system according to claim 1, characterized in that, The multi-objective optimization scheduling algorithm of the collaborative scheduling layer includes particle swarm optimization algorithm and model predictive control algorithm. It constructs a predictive model of model predictive control based on the full-element digital twin model to predict the system state in the next 10 steps. Using the operating parameters of each link equipment as particle codes, the multi-objective particle swarm algorithm is used to optimize and search for control variables in the prediction time domain. The dynamic constraints of the system are processed through model predictive control, and finally the optimal collaborative control command is output.

8. The system according to claim 2, characterized in that, The excavating mechanism, the conveying device, the lifting system, the multiphase separation unit, and the backfilling equipment are all equipped with adaptive control modules. The excavating mechanism adaptively adjusts the milling torque, speed, and excavation direction according to the formation hardness and oil content. The conveying device achieves variable frequency speed regulation and automatic unblocking based on material flow rate and pipeline status. The lifting system automatically avoids gas lock and sand blockage problems based on wellbore pressure and material phase fraction. The multiphase separation unit adjusts separation parameters and chemical dosage based on online material composition analysis results. The backfilling equipment achieves layered backfilling and closed-loop control of compaction based on a full-element digital twin model.

9. A collaborative intelligent control method for the entire process of in-situ oil sands mining based on any one of claims 1-8, characterized in that, Includes the following steps: Based on geological exploration data and equipment parameters, a full-element digital twin initial model is constructed, and the process parameters, equipment operation parameters and multi-objective optimization thresholds for each link are preset. The sensing layer collects geological, equipment, material, environmental and backfill data in real time and transmits the raw data to the edge controller in the corresponding area; The edge controller preprocesses the received raw data, performs local closed-loop control for control requirements with high real-time requirements, and uploads the preprocessed data to the digital twin central layer. The digital twin central layer integrates multi-source data and uses the extended Kalman filter algorithm to iteratively update the digital twin model in real time, realizing a virtual real-time mapping of the physical operation process. The simulation and deduction module completes the prediction of operation trends. The collaborative scheduling layer is based on the updated digital twin model. It uses a multi-objective optimization algorithm to calculate the optimal operating parameters of each link's equipment, realizes dynamic flow matching of each link, generates collaborative control commands and issues them. Each device in the intelligent execution layer receives control commands, automatically adjusts its operating status, and completes the entire process. The perception layer collects the effect parameters after the operation and sends them back to the digital twin central layer to verify the effect and update the model.

10. The method according to claim 9, characterized in that, In the entire process, the intelligent execution layer achieves closed-loop matching of flow at each stage. Specifically, the digging volume of the excavating mechanism matches the lifting capacity of the lifting system in real time, the material conveying volume of the lifting system matches the processing capacity of the multiphase separation unit in real time, and the tailings output of the multiphase separation unit matches the backfilling demand of the backfilling equipment in real time.

Images