Conveying control methods and systems for belt conveyors

Abstract

This invention discloses a conveying control method and system for belt conveyors, relating to the field of conveying control technology. The method includes: performing multi-parameter control collaborative decision-making based on material conveying scenario data of the belt conveyor to obtain an initial conveying control scheme; executing simulated conveying and acquiring a dataset; based on this dataset, performing start-stop impact risk detection, response anomaly risk mining, and fault risk detection to obtain three types of risk detection results; and finally, collaboratively optimizing the initial conveying control scheme based on these risk detection results to obtain an optimized conveying control scheme. This achieves the technical effects of diversified control strategies, improved risk warning mechanisms, and enhanced optimization depth.

Description

Technical Field

[0001] This invention relates to the field of conveying control technology, and in particular to a conveying control method and system for belt conveyors. Background Technology

[0002] Belt conveyors, as a commonly used continuous conveying equipment, are widely used in mining, ports, logistics, and other fields. Traditional belt conveyor control methods mainly rely on monitoring and adjusting single parameters (such as material flow rate and belt speed) to achieve basic material conveying functions. For example, some control systems adjust the conveyor belt speed solely based on feedback signals from material flow sensors, or adjust the tensioning device based solely on motor load current. These methods fail to fully consider the complexity of material conveying scenarios and the interrelationships between multiple parameters, making it difficult to meet the high requirements of modern industrial production for intelligent, safe, and stable conveying systems. Summary of the Invention

[0003] This invention provides a conveying control method and system for belt conveyors to solve the technical problems of single control strategies, lack of risk warning, and insufficient optimization depth in the prior art, and to achieve the technical effects of online monitoring, complete monitoring, and good early warning capability.

[0004] In a first aspect, the present invention provides a conveying control method for a belt conveyor, wherein the conveying control method for the belt conveyor includes:

[0005] Based on the material conveying scenario data of the belt conveyor, multi-parameter conveying control collaborative decision-making is performed to obtain the initial conveying control scheme.

[0006] The simulated conveying of the belt conveyor is executed according to the initial conveying control scheme to obtain a simulated conveying dataset.

[0007] The belt conveyor is subjected to start-up and shutdown impact risk detection based on the simulated conveying dataset, and the first risk detection result of the conveying is obtained.

[0008] Based on the simulated transport dataset, the belt conveyor is subjected to response anomaly risk mining to obtain the second transport risk detection result.

[0009] Based on the simulated transport dataset, the belt conveyor is subjected to fault risk detection to obtain the third risk detection result of the transport.

[0010] Based on the first risk detection result, the second risk detection result, and the third risk detection result, the initial transportation control scheme is collaboratively optimized to obtain an optimized transportation control scheme.

[0011] In one feasible implementation, the belt conveyor is subjected to start-up and shutdown impact risk detection based on the simulated conveying dataset to obtain a first risk detection result for the conveying process, including:

[0012] Start-up and stop feature identification is performed based on the simulated transport dataset to obtain start-up section feature data and stop section feature data.

[0013] Impact risk detection is performed based on the characteristic data of the starting section to obtain the starting impact risk detection result.

[0014] Impact risk detection is performed based on the characteristic data of the shutdown section to obtain the shutdown impact risk detection results.

[0015] The first risk detection result for transportation is generated by combining the startup impact risk detection result and the shutdown impact risk detection result.

[0016] In one feasible implementation, impact risk detection is performed based on the start-up section characteristic data to obtain the start-up impact risk detection result, including:

[0017] Based on the start-up impact risk assessment records of the belt conveyor, a start-up section sample set is obtained, along with the corresponding electrical impact risk sample set, mechanical impact risk sample set, and material impact risk sample set.

[0018] Based on the starting section sample set and the electrical shock risk sample set, a starting electrical shock risk mapping space is constructed.

[0019] Based on the starting section sample set and the mechanical impact risk sample set, a starting mechanical impact risk mapping space is constructed.

[0020] Based on the starting section sample set and the material shock risk sample set, a starting material shock risk mapping space is constructed.

[0021] The starting electrical shock risk mapping space, the starting mechanical shock risk mapping space, and the starting material shock risk mapping space are integrated to generate a starting shock risk detection space.

[0022] The startup segment feature data is input into the startup impact risk detection space to generate the startup impact risk detection result.

[0023] In one feasible implementation, the belt conveyor is subjected to response anomaly risk mining based on the simulated conveying dataset to obtain a second risk detection result for the conveyor, including:

[0024] Tension risk analysis is performed based on the simulated transport dataset to obtain the tension risk coefficient.

[0025] Based on the simulated transport dataset, the risk of deviation and slippage is analyzed to obtain the risk coefficient of deviation and slippage.

[0026] Based on the simulated conveying dataset, the risk of idler jamming is analyzed to obtain the idler jamming risk coefficient.

[0027] The tension risk coefficient, the deviation and slippage risk coefficient, and the idler roller jamming risk coefficient are added to the second risk detection result of the conveyor.

[0028] In one feasible implementation, tension risk analysis is performed based on the simulated transport dataset to obtain the tension risk coefficient, including:

[0029] Based on the simulated transport dataset, tension simulation data is extracted.

[0030] Based on the tension sample dataset and the tension risk sample dataset, train the first tension risk detection model, the second tension risk detection model, and the third tension risk detection model.

[0031] The tension risk detection first model, the tension risk detection second model, and the tension risk detection third model are sorted in descending order of detection accuracy to obtain the risk detection model arrangement distribution.

[0032] Based on the arrangement and distribution of the risk detection model, determine the tension risk detection base learner and the tension risk detection meta learner.

[0033] Based on the tension risk detection base learner, the output reinforcement learning of the tension risk detection meta learner is performed to obtain the tension risk parser.

[0034] The tension simulation data is input into the tension risk analyzer to obtain the tension risk coefficient.

[0035] In one feasible implementation, the belt conveyor is subjected to fault risk detection based on the simulated conveying dataset to obtain a third risk detection result for the conveyor, including:

[0036] Obtain the conveying monitoring record set and the fault risk record set of the belt conveyor.

[0037] Using the conveyor monitoring record set as input information and the fault risk record set as output information, train K conveyor fault risk detection models, where K is a positive integer greater than 1.

[0038] The simulated transport dataset is input into the K conveyor fault risk detection models to obtain K fault risk coefficients. The mean of the K fault risk coefficients is calculated to generate the third risk detection result of the transport.

[0039] In one feasible implementation, multi-parameter conveying control collaborative decision-making is performed based on material conveying scenario data of the belt conveyor to obtain an initial conveying control scheme, including:

[0040] The conveying control constraint information of the belt conveyor is collected to obtain a multi-dimensional conveying control constraint domain.

[0041] Based on the material conveying scenario data, the multidimensional conveying control constraint domain is adjusted according to scenario association to establish a multidimensional control constraint optimization domain.

[0042] Based on the multidimensional control constraint optimization domain, multi-parameter control collaborative decision-making is performed on the belt conveyor to obtain the conveyor control decision set.

[0043] Based on the transport control decision set, energy consumption minimization optimization is performed to generate the initial transport control scheme.

[0044] In one feasible implementation, the multi-dimensional conveying control constraint domain is adjusted based on the material conveying scenario data to establish a multi-dimensional control constraint optimization domain, including:

[0045] Based on the material conveying scenario data, a conveying control scheme is retrieved to obtain a set of retrieved conveying control schemes.

[0046] Based on the retrieved transport control scheme set, interval feature analysis is performed to obtain the multidimensional transport control retrieval domain.

[0047] Based on the multidimensional transport control retrieval domain, the multidimensional transport control constraint domain is mapped and optimized to generate the multidimensional control constraint optimization domain.

[0048] In one feasible implementation, a simulated conveying operation of the belt conveyor is performed according to the initial conveying control scheme to obtain a simulated conveying dataset, including:

[0049] A 3D reconstruction of the belt conveyor was performed to obtain a conveyor model.

[0050] Based on the material conveying scenario data, the initial conveying control scheme is simulated and executed according to the conveyor model to obtain the simulated conveying dataset.

[0051] Secondly, the present invention also provides a conveying control system for a belt conveyor, wherein the conveying control system for the belt conveyor includes:

[0052] The initial scheme acquisition module is used to make multi-parameter conveying control collaborative decisions based on the material conveying scenario data of the belt conveyor, and obtain the initial conveying control scheme.

[0053] The simulated conveying module is used to execute the simulated conveying of the belt conveyor according to the initial conveying control scheme and obtain the simulated conveying dataset.

[0054] The impact risk detection module is used to perform start-stop impact risk detection on the belt conveyor based on the simulated conveying dataset, and obtain the first risk detection result of the conveying.

[0055] The response risk mining module is used to mine response anomaly risks of the belt conveyor based on the simulated transport dataset, and obtain the second risk detection result of the transport.

[0056] The fault risk detection module is used to perform fault risk detection on the belt conveyor based on the simulated conveying dataset and obtain the third risk detection result of the conveying.

[0057] The control coordination optimization module is used to coordinately optimize the initial transportation control scheme based on the first transportation risk detection result, the second transportation risk detection result, and the third transportation risk detection result to obtain an optimized transportation control scheme.

[0058] This invention discloses a conveying control method and system for a belt conveyor, comprising: generating an initial conveying control scheme by performing multi-parameter collaborative decision-making based on material conveying scenario data of the belt conveyor; performing simulated conveying of the belt conveyor based on the initial scheme to obtain a simulated conveying dataset; performing start-stop impact risk detection on the belt conveyor using the simulated conveying dataset to obtain a first risk detection result; performing response anomaly risk mining on the belt conveyor based on the simulated conveying dataset to obtain a second risk detection result; performing fault risk detection on the belt conveyor using the simulated conveying dataset to obtain a third risk detection result; and combining the first, second, and third risk detection results to collaboratively optimize the initial conveying control scheme to obtain an optimized conveying control scheme. The conveying control method and system for a belt conveyor disclosed in this invention solves the technical problems of single control strategy, lack of risk warning, and insufficient optimization depth, and achieves the technical effects of diversified control strategies, improved risk warning mechanism, and enhanced optimization depth. Attached Figure Description

[0059] Figure 1 This is a schematic flowchart of the conveying control method for the belt conveyor of the present invention.

[0060] Figure 2 This is a schematic diagram of the conveying control system of the belt conveyor of the present invention.

[0061] Figure labeling: Initial scheme acquisition module 11, simulated transportation module 12, impact risk detection module 13, response risk mining module 14, fault risk detection module 15, control collaborative optimization module 16. Detailed Implementation

[0062] The above technical solutions will now be described in detail with reference to the accompanying drawings and specific embodiments to provide a better understanding of them. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments used only to explain the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. Furthermore, it should be noted that, for ease of description, only the parts related to the present invention are shown in the drawings, not all of them.

[0063] Example 1, as Figure 1 This is a schematic flowchart of the conveying control method for the belt conveyor of the present invention, wherein the conveying control method for the belt conveyor includes:

[0064] S100: Based on the material conveying scenario data of the belt conveyor, perform multi-parameter conveying control collaborative decision-making to obtain the initial conveying control scheme.

[0065] Specifically, material conveying scenario data includes the physical characteristics of the material (such as particle size distribution, density, humidity, etc.), conveying task requirements (such as conveying volume, conveying distance, conveying time), conveying environmental conditions (such as temperature, humidity, dust concentration), and real-time status information of the conveying equipment (such as conveyor belt tension, motor load, idler speed, etc.).

[0066] Specifically, by comprehensively considering multiple key control parameters of belt conveyor operation (such as conveyor belt speed, tension, material supply, and guide device angle), a parameter correlation model and constraint relationship are established accordingly. Optimization algorithms (such as genetic algorithms and particle swarm optimization) are then used for coordinated adjustment to generate a conveying control strategy. The multi-parameter control model is used to analyze the interactions between different parameters. For example, studying the dynamic relationship between conveyor belt speed and material supply reveals a nonlinear correlation between the two under specific operating conditions.

[0067] Through the above process, comprehensive data collection and analysis of material conveying scenarios can be achieved, enabling refined modeling of the conveying process and improving the accuracy and adaptability of conveying control strategies. Simultaneously, the multi-parameter collaborative decision-making process can uncover potential correlations between various control parameters, reveal hidden optimization opportunities, and reduce energy consumption and equipment wear. The initial conveying control scheme generated based on collaborative decision-making provides a reliable benchmark for subsequent risk assessment and optimization adjustments, enhancing the stability and reliability of belt conveyor operation.

[0068] In some embodiments, multi-parameter conveying control collaborative decision-making is performed based on material conveying scenario data of the belt conveyor to obtain an initial conveying control scheme, including:

[0069] The conveying control constraint information of the belt conveyor is collected to obtain a multi-dimensional conveying control constraint domain; the multi-dimensional conveying control constraint domain is adjusted according to the material conveying scenario data to establish a multi-dimensional control constraint optimization domain; multi-parameter control collaborative decision-making is performed on the belt conveyor according to the multi-dimensional control constraint optimization domain to obtain a conveying control decision set; energy consumption minimization optimization is performed according to the conveying control decision set to generate the initial conveying control scheme.

[0070] Specifically, conveying control constraint information refers to the limitations imposed on control parameters in terms of equipment safety, energy consumption, efficiency, and material characteristics, such as maximum permissible belt speed, maximum load, and minimum tension. The multidimensional conveying control constraint domain is a multidimensional set of all feasible control parameters under various constraint conditions.

[0071] Specifically, scenario-related adjustment refers to the process of dynamically adjusting the parameter ranges of each constraint domain based on actual operating data to adapt to different transportation needs and environments. The multi-dimensional control constraint optimization domain obtained through scenario-related adjustment is a further optimized feasible domain of control parameters, which is more in line with the actual application of the target scenario.

[0072] Specifically, the transport control decision set refers to the set of all feasible combinations of control parameters after optimization and screening. The initial transport control scheme is the final selected control parameter setting scheme that satisfies the constraints and has the lowest energy consumption.

[0073] Specifically, firstly, basic control constraint information of the belt conveyor is collected, such as a belt speed limit of 3.5 m / s, a maximum load of 1000 t / h, a tension range of 10-30 kN, and a maximum motor output power of 200 kW, forming a multi-dimensional conveying control constraint domain. Next, based on real-time material conveying scenario data (e.g., the current conveyed material is wet coal, particle size 50 mm, humidity 15%, ambient temperature 5°C, conveying distance 1.2 km), the above multi-dimensional constraint domain is adjusted according to the scenario to establish a multi-dimensional control constraint optimization domain for the current scenario. For example, wet coal is prone to sticking to the belt, so the belt speed needs to be reduced to below 3.0 m / s, and the tension needs to be increased to 15-25 kN to prevent slippage.

[0074] Then, based on this optimization domain, a multi-parameter collaborative decision-making algorithm (such as genetic algorithm, multi-objective particle swarm optimization, etc.) is adopted to comprehensively consider parameters such as belt speed, tension force, and driving force to generate a series of feasible control parameter combinations (i.e., conveyor control decision set).

[0075] Furthermore, among all feasible decisions, the optimization is performed with the goal of minimizing energy consumption (e.g., belt speed of 2.7 m / s, tension of 18 kN, and driving force of 90 kW) to generate an initial conveying control scheme for the current scenario. For example, after optimization, it is recommended to operate at a belt speed of 2.7 m / s, a tension of 18 kN, and a driving force of 90 kW, with an estimated energy consumption of 0.09 kWh / t per unit of material.

[0076] Through the aforementioned multi-parameter conveying control collaborative decision-making process, adaptive and precise control of belt conveyors under different materials and operating conditions can be achieved. This not only ensures equipment operation safety and material conveying efficiency but also dynamically adapts to changes in material characteristics and the environment, significantly reducing energy consumption.

[0077] In some implementations, the multidimensional conveying control constraint domain is adjusted based on the material conveying scenario data to establish a multidimensional control constraint optimization domain, including:

[0078] Based on the material conveying scenario data, a conveying control scheme retrieval is performed to obtain a set of retrieved conveying control schemes; based on the retrieved conveying control scheme set, interval feature analysis is performed to obtain a multi-dimensional conveying control retrieval domain; based on the multi-dimensional conveying control retrieval domain, the multi-dimensional conveying control constraint domain is mapped and optimized to generate the multi-dimensional control constraint optimization domain.

[0079] Specifically, firstly, a database containing historical conveying control schemes is established, where each scheme records the material conveying scenario data and corresponding control parameter settings at that time. When conveying control is required, based on the current material conveying scenario data, a similarity algorithm (such as Euclidean distance, cosine similarity, etc.) can be used to retrieve historical conveying control schemes with high similarity from the database, forming a retrieved conveying control scheme set.

[0080] Next, the control parameters (such as conveyor belt speed, tension, and material supply) of each scheme in the retrieved conveyor control scheme set are analyzed to determine their range and variation characteristics. For example, analyzing the conveyor belt speed in multiple schemes reveals that it varies between 2.5 and 3.5 m / s, with most concentrated in the 3.0–3.2 m / s range; the tension ranges between 10 and 20 kN; and the material supply ranges between 500 and 800 t / h. This yields a multi-dimensional conveyor control retrieval domain, including the conveyor belt speed range [2.5, 3.5] m / s, the tension range [10, 20] kN, and the material supply range [500, 800] t / h.

[0081] Furthermore, the parameter ranges in the multidimensional conveying control retrieval domain are mapped and matched with the current material conveying scenario data. The parameter ranges are adjusted and optimized considering the specific requirements of the current scenario (such as higher conveying efficiency requirements, specific energy consumption limitations, etc.). Preferably, the intersection calculation of the multidimensional conveying control retrieval domain and the original multidimensional conveying control constraint domain of the equipment is performed to obtain a multidimensional control constraint optimization domain that simultaneously satisfies equipment safety and historical experience.

[0082] Multi-dimensional constraint domain optimization driven by scenario data can fully combine historical experience with the actual capabilities of equipment to achieve precise parameter selection for specific working conditions. This avoids the problem of excessively large parameter ranges and insufficient optimization caused by relying solely on theoretical constraints, and also avoids the safety risks arising from ignoring the actual capabilities of the equipment.

[0083] S200: Perform simulated conveying of the belt conveyor according to the initial conveying control scheme to obtain a simulated conveying dataset.

[0084] Specifically, the conveyor model refers to a digital 3D model of a belt conveyor, including the detailed structure and parameters of each component such as the conveyor belt, idlers, rollers, drive unit, and tensioning device, as well as their connection relationships and kinematic and dynamic characteristics. In a virtual environment, based on the initial conveying control scheme and material conveying scenario data, the conveyor model can be driven to perform operational simulations, simulating the actual conveying process such as material flow, accumulation, and sliding on the conveyor belt.

[0085] Specifically, in the simulated conveying process, real-time data on the operating status of each part of the conveyor model (such as conveyor belt speed, tension, motor speed, power, etc.), material distribution and motion data (such as material flow rate, stacking height, particle speed, etc.), and data on possible abnormal situations (such as slippage, deviation, etc.) are collected to form a simulated conveying dataset.

[0086] In some embodiments, a simulated conveying operation of the belt conveyor is performed according to the initial conveying control scheme to obtain a simulated conveying dataset, including:

[0087] A 3D reconstruction of the belt conveyor is performed to obtain a conveyor model; based on the material conveying scenario data, the initial conveying control scheme is simulated and executed according to the conveyor model to obtain the simulated conveying dataset.

[0088] Specifically, firstly, a laser scanner is used to perform a full-range scan of the belt conveyor to obtain its surface point cloud data. Then, 3D reconstruction software is used to process the point cloud data to generate a 3D model that realistically reflects the structure and motion characteristics of the conveyor. For example, a complete virtual conveyor is created, including components such as the drive unit, tensioning device, conveyor belt, and material inlets and outlets.

[0089] Optionally, the model can be simplified and optimized, retaining details of key components while ensuring that the model can run efficiently in simulation software.

[0090] Optionally, the processed model can be imported into a simulation platform to assign material properties and dynamic parameters to each component, such as the elastic modulus of the conveyor belt, the moment of inertia of the idler rollers, and the power of the motor, thereby establishing a complete conveyor model.

[0091] Specifically, actual material conveying scenario data (e.g., wet coal, 50mm particle size, 15% humidity, and 5℃ ambient temperature) is input into the conveyor model, and simulation parameters are set according to the previously optimized initial conveying control scheme (e.g., belt speed 2.7m / s, tension 18kN, driving force 90kW). Then, the simulation is run in the simulation software to dynamically calculate the motion, forces, and energy consumption of the material on the belt conveyor. For example, the simulation results can output data such as belt speed curves, tension changes, material distribution, cumulative energy consumption, and forces on key equipment components at different time periods. This dataset can provide quantitative basis for subsequent control scheme verification, parameter fine-tuning, and actual operational risk assessment.

[0092] Through 3D reconstruction and virtual simulation, the feasibility and advantages / disadvantages of the initial transport control scheme can be comprehensively evaluated and verified without affecting actual production.

[0093] S300: Based on the simulated conveying dataset, perform start-stop impact risk detection on the belt conveyor to obtain the first risk detection result for the conveying process.

[0094] In some embodiments, the belt conveyor is subjected to start-up and shutdown impact risk detection based on the simulated conveying dataset to obtain a first risk detection result for the conveying process, including:

[0095] Start-up and shutdown feature recognition is performed based on the simulated transport dataset to obtain start-up section feature data and shutdown section feature data; impact risk detection is performed based on the start-up section feature data to obtain start-up impact risk detection results; impact risk detection is performed based on the shutdown section feature data to obtain shutdown impact risk detection results; the start-up impact risk detection results and shutdown impact risk detection results are fused to generate the first transport risk detection result.

[0096] Specifically, the start-up section characteristic data and the shutdown section characteristic data refer to the various characteristic indicators reflecting the start-up and shutdown processes of the conveyor in the simulation dataset. By analyzing the above characteristic data, it is possible to determine whether there are abnormal impact phenomena and quantify the risk level (e.g., low, medium, high).

[0097] Specifically, by integrating the impact risk detection results from both the startup and shutdown phases, an overall startup and shutdown risk assessment conclusion can be obtained.

[0098] Specifically, firstly, time series analysis is performed on the simulated conveying dataset to pinpoint the time segments for conveyor startup and shutdown. For example, by using belt speed curves, driving force curves, and tension curves, data segments from 0 to the set belt speed (start) and from the set belt speed back to 0 (shutdown) are identified, and relevant features are extracted, such as peak acceleration, peak impact force, belt speed change rate, and tension change point, to form characteristic data for the startup and shutdown segments.

[0099] Specifically, the characteristic data of the starting section are then evaluated. For example, if the peak acceleration exceeds the equipment safety threshold (e.g., 0.5 m / s²), the following criteria are considered: 2 If the impact force suddenly exceeds the allowable range, a high risk is identified. Then, the peak impact force is compared with historical safety thresholds to determine the risk level, and the output is the result of initiating impact risk detection.

[0100] Specifically, the same methodology is used to analyze the characteristic data of the stopping section to identify risks such as sudden deceleration and inertial impact during the stopping process. The stopping impact risk detection results are then output. It should be understood that, for the sake of brevity, further elaboration is not provided here.

[0101] Preferably, the impact risk detection includes the detection of electrical impact risk, mechanical impact risk, and material impact risk. Among them, electrical impact risk refers to the risk that during the start-up of the conveyor, due to the sudden increase in motor starting current, power surge, etc., the electrical system components (such as motors, frequency converters, contactors, etc.) will be subjected to abnormal electrical loads, which may lead to overload, overheating, insulation breakdown, or tripping.

[0102] Mechanical shock risk refers to the sudden stress or impact load that occurs on key mechanical components of the conveyor (such as reducers, couplings, idlers, frames, etc.) during startup due to inertial forces, excessive acceleration of the drive system, etc., which may cause fatigue damage, loosening, breakage and other mechanical failures.

[0103] Material impact risk refers to the risk that during the start-up process, the material on the conveyor belt may slip, accumulate, spill, or block due to factors such as inertia and sudden changes in belt speed, which may lead to material loss, environmental pollution, or downstream equipment failure.

[0104] Furthermore, the results of the startup and shutdown impact risk detection will be integrated to form the first risk detection result for delivery. For example, if any stage is high risk, the overall assessment will be high risk; if all stages are low risk, the assessment will be safe.

[0105] Through the above process, potential impact risks during start-up and shutdown can be identified in advance through virtual simulation and data analysis before the actual equipment is put into operation, thus avoiding equipment damage or safety accidents during actual operation.

[0106] In some implementations, impact risk detection is performed based on the characteristic data of the starting section to obtain the starting impact risk detection result, including:

[0107] The starting impact risk assessment records of the belt conveyor are retrieved to obtain a starting section sample set, as well as corresponding electrical impact risk sample sets, mechanical impact risk sample sets, and material impact risk sample sets. A starting electrical impact risk mapping space is constructed based on the starting section sample set and the electrical impact risk sample set. A starting mechanical impact risk mapping space is constructed based on the starting section sample set and the mechanical impact risk sample set. A starting material impact risk mapping space is constructed based on the starting section sample set and the material impact risk sample set. These three mapping spaces are then integrated to generate a starting impact risk detection space. The starting section feature data is input into the starting impact risk detection space to generate the starting impact risk detection result.

[0108] Specifically, the start-up section characteristic data refers to the key operational characteristic parameters collected during the start-up phase of the belt conveyor (such as the process from initial power-on to acceleration and stable operation), including peak start-up current, acceleration, rate of change of tension force, and initial material distribution. The start-up section sample set is a collection of typical data from historical or simulation data that are similar to the current start-up section characteristics, containing multi-dimensional features under multiple start-up conditions.

[0109] Specifically, the electrical, mechanical, and material impact risk sample sets refer to the historical or simulated risk data sets related to electrical systems (such as current surges and overloads), mechanical systems (such as torque peaks and tension surges), and material systems (such as stacking impacts and spillage risks), respectively.

[0110] Specifically, through data modeling (such as multiple regression, machine learning, neural networks, etc.), the relationship between the characteristics of the starting segment and various shock risk indicators can be mapped to a high-dimensional space, forming a shock risk detection space, thereby realizing quantitative prediction from characteristics to risks.

[0111] Specifically, firstly, a predetermined number of typical samples of the belt conveyor startup process are selected from historical databases or simulation data. For each group of startup samples, the corresponding risks are labeled or calculated: electrical impact risk (such as the probability that the startup current exceeds the rated value), mechanical impact risk (such as sudden changes in the force on the idler rollers and reducers), and material impact risk (such as material slippage and spillage at the moment of startup).

[0112] Specifically, supervised learning methods such as multiple regression, decision trees, or neural networks are employed to establish a mapping relationship between start-up characteristics (e.g., starting current, acceleration) and electrical shock risks, thereby obtaining a predictive model. Similarly, a mapping is established between start-up characteristics and mechanical shock risks (e.g., maximum torque, peak tension). Furthermore, a mapping is established between material states (e.g., initial accumulation, flow rate changes) and material-related risks (e.g., spillage, blockage).

[0113] Furthermore, weighted fusion, risk priority ranking, or multi-objective decision-making methods (such as TOPSIS, AHP, etc.) are employed to integrate the three types of risk mapping spaces into a unified startup impact risk detection space. This is achieved by using characteristic data from the current startup (such as a peak startup current of 120A and an acceleration of 0.5m / s²). 2 Inputting data such as initial material accumulation rate of 80% into the integrated detection space can automatically obtain the comprehensive impact risk detection results for this launch, and can further subdivide the risk levels in three aspects: electrical, mechanical, and material.

[0114] By employing the above steps to perform layered modeling and integration of risks related to electrical, mechanical, and material aspects, potential impact hazards during the start-up phase of a belt conveyor can be identified more comprehensively and meticulously. This provides a scientific basis for subsequent control parameter optimization and start-up strategy adjustment, avoiding misjudgments or omissions caused by relying on a single indicator.

[0115] S400: Based on the simulated transport dataset, perform response anomaly risk mining on the belt conveyor to obtain the second transport risk detection result.

[0116] In some embodiments, the belt conveyor is subjected to response anomaly risk mining based on the simulated conveying dataset to obtain a second conveying risk detection result, including:

[0117] Tension risk analysis is performed based on the simulated conveying dataset to obtain a tension risk coefficient; deviation and slippage risk analysis is performed based on the simulated conveying dataset to obtain a deviation and slippage risk coefficient; idler jamming risk analysis is performed based on the simulated conveying dataset to obtain an idler jamming risk coefficient; the tension risk coefficient, the deviation and slippage risk coefficient, and the idler jamming risk coefficient are added to the second risk detection result of the conveying process.

[0118] Specifically, the tension risk coefficient quantifies the degree of risk associated with abnormalities in the conveyor belt tensioning system (such as tensioning devices and tension sensors); both excessive looseness and excessive tightness can lead to malfunctions. The belt misalignment and slippage risk coefficient reflects the risk of the conveyor belt running off-center (the belt deviates from the centerline) or slipping (the belt speed does not match the linear speed of the drive roller). The idler jamming risk coefficient reflects the risk of the idler being obstructed or jammed due to bearing damage, foreign objects, or other reasons.

[0119] Specifically, the analysis of the tension force variation curve, tensioning device operation frequency, extreme values, etc. in the simulation data determines whether there are phenomena such as excessive / insufficient tension force or abnormal fluctuations, and conducts tension risk analysis. The risk coefficient can be calculated using methods such as normalization, fuzzy logic, and threshold determination (e.g., 0~1, the larger the value, the higher the risk).

[0120] Specifically, by using preset rules or mathematical models, information such as belt offset, belt deviation sensor alarm, and the difference between belt speed and roller speed is analyzed to identify belt deviation or slippage trends and calculate the belt deviation and slippage risk coefficient.

[0121] Specifically, by analyzing data such as idler speed, vibration, temperature rise, and energy consumption, it is possible to determine whether there are abnormalities such as idler jamming or poor rotation, and to calculate and output the corresponding idler jamming risk coefficient. It should be understood that the aforementioned idler jamming risk coefficient can also be calculated based on preset rules or mathematical models.

[0122] Furthermore, by summarizing the above three types of risk coefficients to form a second risk detection result, multiple typical operational anomaly risks can be monitored simultaneously, improving safety and serving as a basis for subsequent alarms, maintenance decisions, or risk warnings.

[0123] In some implementations, tension risk analysis is performed based on the simulated transport dataset to obtain the tension risk coefficient, including:

[0124] Based on the simulated transport dataset, tension simulation data is extracted; based on the tension sample dataset and the tension risk sample set, a first tension risk detection model, a second tension risk detection model, and a third tension risk detection model are trained; the first, second, and third tension risk detection models are sorted in descending order of detection accuracy to obtain a risk detection model distribution; based on the risk detection model distribution, a tension risk detection base learner and a tension risk detection meta learner are determined; the tension risk detection meta learner is subjected to output reinforcement learning based on the tension risk detection base learner to obtain a tension risk parser; the tension simulation data is input into the tension risk parser to obtain the tension risk coefficient.

[0125] Specifically, firstly, tension-related feature data are extracted from the simulated conveying dataset to form a tension simulation dataset, which includes: tension force, tensioning device operation, belt speed change, tension sensor output, etc.

[0126] Next, prepare a tension sample dataset, including tension simulation data under normal operation and when tension risks occur (such as slippage due to insufficient tension, or stretching of the conveyor belt due to excessive tension). Prepare a corresponding tension risk sample set, and label each sample with whether a tension risk has occurred and the degree of risk (such as low risk, medium risk, high risk).

[0127] Furthermore, using the acquired sample dataset, three tension risk detection models were trained respectively. These models can be constructed using different machine learning algorithms (such as logistic regression, decision trees, neural networks, etc.). For example, the first model uses a logistic regression algorithm to predict tension risk based on tension and pressure values; the second model uses a decision tree algorithm to assess risk based on the displacement of the tensioning device and the rate of change of tension; and the third model uses a neural network algorithm to perform deep learning by comprehensively considering multiple tension-related parameters.

[0128] Furthermore, the three trained models are tested and evaluated, and the detection accuracy of each model on the test set is calculated, such as precision, recall, and F1 score. The three models are then sorted from highest to lowest detection accuracy to obtain the risk detection model distribution.

[0129] Next, base learners and meta-learners are selected from the risk detection model distribution. Specifically, the tensioned risk detection base learners correspond to the top two risk detection models in terms of detection accuracy, serving as the "base learners" in the ensemble learning. The tensioned risk detection meta-learners are the remaining risk detection models, serving as the "meta-learners" in the ensemble learning to further improve the performance of the ensemble model.

[0130] Furthermore, the output of the base learner is used as the input of the meta-learner for output reinforcement learning. The meta-learner adjusts its parameters based on the prediction results of the base learner and the actual risk labels to optimize the output, thereby improving the accuracy and reliability of the prediction. For example, the meta-learner can use gradient descent to adjust its weights based on the prediction error of the base learner, resulting in a more accurate tension risk coefficient in the final output.

[0131] Finally, the tension simulation data is input into a reinforcement learning-based tension risk parser (including a base learner and a meta-learner). The parser combines the outputs of the base learner and the meta-learner to generate a tension risk coefficient. For example, the tension risk coefficient can be a value between 0 and 1, representing the probability of a risk occurring in the tension system; for example, 0.8 indicates a higher risk, and 0.3 indicates a lower risk.

[0132] The tension risk coefficient obtained through the above process provides a quantitative basis for subsequent risk response and control scheme optimization, which helps to prevent tension failures in advance and ensure the stable operation of the belt conveyor.

[0133] S500: Based on the simulated conveying dataset, perform fault risk detection on the belt conveyor to obtain the third risk detection result of the conveying.

[0134] In some embodiments, fault risk detection is performed on the belt conveyor based on the simulated conveying dataset to obtain a third risk detection result for the conveyor, including:

[0135] Obtain the conveying monitoring record set and the fault risk record set of the belt conveyor; use the conveying monitoring record set as input information and the fault risk record set as output information to train K conveyor fault risk detection models, where K is a positive integer greater than 1; input the simulated conveying dataset into the K conveyor fault risk detection models to obtain K fault risk coefficients, and calculate the mean of the K fault risk coefficients to generate the third risk detection result of the conveying.

[0136] Specifically, the conveyor monitoring record set refers to the collection of various sensor data and monitoring parameters collected during the operation of the belt conveyor, such as motor current, speed, temperature, vibration, and belt misalignment. The fault risk record set is a collection of historically known conveyor fault events and their corresponding operating parameters and risk level labels.

[0137] Specifically, the K conveyor fault risk detection models refer to the K models used for fault risk identification that are trained using different machine learning algorithms (such as decision trees, neural networks, support vector machines, etc.) or the same algorithm with different parameter configurations, where K is a positive integer greater than 1.

[0138] Specifically, the failure risk coefficient is a numerical value (generally a probability value between 0 and 1) output by each risk detection model, reflecting the degree of failure risk under the current operating conditions. The third risk detection result refers to the comprehensive risk assessment result obtained by averaging the K failure risk coefficients, which is used to represent the overall failure risk level of the belt conveyor at present.

[0139] Specifically, firstly, multi-dimensional monitoring data is collected from the actual operation of the belt conveyor to form a conveyor monitoring record set. Past failure events are then organized and labeled to form a failure risk record set. Next, using the conveyor monitoring record set as input and the failure risk record set as output, K failure risk detection models are trained using different algorithms or parameters. Then, a simulated conveyor dataset is used as test input, fed into each of the K trained models to obtain K failure risk coefficients. The average of these K risk coefficients is calculated to obtain the final third-level risk detection result for the conveyor. For example, if K=3, and the outputs are 0.15, 0.22, and 0.19, the final result is (0.15+0.22+0.19) / 3=0.1867.

[0140] By integrating multiple models through the above process, the probability of misjudgment by a single model can be effectively reduced, improving the overall accuracy and robustness of detection. Simultaneously, by using risk coefficients, complex fault risks can be quantified and output, facilitating intuitive understanding and decision-making by operations and maintenance personnel.

[0141] By using fault risk detection and risk assessment with simulated delivery datasets, potential faults can be quickly identified and resolved without actually running the equipment, reducing the risk of equipment damage and maintenance costs, and improving equipment reliability and operating efficiency.

[0142] S600: Based on the first risk detection result, the second risk detection result, and the third risk detection result, the initial transportation control scheme is collaboratively optimized to obtain an optimized transportation control scheme.

[0143] Furthermore, it is determined whether the first risk detection result, the second risk detection result, and the third risk detection result meet the preset control threshold. If any one of them fails to meet the threshold, it can be considered that the initial transmission control scheme cannot complete the expected task and needs to be adjusted and optimized.

[0144] Optionally, the initial conveying control scheme can be randomly fluctuated using optimization algorithms, such as adjusting the conveyor belt speed, tension, and material supply. The optimized conveying control scheme can then be verified in a simulation environment. The aforementioned multiple risk detection processes can be repeated to obtain the first, second, and third risk detection results of the fluctuated scheme. The system can then iteratively determine whether the preset control threshold is met.

[0145] The collaborative optimization process fully considers multiple risk detection results, enabling comprehensive and systematic adjustment and optimization of the initial conveyor control scheme. The optimized conveyor control scheme performs excellently in reducing start-up and shutdown impact risks, minimizing response anomaly risks, and preventing failure risks, effectively improving the safety and stability of belt conveyor operation.

[0146] In summary, the conveying control method for belt conveyors provided by this invention has the following technical effects:

[0147] By performing multi-parameter collaborative decision-making based on material conveying scenario data of the belt conveyor, an initial conveying control scheme is generated. Based on the initial scheme, simulated conveying of the belt conveyor is executed to obtain a simulated conveying dataset. Using the simulated conveying dataset, start-stop impact risk detection is performed on the belt conveyor to obtain a first risk detection result. Based on the simulated conveying dataset, response anomaly risk mining is performed on the belt conveyor to obtain a second risk detection result. Using the simulated conveying dataset, fault risk detection is performed on the belt conveyor to obtain a third risk detection result. Combining the first, second, and third risk detection results, the initial conveying control scheme is collaboratively optimized to obtain an optimized conveying control scheme, thereby achieving the technical effects of diversified control strategies, improved risk warning mechanisms, and enhanced optimization depth.

[0148] Example 2, as Figure 2 This is a schematic diagram of the conveying control system of the belt conveyor of the present invention. For example, Figure 1 The flowchart of the conveying control method for the belt conveyor of the present invention can be illustrated as follows: Figure 2 The structure shown is implemented.

[0149] Based on the same concept as the conveying control method of the belt conveyor in the above embodiments, the present invention also provides a conveying control system for a belt conveyor, comprising:

[0150] The initial scheme acquisition module 11 is used to make multi-parameter conveying control collaborative decisions based on the material conveying scenario data of the belt conveyor, and obtain the initial conveying control scheme.

[0151] The simulated conveying module 12 is used to perform simulated conveying of the belt conveyor according to the initial conveying control scheme and obtain a simulated conveying dataset.

[0152] The impact risk detection module 13 is used to perform start-stop impact risk detection on the belt conveyor based on the simulated conveying dataset, and obtain the first risk detection result of the conveying.

[0153] The response risk mining module 14 is used to perform response anomaly risk mining on the belt conveyor based on the simulated conveying dataset, and obtain the second risk detection result of the conveying.

[0154] The fault risk detection module 15 is used to perform fault risk detection on the belt conveyor based on the simulated conveying dataset and obtain the third risk detection result of the conveyor.

[0155] The control coordination optimization module 16 is used to coordinately optimize the initial transportation control scheme based on the first transportation risk detection result, the second transportation risk detection result, and the third transportation risk detection result to obtain an optimized transportation control scheme.

[0156] In some embodiments, the impact risk detection module 13 includes:

[0157] The start-stop feature recognition unit is used to perform start-stop feature recognition based on the simulated transport dataset to obtain start-up section feature data and stop-down section feature data.

[0158] The impact risk detection unit is activated to perform impact risk detection based on the characteristic data of the activation section and obtain the activation impact risk detection result.

[0159] The shutdown impact risk detection unit is used to perform impact risk detection based on the characteristic data of the shutdown section and obtain the shutdown impact risk detection result.

[0160] A first risk detection result generation unit is used to merge the start-up impact risk detection result and the shutdown impact risk detection result to generate the first risk detection result.

[0161] In some implementations, the shock risk detection module 13 includes the following steps to activate the shock risk detection unit:

[0162] The shock risk assessment record retrieval subunit is used to retrieve the shock risk assessment records of the belt conveyor to obtain the starting section sample set, as well as the electrical shock risk sample set, mechanical shock risk sample set and material shock risk sample set corresponding to the starting section sample set.

[0163] A sub-unit for constructing an electrical shock risk mapping space is initiated, which is used to construct an electrical shock risk mapping space based on the initiation section sample set and the electrical shock risk sample set.

[0164] A sub-unit for constructing the mechanical impact risk mapping space is initiated, which is used to construct the mechanical impact risk mapping space based on the initiation segment sample set and the mechanical impact risk sample set.

[0165] A sub-unit for constructing the material shock risk mapping space is initiated, which is used to construct the material shock risk mapping space based on the sample set of the initiation section and the sample set of material shock risks.

[0166] A startup impact risk detection space generation subunit is used to integrate the startup electrical impact risk mapping space, the startup mechanical impact risk mapping space, and the startup material impact risk mapping space to generate a startup impact risk detection space.

[0167] The startup impact risk detection result generation subunit is used to input the startup segment feature data into the startup impact risk detection space and generate the startup impact risk detection result.

[0168] In some embodiments, the response risk mining module 14 includes:

[0169] The tension risk analysis unit is used to perform tension risk analysis based on the simulated transport dataset to obtain the tension risk coefficient.

[0170] The deviation and slippage risk analysis unit is used to analyze the deviation and slippage risk based on the simulated transport dataset and obtain the deviation and slippage risk coefficient.

[0171] The idler jamming risk analysis unit is used to analyze the idler jamming risk based on the simulated conveying dataset and obtain the idler jamming risk coefficient.

[0172] The second risk detection result generation unit is used to add the tension risk coefficient, the deviation and slippage risk coefficient, and the idler jamming risk coefficient to the second risk detection result.

[0173] In some implementations, the tension risk analysis unit in the response risk mining module 14 includes:

[0174] The tension simulation data extraction subunit is used to extract tension simulation data based on the simulated transport dataset.

[0175] The tension risk detection model training subunit is used to train the first tension risk detection model, the second tension risk detection model, and the third tension risk detection model based on the tension sample dataset and the tension risk sample set.

[0176] The risk detection model arrangement and distribution acquisition subunit is used to sort the first tension risk detection model, the second tension risk detection model, and the third tension risk detection model in descending order of detection accuracy to obtain the risk detection model arrangement and distribution.

[0177] The tension risk detection base learner and meta learner determination subunit is used to determine the tension risk detection base learner and tension risk detection meta learner according to the arrangement distribution of the risk detection model.

[0178] The tension risk parser generation subunit is used to perform output reinforcement learning on the tension risk detection meta-learner based on the tension risk detection base learner to obtain the tension risk parser.

[0179] The tension risk coefficient calculation subunit is used to input the tension simulation data into the tension risk analyzer to obtain the tension risk coefficient.

[0180] In some embodiments, the fault risk detection module 15 includes:

[0181] The conveying monitoring and fault risk record set acquisition unit is used to obtain the conveying monitoring record set and fault risk record set of the belt conveyor.

[0182] The conveyor fault risk detection model training unit is used to train K conveyor fault risk detection models with the conveyor monitoring record set as input information and the fault risk record set as output information, where K is a positive integer greater than 1.

[0183] The fault risk coefficient calculation and third risk detection result generation unit is used to input the simulated transport dataset into the K conveyor fault risk detection models, obtain K fault risk coefficients, calculate the mean of the K fault risk coefficients, and generate the third risk detection result of the transport.

[0184] In some embodiments, the initial scheme acquisition module 11 includes:

[0185] The conveying control constraint information acquisition unit is used to acquire the conveying control constraint information of the belt conveyor and obtain a multi-dimensional conveying control constraint domain.

[0186] The multidimensional control constraint optimization domain establishment unit is used to perform scenario correlation adjustment on the multidimensional conveying control constraint domain based on the material conveying scenario data, and establish the multidimensional control constraint optimization domain.

[0187] The conveying control decision set acquisition unit is used to perform multi-parameter control collaborative decision-making on the belt conveyor based on the multi-dimensional control constraint optimization domain to obtain the conveying control decision set.

[0188] The initial scheme generation unit for transport control is used to perform energy consumption minimization optimization based on the transport control decision set and generate the initial scheme for transport control.

[0189] In some implementations, the multidimensional control constraint optimization domain establishment unit in the initial scheme acquisition module 11 includes:

[0190] The conveying control scheme retrieval subunit is used to retrieve conveying control schemes based on the material conveying scenario data to obtain a set of retrieved conveying control schemes.

[0191] The interval feature parsing and retrieval domain acquisition subunit is used to perform interval feature parsing based on the retrieval transport control scheme set to obtain a multi-dimensional transport control retrieval domain.

[0192] A multidimensional control constraint optimization domain generation subunit is used to map and optimize the multidimensional transport control constraint domain based on the multidimensional transport control retrieval domain, and generate the multidimensional control constraint optimization domain.

[0193] In some embodiments, the execution steps of the simulated conveying module 12 further include: performing three-dimensional reconstruction based on the belt conveyor to obtain a conveyor model; and simulating the execution of the initial conveying control scheme based on the material conveying scenario data and the conveyor model to obtain the simulated conveying dataset.

[0194] It should be understood that the focus of the embodiments mentioned in this specification is their difference from other embodiments. The specific embodiments in the aforementioned Embodiment 1 are also applicable to the conveying control system of the belt conveyor described in Embodiment 2. For the sake of brevity, they will not be elaborated further here.

[0195] It should be understood that the embodiments disclosed in this invention and the above description enable those skilled in the art to implement this invention. However, this invention is not limited to the embodiments mentioned above. It should be understood that those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this invention, and should all be included within the protection scope of this invention.

Claims

1. A conveying control method for a belt conveyor, characterized in that, include: Based on the material conveying scenario data of the belt conveyor, multi-parameter conveying control collaborative decision-making is performed to obtain an initial conveying control scheme, including: Collect the conveying control constraint information of the belt conveyor to obtain a multi-dimensional conveying control constraint domain; Based on the material conveying scenario data, the multidimensional conveying control constraint domain is adjusted according to scenario correlation to establish a multidimensional control constraint optimization domain, including: Based on the material conveying scenario data, a conveying control scheme retrieval is performed to obtain a set of retrieved conveying control schemes; Based on the retrieved transport control scheme set, interval feature analysis is performed to obtain a multi-dimensional transport control retrieval domain; Based on the multidimensional transport control retrieval domain, the multidimensional transport control constraint domain is mapped and optimized to generate the multidimensional control constraint optimization domain; Based on the multidimensional control constraint optimization domain, multi-parameter control collaborative decision-making is performed on the belt conveyor to obtain a conveyor control decision set; Based on the transport control decision set, energy consumption minimization optimization is performed to generate the initial transport control scheme; The simulated conveying of the belt conveyor is executed according to the initial conveying control scheme to obtain a simulated conveying dataset; Based on the simulated conveying dataset, the belt conveyor is subjected to start-up and shutdown impact risk detection to obtain the first risk detection result for the conveying process, including: Based on the simulated transport dataset, start-up and stop features are identified to obtain start-up section feature data and stop section feature data. Based on the characteristic data of the starting section, impact risk detection is performed to obtain the starting impact risk detection results, including: Based on the start-up impact risk assessment record retrieval of the belt conveyor, a start-up section sample set is obtained, as well as the electrical impact risk sample set, mechanical impact risk sample set, and material impact risk sample set corresponding to the start-up section sample set; Based on the starting section sample set and the electrical shock risk sample set, a starting electrical shock risk mapping space is constructed; Based on the starting section sample set and the mechanical impact risk sample set, a starting mechanical impact risk mapping space is constructed; Based on the starting section sample set and the material shock risk sample set, a starting material shock risk mapping space is constructed; The starting electrical shock risk mapping space, the starting mechanical shock risk mapping space, and the starting material shock risk mapping space are integrated to generate a starting shock risk detection space; The startup segment feature data is input into the startup impact risk detection space to generate the startup impact risk detection result; Impact risk detection is performed based on the characteristic data of the shutdown section to obtain the shutdown impact risk detection result; The first risk detection result for the conveying process is generated by combining the startup impact risk detection result and the shutdown impact risk detection result; Based on the simulated transport dataset, the belt conveyor is subjected to response anomaly risk mining to obtain the second transport risk detection result; Based on the simulated conveying dataset, the belt conveyor is subjected to fault risk detection to obtain the third risk detection result of the conveying process; Based on the first risk detection result, the second risk detection result, and the third risk detection result, the initial transportation control scheme is collaboratively optimized to obtain an optimized transportation control scheme.

2. The conveying control method for a belt conveyor as described in claim 1, characterized in that, Based on the simulated transport dataset, the belt conveyor is subjected to response anomaly risk mining to obtain the second risk detection result for the transport, including: Tension risk analysis is performed based on the simulated transport dataset to obtain the tension risk coefficient; Based on the simulated transport dataset, the risk of deviation and slippage is analyzed to obtain the risk coefficient of deviation and slippage. Based on the simulated conveying dataset, the risk of idler jamming is analyzed to obtain the idler jamming risk coefficient. The tension risk coefficient, the deviation and slippage risk coefficient, and the idler roller jamming risk coefficient are added to the second risk detection result of the conveyor.

3. The conveying control method for a belt conveyor as described in claim 2, characterized in that, Tension risk analysis is performed based on the simulated transport dataset to obtain tension risk coefficients, including: Based on the simulated transport dataset, extract the tension simulation data; Based on the tension sample dataset and the tension risk sample dataset, train the first tension risk detection model, the second tension risk detection model, and the third tension risk detection model; The tension risk detection first model, the tension risk detection second model, and the tension risk detection third model are sorted in descending order of detection accuracy to obtain the risk detection model arrangement distribution; Based on the arrangement and distribution of the risk detection model, determine the tension risk detection base learner and the tension risk detection meta learner; Based on the tension risk detection base learner, the output reinforcement learning of the tension risk detection meta learner is performed to obtain the tension risk parser; The tension simulation data is input into the tension risk analyzer to obtain the tension risk coefficient.

4. The conveying control method for a belt conveyor as described in claim 1, characterized in that, Based on the simulated conveying dataset, fault risk detection is performed on the belt conveyor to obtain the third risk detection result for the conveyor, including: Obtain the conveying monitoring record set and fault risk record set of the belt conveyor; Using the conveyor monitoring record set as input information and the fault risk record set as output information, train K conveyor fault risk detection models, where K is a positive integer greater than 1; The simulated transport dataset is input into the K conveyor fault risk detection models to obtain K fault risk coefficients. The mean of the K fault risk coefficients is calculated to generate the third risk detection result of the transport.

5. The conveying control method for a belt conveyor as described in claim 1, characterized in that, The simulated conveying of the belt conveyor is executed according to the initial conveying control scheme to obtain a simulated conveying dataset, including: A three-dimensional reconstruction of the belt conveyor is performed to obtain a conveyor model; Based on the material conveying scenario data, the initial conveying control scheme is simulated and executed according to the conveyor model to obtain the simulated conveying dataset.

6. A conveying control system for a belt conveyor, characterized in that, A conveying control method for implementing the belt conveyor according to any one of claims 1 to 5, comprising: The initial scheme acquisition module is used to make multi-parameter conveying control collaborative decisions based on the material conveying scenario data of the belt conveyor, and obtain the initial conveying control scheme; The simulated conveying module is used to execute the simulated conveying of the belt conveyor according to the initial conveying control scheme and obtain the simulated conveying dataset; The impact risk detection module is used to perform start-stop impact risk detection on the belt conveyor based on the simulated conveying dataset, and obtain the first risk detection result of the conveying process. The response risk mining module is used to mine response anomaly risks of the belt conveyor based on the simulated transport dataset, and obtain the second risk detection result of the transport; The fault risk detection module is used to perform fault risk detection on the belt conveyor based on the simulated conveying dataset and obtain the third risk detection result of the conveying. The control coordination optimization module is used to coordinately optimize the initial transportation control scheme based on the first transportation risk detection result, the second transportation risk detection result, and the third transportation risk detection result to obtain an optimized transportation control scheme.

Images