A silo coal blending closed-loop dynamic scheduling method and system based on factual coal quality data

Abstract

The present application relates to a kind of silo coal blending closed loop dynamic scheduling method and system based on fact coal quality data, the method includes: (1) with the current discharge coal type and real-time physical property index of each silo of silo unloading coal species real-time identification system output as fact coal quality data input dynamic scheduling system;(2) when dynamic scheduling system monitors that any silo coal species switching or into mix material transition section will lead to the quality of mix coal deviate target, the fact coal quality data and current feeding rate are transmitted to intelligent coal blending system;(3) intelligent coal blending system calls built-in coke quality prediction model and intelligent optimization algorithm, and under the constraint of coke quality qualified, with the lowest cost of raw material as target optimization solution, output the optimal feeding rate of each silo to dynamic scheduling system;(4) dynamic scheduling system is issued to the optimal feeding rate to each silo discharge mechanism execution, and continuously monitors feedback, forms closed loop dynamic scheduling.

Description

Technical Field

[0001] This invention relates to the field of coking coal blending technology, and in particular to a closed-loop dynamic scheduling method and system for silo coal blending based on actual coal quality data. Background Technology

[0002] In coking production, the quality of the blended coal directly determines the quality of the coke. Typically, different coal types are stored separately in multiple silos and mixed according to a set ratio using a disc feeder. To save land, more and more coking plants are adopting large silos with multiple coal types stored in a single silo, meaning different coal types are layered from top to bottom within the same silo. The core problem with this storage method is that the coal type and quality at the silo outlet dynamically change over time during the discharge process. When the discharged coal switches from one type to another, it undergoes a gradual mixing transition. During this period, if the silos storing other materials continue to operate according to the original blending ratio, the overall indicators of the blended coal will inevitably deviate from the target, leading to fluctuations in coke quality.

[0003] In addition, existing coal blending control technology has the following two defects: First, the input data is neither real-time nor accurate. Traditional coal blending systems rely on offline laboratory test data or manually set values ​​as the basis for the coal quality of each bin, assuming that the coal quality remains constant during the blending process. However, in the case of multi-coal mixed silos, this assumption is completely invalid, because the coal quality changes in real time with the discharge process, and the system input will inevitably be inconsistent with the actual situation.

[0004] Secondly, the adjustment methods are simple and lack professionalism. When deviations in coke quality are found, simple adjustment methods such as manual experience or proportional increases or decreases are usually used. These methods cannot accurately predict the quality of the coke after adjustment, nor can they optimize raw material costs.

[0005] The invention utilizes "A method and system for real-time identification and quality prediction of coal types in silo discharge based on multi-dimensional information fusion" (patent filed on the same day as this invention), which can output the type and physical properties of the coal currently discharged at each silo nozzle in real time, providing a foundation for solving the aforementioned problems. Currently, some coking enterprises have also established "intelligent coal blending systems," which have built-in coke quality prediction models based on big data and artificial intelligence. These systems can accurately assess the impact of coal blending schemes on core indicators such as coke thermal strength (CSR) and cold strength (M40). However, these intelligent coal blending systems are mostly used for offline coal blending scheme formulation and have not yet been linked with the real-time operating conditions of silo discharge.

[0006] Therefore, there is an urgent need for a method that can deeply integrate the "real coal quality data" detected in real time in silos with a professional "intelligent coal blending system" and form a closed-loop dynamic scheduling. Summary of the Invention

[0007] This invention provides a closed-loop dynamic scheduling method and system for silo coal blending based on actual coal quality data. Driven by actual coal quality data from the silo discharge port, the system is connected to a smart coal blending system for professional optimization calculations and direct closed-loop execution, thereby realizing dynamic scheduling of coal blending.

[0008] To achieve the above objectives, the present invention employs the following technical solution: A closed-loop dynamic scheduling method for silo coal blending based on actual coal quality data includes the following steps: (1) The current coal type and real-time physical property index of each silo discharged from the silo discharge coal type real-time identification system are used as the actual coal quality data input into the dynamic scheduling system; (2) When the dynamic scheduling system detects that switching coal types in any silo or entering the mixing transition section will cause the quality of the blended coal to deviate from the target, the actual coal quality data and the current feeding rate are transmitted to the intelligent coal blending system. (3) The intelligent coal blending system calls the built-in coke quality prediction model and intelligent optimization algorithm, and optimizes the solution with the goal of minimizing raw material cost under the constraint of qualified coke quality, and outputs the optimal feeding rate of each silo to the dynamic scheduling system. (4) The dynamic scheduling system sends the optimal feeding rate to each silo discharge mechanism for execution and continuously monitors and provides feedback to form a closed-loop dynamic scheduling.

[0009] In step (1), the actual coal quality data is updated every second, including the coal type identification, ash content, volatile matter, sulfur content, caking index and other real-time physical property indicators of the actual material discharged at each silo nozzle.

[0010] In step (3), the coke quality prediction model is a nonlinear data-driven model established based on historical production data through machine learning or neural networks, or a mechanism-data fusion model established based on coking mechanism, used to accurately predict the hot strength, cold strength and reactivity index of coke corresponding to blended coal.

[0011] In step (3), the intelligent optimization algorithm is a genetic algorithm, differential evolution algorithm or particle swarm optimization algorithm, which can find the optimal solution in seconds under complex constraints containing nonlinear prediction models.

[0012] In step (4), when issuing the optimal feeding rate, a rate ramp smooth transition method is adopted, and the current rate is gradually adjusted to the target rate within a preset time to reduce the impact on production.

[0013] A closed-loop dynamic scheduling system for silo coal blending based on actual coal quality data is proposed. The closed-loop dynamic scheduling system for silo coal blending consists of a dynamic scheduling system and a smart coal blending system dynamic scheduling system. The dynamic scheduling system includes the following components: The data acquisition module is used to obtain actual coal quality data from the silo coal type identification system. The trigger judgment module is used to determine whether dynamic scheduling should be triggered. The communication interface module is used for data interaction with the intelligent coal blending system; The instruction issuing module is used to issue the optimal feeding rate to the PLC / DCS for execution; The intelligent coal blending system incorporates a coke quality prediction model, an intelligent optimization algorithm library, and a coal type cost database.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1) For the first time, the "real coal quality data" detected in real time at the silo discharge port is used as the input basis for coal blending scheduling, which changes the lag and inaccuracy of traditional coal blending that relies on offline testing, and enables the coal blending system to have the ability to perceive coal quality changes in seconds.

[0015] 2) By connecting to a professional intelligent coal blending system and utilizing existing coke quality prediction models and intelligent optimization algorithms for multi-objective optimization, the system can accurately assess the impact of coal quality changes on key indicators such as coke thermal strength, and achieve optimal cost while ensuring the quality of coke.

[0016] 3) The optimization results are directly sent to drive the equipment for execution, forming a closed-loop control loop of "perception → prediction → optimization → execution → re-perception", which effectively eliminates the coke quality fluctuations caused by coal type switching and ensures the continuous stability of coal blending production.

[0017] 4) Make full use of the company's existing or under-construction smart coal blending system and silo monitoring system, and achieve linkage through data interface, without the need for repeated construction, which has good practicality and economy. Attached Figure Description

[0018] Figure 1 This is a flowchart of a closed-loop dynamic scheduling method for silo coal blending based on actual coal quality data, as described in this invention.

[0019] Figure 2 This is a schematic diagram of the architecture and data flow of a silo coal blending closed-loop dynamic scheduling system based on actual coal quality data, as described in this invention. Detailed Implementation

[0020] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings: like Figure 1 As shown, the silo coal blending closed-loop dynamic scheduling method based on actual coal quality data of the present invention includes the following steps: (1) The current coal type and real-time physical property index of each silo discharged from the silo discharge coal type real-time identification system are used as the actual coal quality data input into the dynamic scheduling system.

[0021] As a preferred option, the actual coal quality data is updated every second, including the coal type identification, ash content, volatile matter, sulfur content, caking index and other real-time physical property indicators of the actual material discharged at each silo nozzle.

[0022] (2) When the dynamic scheduling system detects that switching coal types in any silo or entering the mixing transition section will cause the quality of the blended coal to deviate from the target, the actual coal quality data and the current feeding rate are transmitted to the intelligent coal blending system.

[0023] (3) The intelligent coal blending system calls the built-in coke quality prediction model and intelligent optimization algorithm, and optimizes the solution with the goal of minimizing raw material cost under the constraint of qualified coke quality, and outputs the optimal feeding rate of each silo to the dynamic scheduling system.

[0024] As a preferred option, the coke quality prediction model is a nonlinear data-driven model based on historical production data through machine learning or neural networks, or a mechanism-data fusion model based on coking mechanism, used to accurately predict the hot strength, cold strength and reactivity index of coke corresponding to blended coal.

[0025] As a preferred option, the intelligent optimization algorithm is a genetic algorithm, differential evolution algorithm, or particle swarm optimization algorithm, which can find the optimal solution in seconds under complex constraints containing nonlinear prediction models.

[0026] (4) The dynamic scheduling system sends the optimal feeding rate to each silo discharge mechanism for execution and continuously monitors and provides feedback to form a closed-loop dynamic scheduling.

[0027] As a preferred option, a rate ramp smooth transition method is adopted when issuing the optimal feeding rate, gradually adjusting from the current rate to the target rate within a preset time to reduce the impact on production.

[0028] The silo coal blending closed-loop dynamic scheduling system based on actual coal quality data described in this invention consists of a dynamic scheduling system and a smart coal blending system dynamic scheduling system. The dynamic scheduling system includes the following components: The data acquisition module is used to obtain actual coal quality data from the silo coal type identification system. The trigger judgment module is used to determine whether dynamic scheduling should be triggered. The communication interface module is used for data interaction with the intelligent coal blending system; The instruction issuing module is used to issue the optimal feeding rate to the PLC / DCS for execution; The intelligent coal blending system incorporates a coke quality prediction model, an intelligent optimization algorithm library, and a coal type cost database.

[0029] Preferably, the silo coal blending closed-loop dynamic scheduling method based on actual coal quality data according to the present invention specifically includes the following steps: I. Real-time acquisition of actual coal quality data This invention obtains the current feeding status data of each silo through a real-time coal type identification system for silo feeding. The real-time coal type identification system for silo feeding adopts the method and system for real-time identification and quality prediction of coal types in silos based on multi-dimensional information fusion (patent filed on the same day), which includes: 1. Coal loading information acquisition module, including a coal loading device located at the top of the silo, used to acquire coal loading start and end time, coal type information and loading quantity.

[0030] 2. 3D radar level gauge, installed at the top of the silo, is used to obtain the three-dimensional shape and storage volume data of the material inside the silo.

[0031] 3. Discharge information acquisition module, including a discharge device located at the bottom of the silo, used to acquire discharge speed and cumulative discharge volume.

[0032] 4. Data processing and algorithm unit, electrically connected to the coal loading information acquisition module, 3D radar level gauge and discharge information acquisition module, integrates real-time material algorithm, discharge self-correction mechanism, discrete element simulation model library and online calling module, used to identify coal type and calculate material properties of mixing section online.

[0033] 5. Coal blending control interface module, which is electrically connected to the data processing and algorithm unit, is used to send the identification and prediction results to the coal blending control system.

[0034] 6. Human-computer interaction interface, electrically connected to the data processing and algorithm unit, used for three-dimensional display of material level shape, coal type area and coal quality trend in the silo.

[0035] The method for real-time identification and quality prediction of coal type in silo feeding based on multi-dimensional information fusion includes the following steps: (1) Coal loading information collection: The start time, stop time, coal name and corresponding test information of different coal types are recorded by the coal loading device set on the top of the silo, and the loading amount of each coal type is counted. (2) Material level detection in the silo: The 3D radar level gauge installed on the top of the silo is used to scan and obtain the three-dimensional shape of the material in the silo in real time, and generate material level height and material storage data. (3) Discharge information collection: The discharge speed and discharge volume are recorded in real time through the discharge device set at the bottom of the silo; (4) Determination of coal type for feeding: The data processing and algorithm unit integrates the coal loading information in step (1), the material level information in the silo in step (2), and the material discharge information in step (3). Using the real-time material algorithm, based on the time sequence of the coal type loaded and the material descent trajectory tracked by the three-dimensional material level data and the cumulative discharge amount, it determines in real time the coal type area to which the material currently discharged from the silo nozzle belongs. The output results are: pure A coal type, pure B coal type, or a mixed section of AB coal type. (5) Estimation of mixed material properties: When the output result is the AB coal type mixed section, the data processing and algorithm unit calls the pre-established discrete element simulation model, combines the sampling and testing data of the AB coal type mixed section, calculates and outputs the real-time comprehensive physical property information of the AB coal type mixed section.

[0036] In this invention, the current discharge status data of each silo is obtained from the real-time coal type identification system, including: the current discharged coal type (pure A coal, pure B coal, or a mixture of A and B coal), and the corresponding real-time physical properties such as ash content, volatile matter, sulfur content, and caking index. Simultaneously, the current feeding rate and adjustable range of each silo discharge device (such as a disc feeder) are acquired. This data is updated every second, reflecting the real-time state of the material at the outlet of each silo's nozzle, constituting "real-time coal quality data."

[0037] II. Dynamic scheduling trigger judgment Based on the actual coal quality data of each silo and the current feeding rate, the comprehensive index of the blended coal is calculated in real time. When it is detected that the coal type of any silo is switched or enters the mixing section (AB coal type mixing section), and continued operation according to the current ratio will cause the quality index of the blended coal to deviate from the target range within a preset time, the dynamic scheduling process is triggered.

[0038] III. Optimization and solution by integrating with the intelligent coal blending system The system transmits the current coal quality data, feeding rate, and cost information for each coal type from all participating coal silos to the intelligent coal blending system via a standard data interface. The intelligent coal blending system then performs the following process: 1. Coke Quality Prediction: The built-in coke quality prediction model is invoked to accurately assess the coke quality of the current coal blending scheme. This model is a data-driven model trained using machine learning, neural networks, and other artificial intelligence technologies based on historical production data, or a mechanism-data fusion model based on coking mechanisms. This model can capture the interaction effects and nonlinear relationships between different coal types during the coking process, accurately predicting key indicators such as hot strength, cold strength, reactivity index, and sulfur content of coke. Its prediction accuracy far exceeds that of simple weighted average calculations.

[0039] 2. Multi-objective optimization modeling: Construct an optimization model with the feeding rate of each silo as the decision variable. The model objective is to minimize the total cost of blended coal raw materials while strictly ensuring the quality of coke. Constraints include: predicting coke quality indicators within the acceptable range, ensuring the feeding rate of each silo is within the equipment's allowable range, and ensuring the total feeding volume meets production requirements.

[0040] 3. Intelligent optimization solution: Intelligent optimization algorithms such as genetic algorithm, differential evolution algorithm or particle swarm optimization algorithm are used to quickly solve the above optimization problems containing nonlinear prediction models, and obtain the optimal feeding rate combination of each silo in seconds.

[0041] IV. Closed-Loop Execution and Continuous Monitoring The optimal feeding rate returned by the intelligent coal blending system is directly transmitted to the PLC / DCS in the dynamic scheduling system via industrial communication protocol, driving the disc feeders in each silo to execute the feed. During execution, the actual feeding rate and coal quality changes in each silo are continuously monitored. When the operating conditions change again, a new round of optimization calculation is automatically triggered, forming a closed-loop control.

[0042] To more intuitively illustrate the present invention, the embodiments are further described in conjunction with the examples. The following examples are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any technical solutions that can be obviously obtained by those skilled in the art within the scope of the technology disclosed in the present invention, including simple variations or equivalent substitutions, are all within the scope of protection of the present invention.

[0043] Example:

[0044] like Figure 1 As shown, the flow of the silo coal blending closed-loop dynamic scheduling method based on actual coal quality data in this embodiment is as follows: S401: The real-time coal type identification system for silo feeding can obtain real-time coal quality data (coal type, ash content, volatile matter, etc.) and current feeding rate for each silo at any time. S402: Real-time calculation of comprehensive indicators of blended coal to predict coke quality trends; S403: Determine if scheduling is triggered, i.e., if any silo enters the coal type switching or mixing section and the coke quality will exceed the limit? If yes, execute S404; otherwise, return to S401 to continue monitoring. S404: Send the current actual coal quality data and cost information of each silo to the intelligent coal blending system; S405: The intelligent coal blending system calls upon the coke quality prediction model to perform accurate coke quality assessment; S406: With coke quality as a constraint and the total raw material cost as the objective, intelligent optimization algorithms such as genetic algorithms are used to solve for the optimal coal blending ratio. S407: The intelligent coal blending system returns the optimal feeding rate for each silo; S408: The instruction issuing module smoothly executes the optimal rate to the disc feeders of each silo; Return to S401 to continue monitoring, forming a closed loop.

[0045] like Figure 2 As shown, the silo coal blending closed-loop dynamic scheduling system based on actual coal quality data in this embodiment consists of two collaborative parts: a dynamic scheduling system and a smart coal blending system. The dynamic scheduling system is responsible for data acquisition, trigger judgment, and command issuance; the smart coal blending system, as a professional computing engine, has a built-in coke quality prediction model and intelligent optimization algorithm, and interacts with the scheduling system through a standardized interface. The two work together to form a complete closed loop of "perception-prediction-optimization-execution".

[0046] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A closed-loop dynamic scheduling method for silo coal blending based on actual coal quality data, characterized in that, Includes the following steps: (1) The current coal type and real-time physical property index of each silo discharged from the silo discharge coal type real-time identification system are used as the actual coal quality data input into the dynamic scheduling system; (2) When the dynamic scheduling system detects that switching coal types in any silo or entering the mixing transition section will cause the quality of the blended coal to deviate from the target, the actual coal quality data and the current feeding rate are transmitted to the intelligent coal blending system. (3) The intelligent coal blending system calls the built-in coke quality prediction model and intelligent optimization algorithm, and optimizes the solution with the goal of minimizing raw material cost under the constraint of qualified coke quality, and outputs the optimal feeding rate of each silo to the dynamic scheduling system. (4) The dynamic scheduling system sends the optimal feeding rate to each silo discharge mechanism for execution and continuously monitors and provides feedback to form a closed-loop dynamic scheduling.

2. The silo coal blending closed-loop dynamic scheduling method based on actual coal quality data according to claim 1, characterized in that, In step (1), the actual coal quality data is updated every second, including the coal type identification, ash content, volatile matter, sulfur content, caking index and other real-time physical property indicators of the actual material discharged at each silo nozzle.

3. The silo coal blending closed-loop dynamic scheduling method based on actual coal quality data according to claim 1, characterized in that, In step (3), the coke quality prediction model is a nonlinear data-driven model established based on historical production data through machine learning or neural networks, or a mechanism-data fusion model established based on coking mechanism, used to accurately predict the hot strength, cold strength and reactivity index of coke corresponding to blended coal.

4. The silo coal blending closed-loop dynamic scheduling method based on actual coal quality data according to claim 1, characterized in that, In step (3), the intelligent optimization algorithm is a genetic algorithm, differential evolution algorithm or particle swarm optimization algorithm, which can find the optimal solution in seconds under complex constraints containing nonlinear prediction models.

5. The silo coal blending closed-loop dynamic scheduling method based on actual coal quality data according to claim 1, characterized in that, In step (4), when issuing the optimal feeding rate, a rate ramp smooth transition method is adopted, and the current rate is gradually adjusted to the target rate within a preset time to reduce the impact on production.

6. A silo coal blending closed-loop dynamic scheduling system based on actual coal quality data, used to implement the silo coal blending closed-loop dynamic scheduling method based on actual coal quality data as described in any one of claims 1 to 5; characterized in that, The silo coal blending closed-loop dynamic scheduling system consists of a dynamic scheduling system and a smart coal blending system dynamic scheduling system. The dynamic scheduling system includes the following components: The data acquisition module is used to obtain actual coal quality data from the silo coal type identification system. The trigger judgment module is used to determine whether dynamic scheduling should be triggered. The communication interface module is used for data interaction with the intelligent coal blending system; The instruction issuing module is used to issue the optimal feeding rate to the PLC / DCS for execution; The intelligent coal blending system incorporates a coke quality prediction model, an intelligent optimization algorithm library, and a coal type cost database.

Images