Blast furnace gas prediction balance scheduling system and method

Abstract

The invention relates to the technical field of blast furnace gas scheduling, and discloses a blast furnace gas prediction balance scheduling system and method. The method comprises a topological structure of pipe network nodes and connecting pipelines, and a blast furnace gas network model is constructed. And acquiring real-time pressure and flow parameters of each node in the model, and calculating to obtain a real-time state characterization value of each node. And calculating the global balance degree of the whole gas network according to the characterization values and the topological connection relationship among the nodes. And generating a gas transmission scheduling decision instruction in combination with the planned demand curve of the gas user, the global balance degree and the state characterization value of each node. According to the method, the physical structure of the pipe network and real-time dynamic data are deeply fused, quantitative evaluation of the overall operation state of the pipe network system is achieved, optimal scheduling of comprehensively considering future requirements and the current system bearing capacity is carried out on the basis, and the stability and economical efficiency of the gas system are improved.

Description

Technical Field

[0001] This invention relates to the field of blast furnace gas dispatching technology, specifically to a blast furnace gas prediction and balance dispatching system and method. Background Technology

[0002] In the field of blast furnace gas transmission and distribution and scheduling, traditional methods typically rely on monitoring and controlling independent parameters such as pressure and flow rate at a limited number of key nodes in the pipeline network. Existing scheduling strategies often employ alarm and linkage response modes based on threshold settings for local key point parameters, or simply weighted averages of data from multiple dispersed monitoring points to assess system status. This approach simplifies the complex pipeline system into a collection of independent monitoring points, neglecting the inherent impact of the network topology formed by the pipeline connections between nodes on gas flow, pressure distribution, and system stability. The actual dynamic behavior of gas in the pipeline network is closely related to the network topology; local adjustments may trigger unexpected network fluctuations. Current technologies lack the means to quantify and assess this network effect, resulting in superficial judgments of the overall system "health" and insufficient basis for decision-making.

[0003] Current technologies typically compare real-time collected raw data such as pressure and flow rate with preset planned demand values ​​when making scheduling decisions, and then adjust valve openings or start / stop equipment accordingly based on the magnitude of the deviation. This binary comparison method of "planned-measured" fails to incorporate the dynamic operating status of the pipeline system at the current moment as a holistic constraint into the decision-making process. The generation process of scheduling instructions severs the intrinsic connection between the system's real-time carrying capacity, the network's global stability, and future planned demands. This can easily lead to situations where scheduling instructions, while meeting short-term needs, may push the pipeline network towards a risky state of pressure imbalance and increased fluctuations, failing to achieve preventative optimization and balance. Summary of the Invention

[0004] The purpose of this invention is to provide a blast furnace gas prediction and balance scheduling system and method to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for predictive balancing and scheduling of blast furnace gas, the method comprising: Define a topology structure that includes multiple network nodes and connecting pipelines to construct a blast furnace gas network model; Obtain the real-time pressure and flow parameters of each pipeline node in the blast furnace gas network model; Based on the real-time pressure and flow parameters of each pipeline node, the real-time state characterization values ​​of each pipeline node in the blast furnace gas network model are calculated. Based on the real-time status characteristics of each pipeline node and its topological connection relationship in the blast furnace gas network model, the global balance of the blast furnace gas network model is calculated. The planned demand curves of gas users are obtained, and gas transmission scheduling decision instructions are generated by combining the global balance degree with the real-time status representation values ​​of each pipeline node in the blast furnace gas network model.

[0006] Preferably, the calculation of the real-time state characterization values ​​of each pipeline node in the blast furnace gas network model based on the real-time pressure and flow parameters of each pipeline node specifically includes the following steps: Time-domain analysis is performed on the real-time pressure and flow parameters of a single pipeline node to extract the pressure fluctuation characteristics and flow change trends of the pipeline node. The pressure fluctuation characteristics and the flow rate change trend are fused to form the primary fusion characteristics of the pipeline node; The real-time pressure parameters of neighboring pipeline nodes that have a direct topological connection with the pipeline node are obtained, and the primary fusion feature is corrected by the neighboring pressure to obtain the real-time status characterization value of the pipeline node. For each pipeline node in the blast furnace gas network model, the steps of time-domain analysis, feature fusion processing, and adjacent pressure correction are repeated to obtain the real-time state characterization values ​​of all pipeline nodes.

[0007] Preferably, the calculation of the global balance of the blast furnace gas network model based on the real-time status characteristics of each pipeline node and its topological connection relationship in the blast furnace gas network model specifically includes the following steps: Read the topological connections of the blast furnace gas network model and determine the weight coefficients of each connecting pipeline; Based on the weight coefficients of each connecting pipeline, the weighted difference of the real-time status representation values ​​of the network nodes at both ends of the connecting pipeline is calculated to obtain the local balance index of each connecting pipeline. The local balance indices of all connected pipelines in the blast furnace gas network model are summarized, normalized and aggregated, and the global balance index, which represents the stability of the entire pipeline network, is output.

[0008] Preferably, obtaining the planned demand curve of gas users and combining it with the global balance degree and the real-time status representation values ​​of each pipeline node in the blast furnace gas network model to generate gas transmission scheduling decision instructions specifically includes the following steps: Analyze the planned demand curves of gas users and decompose them into time-period demand sets for different target pipeline nodes; The global balance degree is matched and compared with the real-time status representation value of the target pipeline node in the blast furnace gas network model to determine the matching status between gas supply and planned demand. When the matching status indicates a supply deviation, a gas delivery scheduling decision instruction is generated based on the time-segmented demand set and the supply deviation value, which includes the adjustment target node, adjustment amount, and adjustment timing.

[0009] Preferably, the proximity pressure correction for the primary fusion features specifically includes the following steps: Extract the real-time pressure parameters of all neighboring pipeline nodes that have a direct topological connection with the pipeline node; Calculate the average real-time pressure parameters of all adjacent pipeline nodes as a reference value for adjacent pressure; The primary fusion feature is coupled with the neighboring pressure reference value, and the result of the coupling operation is used to update the primary fusion feature, thereby completing the neighboring pressure correction.

[0010] Preferably, the process of reading the topological connections of the blast furnace gas network model and determining the weight coefficients of each connecting pipeline includes the following steps: Obtain historical flow data and pipe diameter parameters for each connecting pipeline in the blast furnace gas network model; The flow stability index for each connecting pipeline is calculated based on the historical flow data, specifically including the following steps: Extract a sequence of historical traffic data within a specified time period; The standard deviation and trend slope of the historical traffic data sequence are calculated. The calculated standard deviation is combined with the trend slope to map a flow stability index. By combining the pipe diameter parameters and the flow stability index, a corresponding weighting coefficient is assigned to each connecting pipeline using a lookup table method.

[0011] Preferably, matching and comparing the global balance degree with the real-time state representation values ​​of the target pipeline nodes in the blast furnace gas network model specifically includes the following steps: Establish matching comparison rules, which define matching thresholds corresponding to different numerical ranges of global balance. Obtain the real-time status representation value of the target pipeline node, and calculate the expected status value of the target pipeline node based on the planned demand curve of the gas user; Calculate the difference between the real-time state representation value and the expected state value, determine whether the absolute value of the difference exceeds the matching threshold corresponding to the current global balance value range, and then determine the matching state.

[0012] Preferably, after generating the gas transmission scheduling decision instruction, the method further includes the following steps: Send gas delivery scheduling decision instructions to the regulating valves and compressors on the gas delivery pipeline; After the regulating valve and compressor execute the gas delivery scheduling decision command, monitor the changes in real-time pressure and flow parameters of relevant pipeline nodes in the blast furnace gas network model; Based on the monitored changes, the real-time status representation values ​​of relevant pipeline nodes in the blast furnace gas network model are updated, and the global balance of the blast furnace gas network model is recalculated to form a closed-loop feedback.

[0013] Preferably, the definition includes a topology structure of multiple pipeline nodes and connecting pipelines, and the construction of the blast furnace gas network model specifically includes the following steps: Collect equipment data from blast furnaces, gas holders, gas users, regulating valves and compressors at the production site, define them as pipeline nodes, and assign a unique node identifier to each pipeline node. Based on the actual physical connection of the gas transmission pipeline, determine the connection relationship between each pipeline node, and define the pipeline connecting two adjacent pipeline nodes as the connecting pipeline. The entire set of pipeline nodes, their corresponding node identifiers, and the topological connections represented by all connecting pipelines are stored in the form of a graph structure or an adjacency matrix to complete the construction of the blast furnace gas network model.

[0014] Preferably, the present invention also includes a blast furnace gas prediction and balancing scheduling system, the system including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, it implements the steps of the blast furnace gas prediction and balancing scheduling method described above.

[0015] Compared with the prior art, the beneficial effects of the present invention are: By establishing a network model that includes the connections between nodes and pipelines, and calculating the global balance degree based on the real-time status representation values ​​of each node and its specific topological connections, this method achieves a networked and structured assessment of the operational status of the blast furnace gas pipeline system. The global balance degree parameter is no longer a simple statistical representation of individual node data, but rather a profound reflection of the balanced distribution and transmission of gas energy throughout the entire network topology. This enables the scheduling system to perceive vulnerable links in the network structure, identify the pressure transmission and cumulative fluctuation effects caused by topological connections, and thus accurately grasp the stable state of the system as a whole, rather than just at local points. This provides a key state indicator reflecting the inherent interconnectivity of the network, which is not available in traditional methods, for subsequent decision-making.

[0016] This mechanism integrates real-time state characteristics reflecting node dynamic load and safety margin, global balance representing overall network stability, and predicted planned demand curves, all as inputs for generating scheduling instructions. This constructs a multi-layered constraint decision-making model. The decision generation process not only considers "how much is needed" but also simultaneously assesses "how much can be safely and stably supported currently." This enables the generated scheduling instructions to proactively avoid pushing critical network nodes to their limits to meet demand, or triggering drastic fluctuations across the network, achieving a forward-looking optimization balance between demand satisfaction and the real-time dynamic safe operation of the network system. Scheduling behavior shifts from passive deviation correction to proactive, preventative adjustment based on multi-dimensional system state prediction. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the working principle of the blast furnace gas prediction and balance scheduling method described in this invention. Figure 2 Flowchart for adjacent pressure correction; Figure 3 A flowchart for calculating global balance; Figure 4 A diagram illustrating the matching status between gas supply and planned demand. Figure 5 This is a monitoring graph showing the changes in pipeline node parameters after the execution of scheduling instructions. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1This invention provides a method for predictive balancing and scheduling of blast furnace gas. The method includes: defining a topology containing multiple pipeline nodes and connecting pipelines to construct a blast furnace gas network model representing the actual blast furnace gas transportation system; acquiring the pressure and flow parameters of each pipeline node in the blast furnace gas network model in real time; calculating the real-time status representation value of each pipeline node based on the real-time pressure and flow parameters of each node, which comprehensively reflects the node's operating status at a specific moment; and calculating the global balance degree, which reflects the supply and demand and operational stability of the entire pipeline system, by combining the real-time status representation values ​​of each node with their topological connections in the network model. The method also includes acquiring the planned demand curve of downstream gas users for a future period; and combining the planned demand curve, the calculated global balance degree, and the real-time status representation values ​​of each node for analysis and decision-making to generate a gas transportation scheduling decision instruction to guide adjustments to the gas transportation system.

[0020] Example 1: See Figure 2 The time-domain analysis of real-time pressure and flow parameters of a single pipeline node is performed to extract the pressure fluctuation characteristics and flow change trends of the node over a period of time. The extracted pressure fluctuation characteristics and flow change trends are then fused to form the primary fused features of the pipeline node. The real-time pressure parameters of all neighboring pipeline nodes with direct topological connections to the current node are obtained, and the average value of these neighboring nodes' real-time pressure parameters is calculated and used as the neighboring pressure reference value. The primary fused features are coupled with the calculated neighboring pressure reference value, and the result of this coupling operation is used to update the primary fused features, thereby completing the neighboring pressure correction and obtaining the real-time state characterization value of the pipeline node. For each pipeline node in the blast furnace gas network model, the above steps of time-domain analysis, feature fusion processing, and neighboring pressure correction are repeated to obtain the real-time state characterization values ​​of all pipeline nodes.

[0021] In a specific implementation, the calculation of the real-time status characterization value of a pipeline node is achieved through the following steps: In a specific implementation, time-domain analysis is performed on the real-time pressure and flow parameters of a single pipeline node to extract the pressure fluctuation characteristics and flow change trends. The time-domain analysis calculates the statistics of the pressure and flow parameter time series. The pressure fluctuation characteristics are obtained by calculating the standard deviation of the pressure parameter time series, and the flow change trend is obtained by calculating the slope of the linear regression of the flow parameter time series. In some embodiments, the pressure fluctuation characteristics and flow change trends are fused to form the primary fused characteristics of the pipeline node. The feature fusion process weights the pressure fluctuation characteristic values ​​and the flow change trend values. The summation is combined into a single scalar value. Optionally, the weights of the weighted summation are pre-set based on the type of pipeline node. In specific implementations, the real-time pressure parameters of all neighboring pipeline nodes with direct topological connections to the pipeline node are obtained. The average value of the real-time pressure parameters of all neighboring pipeline nodes is calculated as a neighboring pressure reference value. In some embodiments, the identification of neighboring pipeline nodes relies on the topological connections stored in the blast furnace gas network model. All directly connected nodes are determined by querying the connection relationships. It can be understood that the primary fusion feature is coupled with the neighboring pressure reference value, and the result of the coupling operation is used to update the primary fusion feature, thereby completing the neighboring pressure correction and obtaining the real-time status characterization value of the pipeline node. The coupling operation uses the following formula: in: This represents the real-time status indicator value of the pipeline network node. This indicates the initial fusion characteristics of pipeline network nodes. Indicates the nearby pressure reference value. The correction coefficient is a constant between 0 and 1. In practice, the specific value of the correction coefficient is obtained by calibrating the historical operation data of the pipeline network. Optionally, the correction coefficient can take different preset values ​​for different types of pipeline nodes. It can be understood that for each pipeline node in the blast furnace gas network model, the steps of time domain analysis, feature fusion processing and adjacent pressure correction are repeatedly performed to obtain the real-time status characterization value of all pipeline nodes.

[0022] Example 2: See Figure 3The topological connections of the blast furnace gas network model are read to determine the weight coefficients of each connecting pipeline. Specifically, historical flow data and pipe diameter parameters of each connecting pipeline in the blast furnace gas network model are obtained. The flow stability index of each connecting pipeline is calculated based on the historical flow data. The process involves extracting a historical flow data sequence within a set time period, calculating the standard deviation and trend slope of the historical flow data sequence, and combining the calculated standard deviation and trend slope to map a flow stability index. Combining the pipe diameter parameters and the calculated flow stability index, a corresponding weight coefficient is assigned to each connecting pipeline using a lookup table method. After obtaining the weight coefficients of each connecting pipeline, a weighted difference calculation is performed on the real-time state representation values ​​of the network nodes at both ends of the connecting pipeline based on the weight coefficients of each connecting pipeline to obtain the local balance index of each connecting pipeline. The local balance indices of all connecting pipelines in the blast furnace gas network model are summarized, normalized, and aggregated to output the global balance index, which represents the overall stability of the network.

[0023] In a specific implementation, the calculation of the global balance of the blast furnace gas network model is achieved through the following steps: The topological connections of the blast furnace gas network model are read, the weight coefficients of each connecting pipeline are determined, historical flow data and pipe diameter parameters of each connecting pipeline in the blast furnace gas network model are obtained, the flow stability index of each connecting pipeline is calculated based on the historical flow data, a historical flow data sequence within a set time period is extracted, and the standard deviation and trend slope of this historical flow data sequence are calculated. It can be understood that the calculated standard deviation and trend slope are combined and mapped to a flow stability index. In some embodiments, the combination method involves substituting the reciprocal of the standard deviation and the absolute value of the trend slope into a linear combination formula. Optionally, the coefficients in the linear combination formula are... Based on prior knowledge and combined with pipe diameter parameters and calculated flow stability indices, a corresponding weight coefficient is assigned to each connecting pipeline using a lookup table method. In specific implementation, a two-dimensional mapping table is constructed. The row index of the table represents the discrete pipe diameter parameter range, and the column index represents the discrete flow stability index range. Each cell stores a pre-set weight coefficient value. During a query, the range to which the connecting pipeline belongs is determined based on the actual pipe diameter parameters and the calculated flow stability index, thereby retrieving the corresponding weight coefficient from the table. In some embodiments, based on the weight coefficients of each connecting pipeline, a weighted difference is calculated on the real-time status representation values ​​of the network nodes at both ends of the connecting pipeline to obtain the local balance index of each connecting pipeline. The weighted difference calculation follows the following formula: in: Indicates the first The local balance index of the connecting pipeline, Indicates the first The weighting coefficients assigned to each connecting pipeline. This indicates a network node connecting one end of the pipeline. The real-time state representation value, This indicates the network node at the other end of the connecting pipeline. The real-time status representation value can be understood as summing up the local balance indexes of all connected pipelines in the blast furnace gas network model, performing a normalized aggregation operation. The normalized aggregation operation sums up the local balance indexes of all connected pipelines and then divides by the total number of connected pipelines to output the global balance degree, which represents the stability of the entire pipeline network. Optionally, the global balance degree can also be expressed as the reciprocal of the arithmetic mean of the local balance indexes of all connected pipelines.

[0024] Example 3: Analyze the planned demand curve of gas users and decompose it into time-segmented demand sets for different target pipeline nodes. Compare the calculated global balance degree with the real-time status representation values ​​of the target pipeline nodes in the blast furnace gas network model to determine the matching status between gas supply and planned demand. The matching comparison process is as follows: Establish matching comparison rules, which define matching thresholds corresponding to different numerical ranges of the global balance degree; obtain the real-time status representation values ​​of the target pipeline nodes; calculate the expected status value of the target pipeline node based on the planned demand curve of the gas users; calculate the difference between the real-time status representation value and the expected status value; determine whether the absolute value of this difference exceeds the matching threshold corresponding to the current global balance degree numerical range, and thus determine the matching status. When the matching status indicates a supply deviation, generate a gas delivery scheduling decision instruction containing the adjustment target node, adjustment amount, and adjustment timing based on the time-segmented demand set and the supply deviation value.

[0025] In a specific implementation, the generation of gas transmission scheduling decision instructions is achieved through the following steps: First, the planned demand curve of gas users is analyzed and decomposed into time-segmented demand sets for different target pipeline nodes. The planned demand curve is a function of time and total demand. The decomposition process is based on the correspondence between gas user nodes, upstream source nodes, and pipeline nodes in the blast furnace gas network model. The total demand is calculated to each target pipeline node according to a preset allocation ratio and then divided into time-segmented sequences with fixed durations. This can be understood as matching and comparing the global balance with the real-time status representation values ​​of the target pipeline nodes in the blast furnace gas network model to determine the matching status between gas supply and planned demand. The matching and comparison process involves: establishing matching and comparison rules, which define the global balance... The different numerical ranges of the balance degree correspond to different matching thresholds. In some embodiments, the global balance degree numerical range is divided into three continuous ranges: high, medium, and low. Each range is associated with a specific matching threshold value. The matching threshold increases as the global balance degree numerical range decreases. The real-time status representation value of the target pipeline node is obtained, and the expected status value of the target pipeline node is calculated based on the gas user's planned demand curve. The calculation of the expected status value involves mapping the current time-period demand in the time-period demand set for the target pipeline node to a value with the same dimensions as the real-time status representation value through a linear transformation function. In specific implementations, the difference between the real-time status representation value and the expected status value is calculated, and it is determined whether the absolute value of the difference exceeds the matching threshold corresponding to the current global balance degree numerical range. The matching comparison logic can be expressed by the following formula: in: This represents the real-time status indicator value of the target pipeline node. This represents the expected state value of the target pipeline node. Indicates global balance. Indicates global balance The matching threshold function corresponding to the numerical range can be understood as follows: when the matching status indicates a supply deviation, a gas transmission scheduling decision instruction is generated based on the time-segmented demand set and the supply deviation value, which includes the adjustment target node, adjustment amount, and adjustment timing. In some embodiments, the supply deviation value is the difference between the real-time status characterization value and the expected status value. The adjustment target node is the target pipeline node currently judged to be mismatched or its upstream key regulating node. The adjustment amount is calculated based on the supply deviation value and an adjustable gain coefficient. The adjustment timing is directly taken from the time-segmented sequence obtained by decomposing the planned demand curve. Optionally, the gas transmission scheduling decision instruction is generated in the form of a structured data list. Each record in the list includes the target node identifier, the suggested flow or pressure value for adjustment, and the start and end timestamps of the adjustment operation.

[0026] See Figure 4 This chart illustrates the changes in the real-time and expected status values ​​of target pipeline nodes over time. The solid blue line represents the real-time status value, reflecting the current actual operating status of the pipeline node; the dashed red line represents the expected status value, calculated based on the planned demand curve. The green area in the chart represents the matching threshold range; when the real-time status value falls within this range, it indicates a good match between gas supply and planned demand. The orange vertical lines mark the mismatch points, indicating that the difference between the real-time and expected status exceeds the matching threshold, resulting in supply deviations. The upper right corner of the chart displays the overall matching rate, which quantifies the degree of matching between supply and demand. The chart also shows the dynamic adjustment of the matching threshold with changes in global balance; the threshold is smaller in the high balance range, requiring more precise matching; the threshold is larger in the low balance range, allowing for larger deviations. This data provides important information for scheduling decisions. When the matching rate is low or there are significant supply deviations, the system generates corresponding scheduling instructions to adjust gas delivery.

[0027] Example 4: After generating the gas delivery scheduling decision command, the command is sent to the regulating valves and compressors on the gas delivery pipeline. The changes in real-time pressure and flow parameters of relevant pipeline nodes in the blast furnace gas network model are monitored after these regulating valves and compressors execute the gas delivery scheduling decision command. Based on the monitored changes, the real-time state characterization values ​​of relevant pipeline nodes in the blast furnace gas network model are updated, and the global balance of the blast furnace gas network model is recalculated to form a closed-loop feedback.

[0028] In a specific implementation, the closed-loop feedback process of the method is executed after the gas transmission scheduling decision command is generated. In practice, the gas transmission scheduling decision command is sent to the regulating valves and compressors on the gas transmission pipeline. The sending process is completed through an industrial control network or data bus. The gas transmission scheduling decision command is parsed in a predefined data message format. The regulating valves receive the opening adjustment amount and adjustment timing information in the command, and the compressors receive the frequency or power adjustment amount and adjustment timing information in the command. In some embodiments, the changes in real-time pressure parameters and real-time flow parameters of relevant pipeline nodes in the blast furnace gas network model are monitored after the regulating valves and compressors execute the gas transmission scheduling decision command. The monitoring is carried out by sensors deployed near the relevant pipeline nodes. The sensors read pressure and flow data at a fixed sampling period. Optionally, the monitoring continues for a complete adjustment timing cycle. After the adjustment timing cycle ends, the monitoring data is summarized. Refer to Table 1, which shows the monitoring data of some relevant pipeline nodes after one command execution.

[0029] Table 1: Monitoring Table of Pipeline Node Parameters After Command Execution It is understandable that, based on the monitored changes, the real-time status representation values ​​of relevant pipeline nodes in the blast furnace gas network model are updated. The update process calls the calculation method for real-time status representation values, using the newly monitored real-time pressure and flow parameters as inputs to recalculate the real-time status representation values ​​of relevant pipeline nodes. In some embodiments, the global balance of the blast furnace gas network model is also recalculated. The complete calculation process for global balance is executed based on the updated real-time status representation values ​​of each node and the original topological connections to form a closed-loop feedback. Optionally, the recalculated global balance will be compared with the global balance of the previous period, and the change will be recorded. Calculated using the following formula: in: This represents the change in the overall balance. This represents the newly calculated global balance. This represents the overall balance before the execution of gas transmission scheduling decision instructions; this change can be understood as... The data will be recorded and used to evaluate the effectiveness of the current gas transmission scheduling decision, and to provide data input for any subsequent possible new round of scheduling decisions.

[0030] See Figure 5 This chart displays the monitoring of changes in pipeline node parameters after the execution of a dispatch command. It employs a dual Y-axis design, with the left Y-axis displaying pressure parameters (unit: kPa) and the right Y-axis displaying flow parameters (unit: m³ / h). The blue solid line connects pressure measurements at different monitoring time points, and the orange solid line connects flow measurements, reflecting the trends of both parameters over time. The red dashed line represents the original pressure value before adjustment, and the green dashed line represents the target pressure value after adjustment; the red dotted line represents the original flow value before adjustment, and the green dotted line represents the target flow value after adjustment. Monitoring points are distributed at different time intervals (5 minutes, 15 minutes, 30 minutes, and 60 minutes) after the command execution, demonstrating the dynamic process of parameters gradually approaching the target value from their original values. The text label in the upper left corner of the chart shows the percentage changes in pressure and flow, quantifying the adjustment effect. The upper right corner indicates the monitoring time point information. This chart allows for a direct assessment of the execution effect of dispatch commands, determining whether regulating valves and compressors are operating as expected, and whether pipeline node parameters have reached the target state. These monitoring data are used to update the real-time status representation values ​​of the nodes and recalculate the global balance, forming a complete closed-loop feedback control loop to ensure that the gas transmission system can be continuously optimized and adaptively adjusted.

[0031] Example 5: Collect equipment data from the blast furnace, gas holder, gas users, regulating valves, and compressors at the production site, define them as pipeline network nodes, and assign a unique node identifier to each pipeline network node. Based on the actual physical connections of the gas transmission pipelines, determine the connection relationships between each pipeline network node, and define the pipeline connecting two adjacent pipeline network nodes as connecting pipelines. Store the set of all pipeline network nodes, their corresponding node identifiers, and the topological connection relationships represented by all connecting pipelines in the form of a graph structure or adjacency matrix to complete the construction of the blast furnace gas network model.

[0032] In a specific implementation, the blast furnace gas network model is constructed through the following steps: Equipment data from the blast furnace, gas holder, gas users, regulating valves, and compressors at the production site are collected. The blast furnace equipment data includes at least the blast furnace number and its corresponding process unit; the gas holder equipment data includes at least the gas holder number and volume; the gas user equipment data includes at least the user number and maximum consumption; the regulating valve equipment data includes at least the valve number and installation location; and the compressor equipment data includes at least the compressor number and rated power. Essentially, the blast furnace, gas holder, gas users, regulating valves, and compressors are defined as network nodes, and each network node is assigned a unique node identifier. The node identifier assignment rule adopts a "type code-serial number" encoding format. In some embodiments, the type code is represented by two uppercase letters, and the serial number is represented by four digits. For example, blast furnace gas holders, gas users, regulating valves, and compressors are defined as network nodes. The furnace node is identified as "BF-0001", and the gas holder node is identified as "TG-0001". Fields related to node identity in the collected equipment data will be associated with and stored with these node identifiers. In practice, the connection relationships between each network node are determined based on the actual physical connections of the gas transmission pipelines. This determination process is based on the plant's Piping & Instrumentation Flow Diagram (P&ID) and on-site inspection records. It can be understood that the pipeline connecting two adjacent network nodes is defined as a connecting pipeline. Optionally, the connecting pipeline records the node identifiers of the network nodes it connects to at both ends, and includes the pipeline length and initial design flow parameters. The set of all network nodes, their corresponding node identifiers, and the topological connection relationships represented by all connecting pipelines are stored in the form of a graph structure or adjacency matrix to complete the construction of the blast furnace gas network model. In some embodiments, an adjacency matrix is ​​used for storage. Dimensions Equal to the total number of network nodes, matrix elements The possible values ​​of are defined as follows: in: Represents the adjacency matrix of the nth element. Line 1 The elements of the column are used to represent the nodes. With nodes Does a direct topological connection exist between them? When, it indicates the existence of a connecting pipeline; when When there is no connection, it means there are no connecting pipelines. Optionally, the adjacency matrix of the undirected pipeline network is a symmetric matrix, and the connection relationships of the pipeline nodes themselves are not stored, that is, all the elements on the main diagonal of the matrix are 0. It can be understood that after the construction of the adjacency matrix or graph structure is completed, the blast furnace gas network model has the topological structure foundation that can be parsed and calculated by computer programs. Subsequent calculations of real-time state representation values ​​and global balance are all based on this model.

[0033] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for predictive balancing and scheduling of blast furnace gas, characterized in that, Perform the following operations: Define a topology structure that includes multiple network nodes and connecting pipelines to construct a blast furnace gas network model; Obtain the real-time pressure and flow parameters of each pipeline node in the blast furnace gas network model; Based on the real-time pressure and flow parameters of each pipeline node, the real-time state characterization values ​​of each pipeline node in the blast furnace gas network model are calculated. Based on the real-time status characteristics of each pipeline node and its topological connection relationship in the blast furnace gas network model, the global balance of the blast furnace gas network model is calculated. The planned demand curves of gas users are obtained, and gas transmission scheduling decision instructions are generated by combining the global balance degree with the real-time status representation values ​​of each pipeline node in the blast furnace gas network model.

2. The blast furnace gas prediction and balance scheduling method as described in claim 1, characterized in that, The calculation of the real-time state characterization values ​​of each pipeline node in the blast furnace gas network model based on the real-time pressure and flow parameters of each pipeline node includes the following steps: Time-domain analysis is performed on the real-time pressure and flow parameters of a single pipeline node to extract the pressure fluctuation characteristics and flow change trends of the pipeline node. The pressure fluctuation characteristics and the flow rate change trend are fused to form the primary fusion characteristics of the pipeline node; The real-time pressure parameters of neighboring pipeline nodes that have a direct topological connection with the pipeline node are obtained, and the primary fusion feature is corrected by the neighboring pressure to obtain the real-time status characterization value of the pipeline node. For each pipeline node in the blast furnace gas network model, the steps of time-domain analysis, feature fusion processing, and adjacent pressure correction are repeated to obtain the real-time state characterization values ​​of all pipeline nodes.

3. The blast furnace gas prediction and balance scheduling method as described in claim 1, characterized in that, The global balance of the blast furnace gas network model is calculated based on the real-time status characteristics of each pipeline node and its topological connectivity in the blast furnace gas network model, specifically including the following steps: Read the topological connections of the blast furnace gas network model and determine the weight coefficients of each connecting pipeline; Based on the weight coefficients of each connecting pipeline, the weighted difference of the real-time status representation values ​​of the network nodes at both ends of the connecting pipeline is calculated to obtain the local balance index of each connecting pipeline. The local balance indices of all connected pipelines in the blast furnace gas network model are summarized, normalized and aggregated, and the global balance index, which represents the stability of the entire pipeline network, is output.

4. The blast furnace gas prediction and balance scheduling method as described in claim 1, characterized in that, Obtaining the planned demand curves of gas users, and combining the global balance with the real-time status representation values ​​of each pipeline node in the blast furnace gas network model, the specific steps for generating gas transmission scheduling decision instructions include: Analyze the planned demand curves of gas users and decompose them into time-period demand sets for different target pipeline nodes; The global balance degree is matched and compared with the real-time status representation value of the target pipeline node in the blast furnace gas network model to determine the matching status between gas supply and planned demand. When the matching status indicates a supply deviation, a gas delivery scheduling decision instruction is generated based on the time-segmented demand set and the supply deviation value, which includes the adjustment target node, adjustment amount, and adjustment timing.

5. The blast furnace gas prediction and balance scheduling method as described in claim 2, characterized in that, The process of performing proximity pressure correction on the primary fusion features specifically includes the following steps: Extract the real-time pressure parameters of all neighboring pipeline nodes that have a direct topological connection with the pipeline node; Calculate the average real-time pressure parameters of all adjacent pipeline nodes as a reference value for adjacent pressure; The primary fusion feature is coupled with the neighboring pressure reference value, and the result of the coupling operation is used to update the primary fusion feature, thereby completing the neighboring pressure correction.

6. The blast furnace gas prediction and balance scheduling method as described in claim 3, characterized in that, The steps to read the topological connections of the blast furnace gas network model and determine the weight coefficients of each connecting pipeline include: Obtain historical flow data and pipe diameter parameters for each connecting pipeline in the blast furnace gas network model; The flow stability index for each connecting pipeline is calculated based on the historical flow data, specifically including the following steps: Extract a sequence of historical traffic data within a specified time period; The standard deviation and trend slope of the historical traffic data sequence are calculated. The calculated standard deviation is combined with the trend slope to map a flow stability index. By combining the pipe diameter parameters and the flow stability index, a corresponding weighting coefficient is assigned to each connecting pipeline using a lookup table method.

7. The blast furnace gas prediction and balance scheduling method as described in claim 4, characterized in that, The process of matching and comparing the global balance with the real-time state representation values ​​of the target pipeline nodes in the blast furnace gas network model specifically includes the following steps: Establish matching comparison rules, which define matching thresholds corresponding to different numerical ranges of global balance. Obtain the real-time status representation value of the target pipeline node, and calculate the expected status value of the target pipeline node based on the planned demand curve of the gas user; Calculate the difference between the real-time state representation value and the expected state value, determine whether the absolute value of the difference exceeds the matching threshold corresponding to the current global balance value range, and then determine the matching state.

8. The blast furnace gas prediction and balance scheduling method as described in claim 1, characterized in that, After generating the gas transmission scheduling decision instruction, the method further includes the following steps: Send gas delivery scheduling decision instructions to the regulating valves and compressors on the gas delivery pipeline; After the regulating valve and compressor execute the gas delivery scheduling decision command, monitor the changes in real-time pressure and flow parameters of relevant pipeline nodes in the blast furnace gas network model; Based on the monitored changes, the real-time status representation values ​​of relevant pipeline nodes in the blast furnace gas network model are updated, and the global balance of the blast furnace gas network model is recalculated to form a closed-loop feedback.

9. The blast furnace gas prediction and balance scheduling method as described in claim 1, characterized in that, The definition includes a topology structure of multiple network nodes and connecting pipelines. The specific steps for constructing a blast furnace gas network model are as follows: Collect equipment data from blast furnaces, gas holders, gas users, regulating valves and compressors at the production site, define them as pipeline nodes, and assign a unique node identifier to each pipeline node. Based on the actual physical connection of the gas transmission pipeline, determine the connection relationship between each pipeline node, and define the pipeline connecting two adjacent pipeline nodes as the connecting pipeline. The entire set of pipeline nodes, their corresponding node identifiers, and the topological connections represented by all connecting pipelines are stored in the form of a graph structure or an adjacency matrix to complete the construction of the blast furnace gas network model.

10. A blast furnace gas prediction and balancing scheduling system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the blast furnace gas prediction and balance scheduling method as described in any one of claims 1 to 9.

Images