Coking moisture control system based on atomized spraying

Abstract

The present application relates to the technical field of industrial process control, in particular to a coking moisture control system based on atomized spraying, in which the spatial distribution state of heat in the oven can be accurately judged by collecting the temperature data of the coke oven in a specified period and constructing the heat flux change trend, and the abnormal heat load area is identified based on the heat accumulation change difference, and then the target section with potential moisture regulation value is locked, before the execution of the spraying action, the heat conduction condition and the disturbance propagation direction are comprehensively judged by combining the heat conduction response time of the spraying point to the oven wall direction and the spatial extension structure of the heat disturbance path, so that the execution order and the continuous parameters of the spraying task are dynamically adjusted, the spraying area selection error caused by resource limitation is avoided, and the spraying target is screened and rearranged through the path connectivity and the temperature zone coverage relationship before the spraying instruction is formed, so that the moisture regulation behavior not only has the ability to respond to the heat state change in real time.

Description

Technical Field

[0001] This invention relates to the field of industrial process control technology, and in particular to a coking moisture control system based on atomized spraying. Background Technology

[0002] The field of industrial process control technology involves monitoring, adjusting, and optimizing various parameters in industrial production processes to ensure the stability, safety, and controllability of product quality. Its core aspects include online detection and feedback control of physical or chemical quantities such as temperature, pressure, flow rate, concentration, and moisture.

[0003] Traditional coking moisture control refers to the control of the moisture content of coal during the coking process in order to ensure the quality of coke and the stability of coke oven operation, especially the moisture adjustment before and after coal charging or during the coking and exiting the oven.

[0004] Traditional coking moisture control relies solely on static adjustments before and after coal charging or during the tapping stage, lacking the ability to continuously sense the heat load within the furnace. Under variable thermal conditions, it cannot acquire the thermal response characteristics of different areas within the coke oven in real time, resulting in lagging and vague control strategies. In actual operation, it is easy for the spraying location to be misaligned with the thermal anomaly area. Especially during high-temperature operation, if the abnormal heat zone is not accurately covered, the spraying behavior not only fails to suppress excessive moisture evaporation but may also trigger the spread of thermal disturbance, causing aggravated furnace temperature fluctuations and affecting coke maturity and structural stability. For example, in situations where water resources are scarce or the heat conduction path within the furnace is complex, traditional methods struggle to rationally allocate the rhythm and intensity of moisture control, thereby increasing the operational burden and reducing control efficiency. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the existing technology and propose a coking moisture control system based on atomized spraying.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a coking moisture control system based on atomized spraying includes:

[0007] The heat storage assessment module obtains the coke oven temperature for a specified period during the coking process and establishes a heat storage index sequence by calculating the cumulative change in heat flux of the coke oven bricks.

[0008] The heat load area identification module identifies abnormal heat load areas in the coke oven by referring to the heat storage index sequence, extracts abnormal load areas as spraying targets and constructs a spraying task list.

[0009] The thermal feedback triggering module determines the spray target switching logic based on the spraying task list, collects heat flux density data from the spraying point to the corresponding furnace wall, calculates the thermal conduction response time of the coke oven brick, and generates a corresponding thermal feedback triggering signal group.

[0010] The thermal disturbance path construction module constructs the thermal disturbance path within the coke oven temperature range through the thermal conduction feedback trigger signal group, and filters the set of thermal disturbance response segments along the direction of the thermal disturbance path;

[0011] The spraying instruction generation module modifies the spraying targets in the spraying task list based on the set of thermal disturbance response segments, and constructs spraying execution instructions for quantitative control of coal moisture release during the coking process by atomized spraying behavior.

[0012] As a further aspect of the present invention, the heat storage index sequence includes heat flux change distribution, coke oven temperature zone heat energy accumulation structure, and time-series temperature response characteristics; the spraying task list includes the distribution location of abnormal heat load areas, target area screening range, and spraying control cycle; the heat conduction feedback trigger signal group includes response effective area identification marker, heat conduction path confirmation information, and spraying response critical conditions; the thermal disturbance response segment set includes continuous thermal disturbance path area, heat conduction connection area, and feedback propagation direction structure; and the spraying execution command includes spraying target area, spraying duration, and spraying intensity set value.

[0013] As a further aspect of the present invention, the thermal storage evaluation module includes:

[0014] The temperature data acquisition submodule acquires coke oven temperature data for a specified period during the coking process, and collects continuous temperature sequences recorded by thermocouples set in the upper, middle and lower layers after the coal is fed into the furnace, generating a temperature time series dataset.

[0015] The heat flux integration submodule obtains the temperature difference between adjacent time points in the time-series temperature dataset. Combining the thermal conductivity of the measuring point with the corresponding area parameter, it calculates the total heat flux change of each temperature difference within a specified time period using the Simpson integral method to obtain the cumulative heat flux value.

[0016] The index sequence construction submodule extracts the temperature change rate of each measuring point, pairs the temperature change rate with the cumulative heat flux value according to the correspondence of the measuring points, and completes the combination arrangement according to the spatial layout order of the coke oven to establish a heat storage index sequence.

[0017] As a further aspect of the present invention, the heat load area identification module includes:

[0018] The abnormal symptom extraction submodule extracts the heat storage value corresponding to each measuring point based on the heat storage index sequence, and obtains the local difference of each group of measuring points under the condition of regional continuity, and constructs the heat storage change distribution information.

[0019] The critical section identification submodule uses the local outlier factor algorithm to calculate the relative density ratio between each measuring point and its adjacent measuring points in the heat storage change distribution, and delineates the areas that meet the outlier criteria as heat load abnormal sections, thus obtaining a set of abnormal areas.

[0020] The spraying target generation submodule excludes areas with limited water supply in the current period based on the abnormal area set, and sets task numbers and period markers for the remaining areas to create a spraying task list.

[0021] As a further aspect of the present invention, the thermal feedback triggering module includes:

[0022] The spray point positioning submodule extracts the action position corresponding to each spray target in the spray task list, determines the coordinates of each spray point on the coke oven brick and the corresponding contact surface of the adjacent furnace wall, and obtains the spray-furnace wall mapping information.

[0023] The thermal response calculation submodule obtains the data on the change of heat flux density over time on the path from the spray contact point to the adjacent furnace wall based on the spray-furnace wall mapping information, and calculates the time required for the heat flux response of a specified area by combining the brick thickness and thermal diffusivity parameters under the path, and establishes a thermal conduction response time set.

[0024] The feedback signal generation submodule compares each time point in the thermal conductivity response time set with the continuous spraying time set in the spraying task, filters out areas where the spraying time is greater than the response time, marks them as having the conditions for thermal conductivity feedback response, and forms a thermal conductivity feedback trigger signal group.

[0025] As a further aspect of the present invention, the thermal disturbance path construction module includes:

[0026] The disturbance source location submodule extracts all thermal response points that meet the thermal response conditions based on the thermal feedback trigger signal group, and confirms the relative position of each point in space by combining the coke oven structure layout diagram and temperature zone division diagram, thereby obtaining the disturbance initiation position set.

[0027] The path search operation submodule constructs a temperature zone connection graph based on the set of disturbance start positions, sets the heat flux reachability condition in the temperature zone connection graph as an edge weight constraint, and uses Dijkstra's shortest path algorithm to search for the shortest heat transfer path from the disturbance start point to other temperature zones, thereby obtaining the set of shortest disturbance paths.

[0028] The response segment extraction submodule extracts all temperature zone nodes with heat flux continuity requirements from the shortest disturbance path set in path order, excludes non-connected areas of the structure, and re-numbers and sorts the selected points to establish a thermal disturbance response segment set.

[0029] As a further aspect of the present invention, the spraying instruction generation module includes:

[0030] The target filtering submodule performs cross-matching with the spraying task list based on the set of thermal disturbance response sections to identify the intersection area that has thermal disturbance feedback and has been included in the spraying schedule, and obtains the spraying target section.

[0031] The regional sorting submodule extracts the connecting path of each segment according to the spatial distribution relationship of the spraying target segments, and rearranges the numbers of each spraying target according to the path order to form a spraying scheduling sequence table.

[0032] Based on the spraying schedule table, the parameter setting submodule assigns control parameters to the spraying tasks corresponding to each area, sets the spraying duration and spraying intensity value for each segment, and forms a spraying execution instruction.

[0033] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0034] In this invention, by collecting temperature data of the coke oven within a specified period and constructing a trend of heat flux changes, the spatial distribution of heat within the oven can be accurately determined. Based on differences in heat accumulation, abnormal heat load areas can be identified, thereby locking in target sections with potential moisture control value. Before the spraying action is executed, the heat conduction response time from the spraying point to the oven wall and the spatial extension structure of the thermal disturbance path are considered to comprehensively judge the heat conduction conditions and the direction of disturbance propagation. This allows for dynamic adjustment of the execution sequence and continuous parameters of the spraying task, ensuring the matching between the spraying behavior and the heat response. It also avoids errors in selecting spraying areas due to resource constraints. Before forming the spraying command, the spraying targets are screened and rearranged based on path connectivity and temperature zone coverage. This enables the moisture control behavior to not only respond to changes in thermal state in real time but also achieve targeted moisture release control in space. Ultimately, this improves the accuracy of moisture control and the ability to balance heat load during the coking process, enhancing the thermal stability of coke oven operation and the consistency of coke quality. Attached Figure Description

[0035] Figure 1 This is a system flowchart of the present invention;

[0036] Figure 2 This is a flowchart of the thermal storage evaluation module of the present invention;

[0037] Figure 3 This is a flowchart of the heat load area identification module of the present invention;

[0038] Figure 4 This is a flowchart of the thermal feedback triggering module of the present invention;

[0039] Figure 5This is a flowchart of the thermal disturbance path construction module of the present invention;

[0040] Figure 6 This is a flowchart of the spraying instruction generation module of the present invention. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0042] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0043] Please see Figure 1 The coking moisture control system based on atomized spraying includes:

[0044] The heat storage assessment module obtains the coke oven temperature for a specified period during the coking process and establishes a heat storage index sequence by calculating the cumulative change in heat flux of the coke oven bricks.

[0045] The heat load area identification module identifies abnormal heat load areas in the coke oven by referring to the heat storage index sequence, extracts abnormal load areas as spraying targets and constructs a spraying task list.

[0046] The thermal feedback triggering module determines the spray target switching logic based on the spraying task list, collects heat flux density data from the spraying point to the corresponding furnace wall, calculates the thermal conduction response time of the coke oven brick, and generates a corresponding thermal feedback triggering signal group.

[0047] The thermal disturbance path construction module constructs the thermal disturbance path within the coke oven temperature range through a set of thermal feedback trigger signals, and filters the set of thermal disturbance response segments along the direction of the thermal disturbance path;

[0048] The spraying instruction generation module corrects the spraying targets in the spraying task list based on the set of thermal disturbance response segments, and constructs spraying execution instructions for quantitative control of coal moisture release during the coking process by atomized spraying behavior.

[0049] The heat storage index sequence includes heat flux change distribution, coke oven temperature zone heat energy accumulation structure, and time-series temperature response characteristics. The spraying task list includes the distribution location of abnormal heat load areas, target area screening range, and spraying control cycle. The heat conduction feedback trigger signal group includes the effective response area identification mark, heat conduction path confirmation information, and spraying response critical conditions. The thermal disturbance response segment set includes the continuous area of ​​thermal disturbance path, the heat conduction connection area, and the feedback propagation direction structure. The spraying execution command includes the spraying target area, spraying duration, and spraying intensity set value.

[0050] Please see Figure 2 The thermal storage assessment module includes:

[0051] The temperature data acquisition submodule acquires coke oven temperature data for a specified period during the coking process, and collects continuous temperature sequences recorded by thermocouples set in the upper, middle and lower layers after the coal is fed into the furnace, generating a temperature time series dataset.

[0052] To acquire coke oven temperature data for a specified period during the coking process, the process first requires setting the coal feeding time as the zero point. Then, temperature data is collected through a multi-layered spatial distribution using thermocouples embedded in the upper, middle, and lower layers of the coke oven. Thermocouples should be positioned at different heights and depths within the coke oven wall to represent the overall temperature distribution characteristics. For example, a set of thermocouples can be placed at heights of 1 meter, 2 meters, and 3 meters to ensure coverage of typical heat transfer paths. Each thermocouple must continuously record temperature values ​​according to a set period (e.g., every 1 second) and store the data as raw time-series data. To ensure the stability and accuracy of temperature acquisition, high-temperature resistant thermocouples with specific characteristics are required. With excellent insulation performance, the thermocouple signal is amplified by an amplifier and transmitted to the acquisition module. At the same time, the time of each sampling point is recorded by the time synchronization module. During the acquisition process, each thermocouple will generate a temperature-time series, which will eventually form a multi-point temperature time series dataset. For example, if the acquisition time is 600 seconds and sampling is performed once per second, each measuring point will generate 600 sets of temperature values. The temperature data of each measuring point are summarized by number to form an overall sequence data file. This file serves as the basis for subsequent heat flux calculation. The temperature sampling values ​​are as follows: 1080℃ for the upper layer thermocouple, 980℃ for the middle layer, and 920℃ for the lower layer. The temperature of each layer changes continuously over time.

[0053] The heat flux integration submodule obtains the temperature difference between adjacent time points in the time-series temperature dataset. Combining the thermal conductivity of the measuring point with the corresponding area parameter, it calculates the total heat flux change of each temperature difference within a specified time period using the Simpson integral method to obtain the cumulative heat flux value.

[0054] To obtain the temperature difference between adjacent time points in a time-series temperature dataset, the temperature difference needs to be calculated for each sampling point. First, the time-series data collected by each thermocouple is processed, and the instantaneous temperature difference is obtained by subtracting the values ​​of two adjacent time points. For example, if the temperature is 1050℃ at second t and 1060℃ at second t+1, then ΔT = 10℃. Then, combining the thermal conductivity and thermally conductive area of ​​the measuring point, these parameters are input into the integration calculation module. The Simpson integral method is used to process the product of the temperature difference and the thermally conductive area over all consecutive time periods. Weighted integration is performed by selecting points according to the odd / even rule of the number of segmented intervals, specifically using the following formula:

[0055]

[0056] The parameters are explained as follows: Q: The cumulative heat flux of the measuring point within the interval obtained by integration, in joules (J), used to measure the change in heat over a specified time. Δt: The length of the integration time period, in seconds (s), obtained by the sampling period between every two sampling points. For example, if sampling is set to once per second, then Δt = 1. M: The coefficient in the denominator of the Simpson formula, which is constant at 3. Here, it is replaced with the letter M = 3 to simplify the expression and algebraic processing. k: Thermal conductivity, in W / m·K, representing the amount of heat transferred by a material under a unit temperature difference and unit length. It is obtained by looking up the material in a table. For example, for steel, k can be set to 1.5. A: The thermally active area of ​​the measuring point, in m². 2 ΔT1 represents the size of the heat transfer surface, obtained by measuring the area of ​​the coke oven wall in contact with the thermocouple, and is set to A = 0.002. ΔT1, ΔT2, ΔT3: represent the temperature difference values ​​at three consecutive time sampling points, in degrees Celsius (°C), obtained by subtracting from the temperature time series. For example: t1 = 1s, T = 1045°C; t2 = 2s, T = 1050°C; t3 = 3s, T = 1060°C, then ΔT1 = T2 - T1 = 5°C, ΔT2 = T3 - T2 = 10°C, ΔT3 = T4 - T3 = 8°C.

[0057] Based on the above data, substitute it into the formula to perform the calculation:

[0058]

[0059] This indicates that the accumulated heat flux at this measuring point within the 3-second interval is 0.053 J. The above calculation steps are executed in the data processing system every three consecutive time points until the entire data segment ends. If the total time is 600 seconds, approximately 199 sets of such integral values ​​will be generated. Finally, the sum of each integral value yields the overall accumulated heat flux value. For example, the accumulated heat flux calculated for this measuring point is Q. total =198×0.053=10.494J, this value is recorded in the dataset as the final heat flux index and used in combination with the temperature change rate index for processing.

[0060] The index sequence construction submodule extracts the temperature change rate of each measuring point, pairs the temperature change rate with the cumulative heat flux value according to the corresponding relationship of the measuring points, and completes the combination arrangement according to the spatial layout order of the coke oven to establish a heat storage index sequence.

[0061] To extract the temperature change rate at each measuring point, the temperature time series of each thermocouple must first be obtained and continuously recorded in the system in seconds. Then, the temperature difference between adjacent points is calculated point by point in chronological order and divided by the sampling period to obtain the rate of change per unit time. For example, if the temperature is 1120℃ at 360 seconds and 1115℃ at 359 seconds, with a sampling period of 1 second, the rate of change in that second is 5℃ / second. This process needs to continuously traverse the entire time series data of each measuring point to form a temperature rate sequence for each measuring point. Subsequently, the average rate value is calculated for the temperature rate sequence of each measuring point. For example, if the total change at a measuring point in 600 seconds is 2500℃, the average rate is 4.17℃ / second. Then, the average rate obtained from the previous stage is used to calculate the average rate for that measuring point. The accumulated heat flux value is extracted, for example, 13400 joules. These two values ​​are combined to form a two-dimensional data set, which represents the thermal behavior characteristics of the measuring point throughout the coking process. Then, all measuring points are arranged according to the spatial arrangement of the coke oven, generally numbered from top to bottom and from left to right. For example, the upper layer is numbered L1-L5, the middle layer is M1-M5, and the lower layer is B1-B5, forming 15 numbered points, corresponding to 15 pairs of temperature rate and heat flux values. For example, the upper layer measuring point L1 is 4.3℃ / s and 12700 joules, and L2 is 4.0℃ / s and 13000 joules, and so on. The resulting index sequence is a one-dimensional array arranged in numbered order. The system stores this as a heat storage index sequence in the database.

[0062] Please see Figure 3 The heat load area identification module includes:

[0063] The abnormal symptom extraction submodule extracts the heat storage value corresponding to each measuring point based on the heat storage index sequence, and obtains the local difference of each group of measuring points under the condition of regional continuity, thus constructing the heat storage change distribution information.

[0064] Based on the heat storage index sequence, the heat storage value corresponding to each measuring point is extracted. First, the combination information of each measuring point number and its corresponding heat storage value is read from the database or cache file. For example, the sequence A1 to A15 is read, and the corresponding heat storage values ​​are 12700, 13000, 12900, 14100, 14500, etc. Then, the spatial arrangement of the coke oven structure is used as the basis for regional division. For example, it can be divided into regional groups of five measuring points each, such as group 1 (A1 to A5), group 2 (A6 to A10), and group 3 (A11 to A15). Within each group, the heat storage values ​​of the measuring points are sorted and the difference between each pair of adjacent measuring points is calculated. For example, the difference between A1 and A2 is 300, between A2 and A3 it is -100, and between A3 and A4 it is 1200. This process is repeated sequentially to form the local difference sequence for the group. This process requires that each group of measuring points be spatially continuous. If a measuring point is missing for any reason, the area will not participate in this round of calculation. After obtaining the local difference sequence, the system performs statistical characteristic analysis on it, such as calculating the mean, range, and variance, and compares it with the difference sequence under the previous period or historical stable period to determine whether there is a region with significant fluctuations in heat accumulation under the current period. For example, if the average local difference of group 1 exceeds three times the historical stable value, or the range exceeds 5000, it is preliminarily determined that the region has abnormal signs. This process provides input information for subsequent outlier detection and target recognition, and finally generates a heat accumulation change distribution information set containing all region numbers and the distribution of local heat accumulation differences within them.

[0065] The critical section identification submodule uses the local outlier factor algorithm to calculate the relative density ratio between each measuring point and its adjacent measuring points in the heat storage change distribution, and delineates the areas that meet the outlier criteria as heat load abnormal sections, thus obtaining a set of abnormal areas.

[0066] The local outlier factor algorithm is used to analyze each measuring point in the thermal accumulation variation distribution. First, the system reads the thermal accumulation variation distribution information set generated by the previous module. This information set contains the spatial numbers of all measuring points and their corresponding local thermal accumulation differences. For example, taking points A1 to A9, their differences are 300, 250, 1200, 310, 290, 1150, 305, 280, and 1180, respectively. The system maps these one-dimensional scalar data to a one-dimensional thermal accumulation variation space, forming a point set (x... i ), where x i This represents the thermal accumulation difference at the measuring point numbered i. For example, the coordinates of point A1 are x1 = 300.

[0067] In this one-dimensional space, a neighborhood is constructed for each measurement point, and a neighborhood size parameter k is set, which means that when calculating the density of each measurement point, its k nearest neighbors are considered. For example, setting k=2 means that when the system calculates the density of any measurement point i, it uses the two nearest neighbors of that point for comparison. The symbol j∈N k(i) indicates that the measuring point j is one of the k nearest neighbors of the measuring point i, and the neighbor set N k (i) is the set of the k measuring points that are closest to measuring point i in Euclidean distance.

[0068] In this one-dimensional numerical space, the Euclidean distance simplifies to the absolute difference, defined as:

[0069] dist(i,j)=|x i -x j |;

[0070] The system first calculates the Euclidean distance from measurement point i to each of its neighbors j, and then introduces the reachability distance, denoted as ReachDist(i,j). This metric is used to reduce the impact of distance fluctuations between dense points on density assessment, and is defined as follows:

[0071] ReachDist(i,j)=max(dist(i,j),k-dist(j));

[0072] Where: ReachDist(i,j): the reachable distance from measurement point i to its neighbor j; dist(i,j): the Euclidean distance between measurement point i and its neighbor j; k-dist(j): the Euclidean distance between neighbor point j and its k-th nearest neighbor measurement point, representing the density boundary scale of the region where point j is located.

[0073] Next, the local reachability density LRD(i) of measurement point i is calculated, which represents the reciprocal of the average reachability density of the neighborhood surrounding measurement point i, as shown in the following formula:

[0074]

[0075] The meanings of each symbol are explained below: LRD(i): Local reachability density of measurement point i; k: Number of neighboring points; N k (i): The set of k nearest neighbors of the test point i; ReachDist(i,j): As above, it is the maximum distance from the test point i to its neighbor j.

[0076] Taking A2 as an example, its thermal storage difference is 250. The system calculates its neighbors as A1 (300) and A3 (1200), with corresponding distances of |250-300|=50 and |250-1200|=950, respectively. Let k-dist(A1)=100 for A1 and k-dist(A3)=1000 for A3, then:

[0077] ReachDist(A2,A1)=max(50,100)=100;

[0078] ReachDist(A2,A3)=max(950,1000)=1000.

[0079] The average value is (100+1000) / 2=550, therefore:

[0080]

[0081] Then calculate the LRD values ​​of the two neighboring points of A2 respectively: A1's neighbors are A2 and A3, ReachDist(A1,A2)=max(50,100)=100, ReachDist(A1,A3)=max(900,1000)=1000, with an average of 550, LRD(A1)=1 / 550=0.00182, A3's neighbors are A2 and A4 (310), ReachDist(A3,A2)=max(950,1000)=1000, ReachDist(A3,A4)=max(890,900)=900, with an average of 950, LRD(A3)=1 / 950=0.00105.

[0082] Finally, the Local Outlier Factor (LOF) of measurement point A2 is calculated to determine whether it belongs to an outlier region. The formula is as follows:

[0083]

[0084] Substitute into the calculation:

[0085]

[0086] A higher LOF value indicates a greater deviation of the measuring point from its neighborhood density. If the LOF of a point is greater than 1.6, the system classifies it as a measuring point with abnormal heat load. An LOF value greater than 1.6 is set as the threshold for outlier detection, based on experience gained from the application of the Local Outlier Factor method in engineering monitoring and anomaly detection practices. In one-dimensional or low-dimensional numerical spaces, LOF = 1 indicates that the local density of the measuring point is consistent with the mean of its neighborhood, which is normal. If LOF > 1, it indicates that the density of the measuring point is sparser than its neighborhood, showing an abnormal trend. In industrial scenarios, especially in objects with relatively stable thermophysical changes such as coke oven temperature and heat accumulation monitoring, density changes caused by normal fluctuations usually correspond to LOFs between 1.0 and 1.5, reflecting spatial differences in heat load distribution during normal operation. When LOF > 1.6, it indicates that the density difference between the measuring point and its neighborhood is greater than 60%, which is considered a structural anomaly or operational deviation exceeding the system's fluctuation tolerance range. Therefore, 1.6 is empirically set as a sensitive anomaly detection threshold that avoids excessive false alarms. This threshold can be dynamically adjusted based on actual operating data, or it can be referenced to the maximum LOF value of the system's historical normal cycle as a benchmark. For example, if the maximum LOF value of all measuring points under normal operating conditions is 1.4, then the threshold can be set to be 15% higher than that, i.e., 1.4 × 1.15 = 1.61, further confirming the rationality of setting 1.6. This value can also be cross-validated with manual inspection data or thermal imaging results. The system will group adjacent abnormal points into abnormal segments based on location continuity.

[0087] The spraying target generation submodule excludes areas with limited water supply in the current period based on the abnormal area set, and sets task numbers and period markers for the remaining areas to create a spraying task list;

[0088] Based on the identified set of abnormal areas, target screening and task construction are performed. First, the system imports water supply status data for the current period. This data is updated in real-time by the plant's water supply monitoring system, including water supply status markers for each spraying area. For example, areas B1, B3, and C2 have normal water supply, while areas A2, C1, and C4 are in a restricted state due to equipment maintenance or water source scheduling. Then, the system matches the set of abnormal areas with the water supply status data, excluding all area numbers in a "restricted" state. For example, if the abnormal area includes three segments: A2-A3, B1-B2, and C1-C2-C3, then A2... C1 will be removed from the spray candidate areas. The remaining areas will be merged and reconstructed into valid spray target areas such as A3, B1-B2, and C2-C3. The system will then generate a unique task number for each remaining area, such as TSK-001 and TSK-002, and set a period marker to identify which data acquisition period the task belongs to, such as the period number CY-056. Finally, the system will output a structured task entry, which includes: task number, area number, period number, continuous spraying indicator, priority level, etc., forming a spray task list, and will send it to the control layer through the system interface or API.

[0089] Please see Figure 4 The thermal feedback triggering module includes:

[0090] The spray point positioning submodule extracts the action position corresponding to each spray target in the spray task list, determines the coordinates of each spray point on the coke oven brick and the corresponding contact surface of the adjacent furnace wall, and obtains the spray-furnace wall mapping information.

[0091] The system extracts the target location for each spraying target from the spraying task list. This process first calls the task number and area coordinate information fields from the task list and matches them against the furnace geometry model data. For example, task TSK-003 points to area B2, which corresponds to spatial coordinates x = 6.2m, y = 2.4m, z = 1.5m in the coke oven structure model. The system uses these three-dimensional coordinates as the spraying point location. Then, based on the brick assembly relationship and wall number defined in the coke oven structure database, it retrieves the furnace wall number directly contacted by the spraying point. If spraying point B2 is located in the upper passage... The contact surface is furnace wall number W5. The system then calls the furnace wall grid division information to confirm the contact area range. For example, if the coverage radius of the spray nozzle is 0.6m, the corresponding wall grid covers numbers W5-11 to W5-13. The system binds the spatial coordinates of the spray point with the furnace wall contact surface area, which is the spray-furnace wall mapping information. The structured representation is: TSK-003 → coordinates (6.2, 2.4, 1.5) → contact wall W5 → coverage area W5-11 to W5-13. This process is executed in batches in all spray task entries to complete the mapping information of spray target coordinates and wall number.

[0092] The thermal response calculation submodule obtains the data on the change of heat flux density over time on the path from the spray contact point to the adjacent furnace wall based on the spray-furnace wall mapping information, and calculates the time required for the heat flux response of a specified area by combining the brick thickness and thermal diffusivity parameters under the path, and establishes a thermal conduction response time set.

[0093] Based on the spray-furnace wall mapping information, the heat flux density over time along the path from the spray contact point to the adjacent furnace wall is obtained. This process first extracts the corresponding heat flux density record sequence from the system temperature monitoring database based on the spray coordinates and corresponding contact wall numbers in the mapping information. For example, if the thermocouple number in wall W5 is W5-12, its heat flux density change curve over the past 10 minutes is obtained, with data intervals of 1 second, forming a sequence such as: [310, 312, 309, 308, ...] W / m³. 2 The system then reads the furnace wall structure thickness and material thermal diffusivity parameters along the spraying point direction. For example, if the path thickness is 0.35m, the thermal diffusivity is set to 1.4×10⁻⁶. -6 m 2 / s, and then calculate using the standard heat transfer and diffusion time estimation model, that is, the response time required for thermal disturbance to be conducted from the spray contact surface to the target wall under the material and thickness conditions. According to the thermal conductivity response time approximation model, the system traverses each time point in the heat flux density curve, estimates the corresponding time for heat to be transferred to the wall, and generates a time series. Each value in the series represents the conduction delay that thermal disturbance needs to experience at that moment. The system classifies the response time of all points into the thermal conductivity response time set and establishes a thermal response time table for each spray point.

[0094] The feedback signal generation submodule compares each time point in the heat conduction response time set with the continuous spraying time set in the spraying task, filters out areas where the spraying time is greater than the response time, marks them as having the conditions for heat conduction feedback response, and forms a heat conduction feedback trigger signal group.

[0095] The system compares each time point in the heat conduction response time set with the continuous spraying time set in the spraying task, filtering out areas where the spraying time is greater than the response time. The system first iterates through the set spraying duration field of each task in the spraying task list. For example, task TSK-003 is set to 120 seconds, TSK-004 to 75 seconds, and TSK-005 to 90 seconds. Then, these times are compared one by one with the response time values ​​of the corresponding spraying points in the previous module. For example, the heat response time of TSK-003 is 84 seconds, TSK-004 is 88 seconds, and TSK-005 is 95 seconds. The judgment logic is: if the spraying time value is greater than the response time value, it is marked as having heat conduction conditions; if it is less than or equal to, it is discarded. The system establishes a Boolean judgment matrix, such as TSK-003 being True, TSK-004 being False, and TSK-005 being False. Then, all task numbers marked as True are extracted to form a list, and a feedback response flag field value of 1 is added to them, while the rest are 0, thus constructing a feedback trigger signal group.

[0096] Please see Figure 5 The thermal disturbance path construction module includes:

[0097] The disturbance source location submodule extracts all thermal response points that meet the thermal response conditions based on the thermal feedback trigger signal group, and confirms the relative position of each point in space by combining the coke oven structure layout diagram and temperature zone division diagram, thus obtaining the set of disturbance start positions.

[0098] Based on the thermal conductivity feedback trigger signal group, all thermal response points that meet the thermal conductivity response conditions are extracted. First, the system traverses all spraying task records in the feedback signal group and filters out task numbers marked with a value of 1, i.e., points whose spraying time is confirmed to be greater than the response time. For example, the feedback values ​​of tasks TSK-003, TSK-006, and TSK-007 in the records are 1. The system extracts the spatial coordinates of the corresponding spraying points and the furnace wall numbers they are connected to, and summarizes them to form a set of response points. Then, the spatial location is confirmed by combining the coke oven structural layout diagram and the temperature zone division diagram. The system retrieves the three-dimensional layout... The model is used to map the coordinates of the response points onto the graph. This further confirms the furnace body number, temperature zone number, and the positional relationship of the point relative to the entire furnace. For example, TSK-003 corresponds to the 3rd row of the middle layer in section B of the furnace body, and the temperature zone number is ZB-M-03. The system marks its displacement in the coke oven volume space as 6.2m in the X direction, 2.4m in the Y direction, and 1.5m in the Z direction. This displacement is then bound to the block number in the temperature zone graph. All points that meet the conditions are mapped point by point, duplicate coordinates or overlapping numbers are removed, and the set of disturbance start positions is output.

[0099] The path search operation submodule constructs a temperature zone connection graph based on the set of disturbance initiation positions, sets the heat flux reachability condition in the temperature zone connection graph as an edge weight constraint, and uses Dijkstra's shortest path algorithm to search for the shortest heat transfer path from the disturbance initiation point to other temperature zones, thereby obtaining the set of shortest disturbance paths.

[0100] Based on the set of disturbance initiation locations, a temperature zone connection graph is constructed. First, the system imports the coke oven temperature zone division map and structural connection relationship graph. Each hot zone node is abstracted as a vertex in the graph structure. Edge connections are established between all adjacent nodes with heat transfer paths. Each edge represents a potential heat transfer path, and edge weights are assigned for path optimization search. The edge weights are designed as heat transfer delay time, i.e., the response time required for a thermal disturbance to propagate from one temperature zone to an adjacent temperature zone, as shown in the following formula:

[0101]

[0102] Where: w ij : Thermal disturbance response time from node i to j (in seconds); d ij α: Path length or equivalent heat diffusion path between the centers of two temperature zones (unit: meters); ij Thermal diffusivity of the material connecting two nodes (unit: m) 2 / s). The formula is derived from the thermal diffusion approximation model and is used to describe the time scale estimation of the thermal front passing through the material. It is dimensionally consistent, i.e., d 2 The result of / α is in seconds, representing the time it takes for the heat signal to travel.

[0103] Let the initial temperature region of the disturbance be ZB-M-03, and the three adjacent temperature regions be ZB-M-04, ZB-U-03, and ZB-L-03. The path parameters are as follows:

[0104] ZB-M-03 and ZB-M-04 are adjacent, with a path length d = 0.35 m and a thermal diffusivity α = 1.4 × 10⁻⁶. -6 m 2 / s, substitute into the calculation formula have to

[0105] ZB-M-03 and ZB-U-03 are adjacent, with a path length d = 0.28 m and a thermal diffusivity α = 1.6 × 10⁻⁶. -6 m 2 / s, calculated

[0106] ZB-M-03 and ZB-L-03 are adjacent, with a path length d = 0.40 m and a thermal diffusivity α = 1.5 × 10⁻⁶. -6 m 2 / s, calculated

[0107] The system writes the aforementioned edge weights into the connectivity graph. Each node maintains the path weights of its adjacent nodes and initializes the path record matrix and shortest path list. Then, starting from the perturbation initiation point ZB-M-03, it executes Dijkstra's shortest path algorithm. In each iteration, the system updates the cumulative shortest thermal response time from the initiation point to each target point and records the node numbers traversed by the path. Finally, it outputs the shortest path result, such as: -ZB-M-03→ZB-U-03 (path weight: 49000s, path sequence: [ZB-M-03, ZB-U-03]).

[0108] -ZB-M-03→ZB-M-04→ZB-U-04 (Path weight: 87500s+65000s=152500s).

[0109] After all calculations are completed, the system summarizes the shortest thermal disturbance paths from all starting points to the remaining temperature zone nodes, and generates a shortest disturbance path set table, which includes the starting point number, path node sequence, total path response time, and path segment weight details.

[0110] The response segment extraction submodule extracts all temperature zone nodes with heat flux continuity requirements from the shortest disturbance path set in path order, excludes non-connected areas of the structure, and re-numbers and sorts the selected points to establish a set of thermal disturbance response segments.

[0111] The system extracts all temperature zone nodes that meet the heat flux continuity requirement from the shortest perturbation path set in path order. First, it traverses all path node sequences in the shortest perturbation path set, evaluating whether the heat flux change between adjacent temperature zone nodes on each path meets the continuity requirement. Specifically, the system calls the heat flux time series data corresponding to each node in the temperature database. For example, subsequent nodes in the ZB-M-03 path are ZB-U-03, ZB-M-04, and ZB-U-04. The system calculates the gradient of heat flux change between each pair of connected nodes (e.g., ZB-M-03 and ZB-U-03), setting a threshold condition that the rate of change of heat flux per unit time between two points does not exceed 200 W / m². 2If a segment exceeds this upper limit ( / s), it is considered a heat flux interruption region, and the extraction of subsequent nodes on that path is terminated. Taking the path ZB-M-03→ZB-U-03→ZB-U-04 as an example, if the gradient between ZB-M-03 and ZB-U-03 is 180, and between ZB-U-03 and ZB-U-04 it is 220, then the system extracts the first two segments of nodes, discarding the latter and retaining only the first segment. After scanning all paths, the system merges the temperature zone nodes that meet the continuity condition in each path and removes duplicate nodes. Then, it calls the coke oven structure diagram to verify the connectivity of the node spatial positions, excluding structurally non-directly connected area nodes. For example, if ZB-M-03 and ZC-M-03 have continuous heat flux but there is a partition wall or fault area in between, then ZC-M-03 will be discarded. Finally, the system renumbers and sorts the nodes according to their vertical, horizontal, and vertical sequence in the spatial structure to form a standard set of thermal disturbance response segments.

[0112] Please see Figure 6 The spraying instruction generation module includes:

[0113] The target selection submodule cross-matches the set of thermal disturbance response sections with the spraying task list to identify the intersection area that has thermal disturbance feedback and is included in the spraying schedule, and obtains the spraying target section.

[0114] Based on the set of thermal disturbance response zones, the system performs cross-matching with the spraying task list. First, the system extracts the node numbers of each region in the response zone set and generates an index table. For example, the response zone set includes numbers such as ZB-M-03, ZB-U-03, and ZB-L-03. At the same time, the system extracts the target region number field corresponding to all tasks in the spraying task list. For example, the target for TSK-003 is ZB-M-03, for TSK-005 it is ZB-U-03, and for TSK-007 it is ZC-M-04. The system performs cross-searching on the two sets of numbers and retains only the common region number that appears in both lists and its corresponding task record. This confirms the task regions that have thermal disturbance feedback response and are also in the spraying schedule. The filtering process uses a hash matching mechanism to improve the comparison efficiency and forms a set of spraying target zones.

[0115] The regional sorting submodule extracts the connecting paths of each segment based on the spatial distribution of the spraying target segments, and rearranges the numbers of each spraying target according to the path order to form a spraying scheduling sequence table.

[0116] Based on the spatial distribution of the target spraying sections, the system extracts the connecting paths of each section. The system reads the three-dimensional coordinate information of each target section node in the structural model and determines the physical connection relationship between each node. For example, ZB-M-03 is located at coordinates (6.2, 2.4, 1.5), and ZB-U-03 is located at (6.2, 2.4, 2.2). Since they are vertically adjacent in the Z-axis direction, the system determines that they are connected. The system then constructs a path starting from ZB-M-03, and sorts the nodes according to the priority of the coordinate axis as the main line direction, such as ascending order of X, then Y, and then Z. The final path order is ZB-M-03 → ZB-U-03. The system assigns a sequence number to each node on the path and associates it with the corresponding task number, such as TSK-003 as sequence number 1 and TSK-005 as sequence number 2. All sorting information generates a spraying scheduling sequence table, which includes fields such as task number, spatial coordinates, connecting path sequence, and execution priority identifier.

[0117] The parameter setting submodule assigns control parameters to the spraying tasks corresponding to each area based on the spraying schedule order table, sets the spraying duration and spraying intensity value of each segment, and forms a spraying execution command.

[0118] Based on the spraying schedule, control parameters are assigned to the spraying tasks corresponding to each area. The system first reads each task number and its area number from the schedule and calls historical response data of the area, such as response time and cumulative heat flux, to calculate parameters. For example, the historical heat flux of area ZB-M-03 is 13000J and the response time is 84s. The system sets the spraying duration to be 10% higher than the response time, i.e., 92 seconds. At the same time, the spraying intensity is set to the ratio of heat flux to time, i.e., 13000J / 92s≈141.3W. In addition, combined with the physical properties of the nozzle and the pump pressure curve, the intensity value is converted into the pump pressure setpoint and the spraying rate value to complete the parameter table construction. For example, TSK-003 is set to: spraying time 92s, intensity 141.3W, corresponding pump pressure 1.6MPa, and spraying rate 0.5L / s. Finally, all task parameters form a spraying execution instruction list and are output to the control system interface.

[0119] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A mist-based spray coking moisture control system, characterized in that, The system includes: The heat storage assessment module obtains the coke oven temperature for a specified period during the coking process and establishes a heat storage index sequence by calculating the cumulative change in heat flux of the coke oven bricks. The heat load area identification module identifies abnormal heat load areas in the coke oven by referring to the heat storage index sequence, extracts abnormal load areas as spraying targets and constructs a spraying task list. The thermal feedback triggering module determines the spray target switching logic based on the spraying task list, collects heat flux density data from the spraying point to the corresponding furnace wall, calculates the thermal conduction response time of the coke oven brick, and generates a corresponding thermal feedback triggering signal group. The thermal disturbance path construction module constructs the thermal disturbance path within the coke oven temperature range through the thermal conduction feedback trigger signal group, and filters the set of thermal disturbance response segments along the direction of the thermal disturbance path; The spraying instruction generation module corrects the spraying targets in the spraying task list based on the set of thermal disturbance response segments and constructs spraying execution instructions for quantitative control of coal moisture release during the coking process by atomized spraying behavior. The thermal storage assessment module includes: The temperature data acquisition submodule acquires coke oven temperature data for a specified period during the coking process, and collects continuous temperature sequences recorded by thermocouples set in the upper, middle and lower layers after the coal is fed into the furnace, generating a temperature time series dataset. The heat flux integration submodule obtains the temperature difference between adjacent time points in the time-series temperature dataset. Combining the thermal conductivity of the measuring point with the corresponding area parameter, it calculates the total heat flux change of each temperature difference within a specified time period using the Simpson integral method to obtain the cumulative heat flux value. For the cumulative heat flux, the formula is used: ； in, This is the cumulative heat flux at that measuring point within that interval, obtained by integration. The length of the integration time period. The coefficients in the denominator of Simpson's formula are... Thermal conductivity, For the heat-affected area of ​​the measuring point, This represents the temperature difference between three consecutive time sampling points; The index sequence construction submodule extracts the temperature change rate of each measuring point, pairs the temperature change rate with the cumulative heat flux value according to the corresponding relationship of the measuring points, and completes the combination arrangement according to the spatial arrangement order of the coke oven to establish a heat storage index sequence. The heat load area identification module includes: The abnormal symptom extraction submodule extracts the heat storage value corresponding to each measuring point based on the heat storage index sequence, and obtains the local difference of each group of measuring points under the condition of regional continuity, and constructs the heat storage change distribution information. The critical section identification submodule uses the local outlier factor algorithm to calculate the relative density ratio between each measuring point and its adjacent measuring points in the heat storage change distribution, and delineates the areas that meet the outlier criteria as heat load abnormal sections, thus obtaining a set of abnormal areas. The spraying target generation submodule excludes areas with limited water supply in the current period based on the abnormal area set, and sets task numbers and period markers for the remaining areas to create a spraying task list; The thermal feedback triggering module includes: The spray point positioning submodule extracts the action position corresponding to each spray target in the spray task list, determines the coordinates of each spray point on the coke oven brick and the corresponding contact surface of the adjacent furnace wall, and obtains the spray-furnace wall mapping information. The thermal response calculation submodule obtains the data on the change of heat flux density over time on the path from the spray contact point to the adjacent furnace wall based on the spray-furnace wall mapping information, and calculates the time required for the heat flux response of a specified area by combining the brick thickness and thermal diffusivity parameters under the path, and establishes a thermal conduction response time set. The feedback signal generation submodule compares each time point in the thermal conductivity response time set with the continuous spraying time set in the spraying task, filters out areas where the spraying time is greater than the response time, marks them as having the conditions for thermal conductivity feedback response, and forms a thermal conductivity feedback trigger signal group.

2. The atomized spray based coking moisture control system of claim 1, wherein, The heat storage index sequence includes heat flux change distribution, coke oven temperature zone heat energy accumulation structure, and time-series temperature response characteristics. The spraying task list includes the distribution location of abnormal heat load areas, target area screening range, and spraying control cycle. The heat conduction feedback trigger signal group includes response effective area identification markers, heat conduction path confirmation information, and spraying response critical conditions. The thermal disturbance response segment set includes continuous thermal disturbance path areas, heat conduction connection areas, and feedback propagation direction structure. The spraying execution command includes the spraying target area, spraying duration, and spraying intensity set value.

3. The atomized spray based coking moisture control system of claim 1, wherein, The thermal disturbance path construction module includes: The disturbance source location submodule extracts all thermal response points that meet the thermal response conditions based on the thermal feedback trigger signal group, and confirms the relative position of each point in space by combining the coke oven structure layout diagram and temperature zone division diagram, thereby obtaining the disturbance initiation position set. The path search operation submodule constructs a temperature zone connection graph based on the set of disturbance start positions, sets the heat flux reachability condition in the temperature zone connection graph as an edge weight constraint, and uses Dijkstra's shortest path algorithm to search for the shortest heat transfer path from the disturbance start point to other temperature zones, thereby obtaining the set of shortest disturbance paths. The response segment extraction submodule extracts all temperature zone nodes with heat flux continuity requirements from the shortest disturbance path set in path order, excludes non-connected areas of the structure, and re-numbers and sorts the selected points to establish a thermal disturbance response segment set.

4. The atomized spray based moisture control system for coke making as claimed in claim 3 wherein, The spraying instruction generation module includes: The target filtering submodule performs cross-matching with the spraying task list based on the set of thermal disturbance response sections to identify the intersection area that has thermal disturbance feedback and has been included in the spraying schedule, and obtains the spraying target section. The regional sorting submodule extracts the connecting path of each segment according to the spatial distribution relationship of the spraying target segments, and rearranges the numbers of each spraying target according to the path order to form a spraying scheduling sequence table. Based on the spraying schedule table, the parameter setting submodule assigns control parameters to the spraying tasks corresponding to each area, sets the spraying duration and spraying intensity value for each segment, and forms a spraying execution instruction.

Images