Automatic proportioning control system for catalyst mixed materials

Abstract

The invention relates to the technical field of catalyst processing, in particular to an automatic proportioning control system for catalyst mixed materials. According to the invention, the physical distance of each material conveying pipeline and the opening and closing delay of the control valve are obtained, and the task issuing time point is combined, so that the effective response time intervals corresponding to different pipelines are established, and the judgment of the validity fed back by the weighing sensor is more accurate; furthermore, through dynamic trend fitting between historical matching data and current adjacent pipeline data, it is ensured that accurate substitution weighing information can still be obtained under the condition of abnormal feedback, and the matching precision deviation caused by the abnormal feedback is effectively avoided; the actual emptying state of the residual materials in the pipeline is jointly judged by combining the tail section flow velocity and the tail end storage capacity change, scheduling resource configuration has real-time effectiveness, the pipeline priority is calculated through multiple factors such as the conveying distance, the flow velocity fluctuation and the historical idle period, and the reasonability and accuracy of a scheduling path are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of catalyst processing, in particular to an automatic proportioning control system for catalyst mixed materials. BACKGROUND

[0002] The technical field of catalyst processing involves key links such as design, preparation, modification and application of catalysts, and its core matters include selection of catalyst raw materials, proportioning of active components, preparation of carriers, regulation of physical and chemical properties, and optimization of stability and reaction selectivity of catalysts in industrial processes.

[0003] Among them, the automatic proportioning control system for traditional catalyst mixed materials refers to a system for automatically mixing and proportioning catalyst raw materials of multiple different components according to a set proportion.

[0004] The automatic proportioning control system for traditional catalyst mixed materials only controls the proportioning of components according to the set proportion, lacks dynamic sensing ability for the response time of the actual conveying process, and when there is a delay in the opening and closing of the control valve or there is a significant difference in the length of the conveying path, it is easy to cause the time offset of the weighing feedback, and the system will misrecognize abnormal feedback as valid data without effective judgment mechanism, resulting in a decrease in the accuracy of material proportioning. In addition, the system lacks a substitute value generation mechanism when the weighing data is abnormal, and cannot perform trend fitting correction, which easily causes the interruption or misfeeding of the entire proportioning process. At the same time, the judgment standard for whether the residual material in the pipeline is completely emptied is not clear, and it only relies on the feeding completion signal to determine its state, which may cause old material mixing or incomplete feeding, especially in the multi-pipeline alternating operation scene, which is more likely to accumulate errors, and ultimately leads to systematic deviation of the overall proportioning result. SUMMARY

[0005] The purpose of the present application is to solve the problems existing in the prior art and to provide an automatic proportioning control system for catalyst mixed materials.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: an automatic proportioning control system for catalyst mixed materials comprises: A pipeline response analysis module: obtains the length of each pipeline conveying catalyst mixed materials and the opening and closing delay time of the corresponding electric control valve, calculates the shortest response time and the maximum allowed lag time difference of each pipeline, compares with the weighing sensor feedback time, and screens abnormal feedback nodes; A substitute weighing value calculation module: obtains the weighing trend of the pipeline where the abnormal feedback node is located in the last batch of catalyst mixed material proportioning process and the real-time weighing trend of the adjacent pipeline, and calculates the substitute weighing value by fitting the two trends; Empty state judgment module: obtain the instantaneous flow rate of the abnormal feedback node pipeline tail section and the capacity change data of the end storage section, judge the emptying state of the residual catalyst material in the pipeline, and mark the schedulable pipeline; Scheduling sequence adjustment module: taking the replacement weighing value as the comparison basis, calculating the scheduling priority order of the schedulable pipeline according to the target feeding quality of each component raw material in the current material proportioning task, and obtaining a pipeline scheduling queue; Activation instruction generation module: generating a path activation instruction according to the pipeline scheduling queue and the feeding target of the catalyst mixed material proportioning task.

[0007] As a further scheme of the application, the abnormal feedback node includes a feedback time abnormality label, a pipeline number to which the node belongs, and a node weighing value, the replacement weighing value is a value calculated through trend fitting and adjacent pipeline data, the schedulable pipeline includes a pipeline state label, an end flow rate determination result, and a storage capacity change determination result, the pipeline scheduling queue includes a path sorting position, a target raw material corresponding relationship, and a priority reference index, and the path activation instruction includes a control valve opening and closing instruction, a feeding start instruction, and a feeding amount adjustment instruction.

[0008] As a further scheme of the application, the pipeline response analysis module includes: Distance detection submodule: obtaining the conveying distance of each raw material pipeline, detecting the opening and closing response delay time of the connected electric control valve, and calculating the shortest conveying response time of each pipeline under the task condition according to the issue time of the current catalyst mixed material proportioning task as the reference starting point; Response time difference construction submodule: obtaining the response time sequence of each pipeline in the current task, calculating the time difference between the shortest conveying response time and the maximum response lag time, constructing the weighing feedback effective time range of each pipeline under the current task, and generating a feedback time effective interval; Feedback screening submodule: obtaining the feedback time of each raw material weighing sensor, calculating the relative position of the feedback time and the feedback time effective interval corresponding to the pipeline, judging whether there is a case earlier than the shortest response time or later than the maximum lag time difference, and screening abnormal feedback nodes.

[0009] As a further scheme of the application, the replacement weighing value calculation module includes: Number identification submodule: extracting the raw material pipeline number corresponding to the abnormal feedback node, calling the task identification in the last batch of catalyst mixed material proportioning task, determining the adjacent raw material pipeline numbers arranged in parallel, and establishing a target and adjacent pipeline number mapping table; Trend feature extraction submodule: based on the target and adjacent pipeline number mapping table, the weighing trend curve of the line where the abnormal node is located in the last batch of catalyst mixture material proportioning process and the real-time weighing trend curve in the current period are obtained, and a multi-source weighing trend data set is generated; Fitting calculation submodule: according to the multi-source weighing trend data set, the weighing trend of the abnormal feedback node in the last batch and the real-time weighing trend of the adjacent pipeline in the current period are taken as joint input variables, which are input into the Bayesian dynamic regression model for curve fitting, and the alternative weighing value of the abnormal feedback node in the current period is calculated.

[0010] As a further scheme of the present application, the emptying state judgment module comprises: Flow rate acquisition submodule: the pipeline number corresponding to the raw material of the abnormal feedback node is obtained, the instantaneous flow rate data of the tail section of each abnormal pipeline in the last three periods is collected, it is judged whether the flow rate in each period is lower than the minimum instantaneous flow rate threshold, and a flow rate period identification result is generated; Capacity change extraction submodule: the capacity change of the raw material pipeline end storage section in the current period is obtained, the change value is compared and judged to determine whether the capacity fluctuation is in a stable range, and a capacity change determination result is generated; State determination submodule: screening schedulable pipelines that meet the flow rate period identification result and the capacity change determination result at the same time.

[0011] As a further scheme of the present application, the scheduling sequence adjustment module comprises: Target parameter extraction submodule: the alternative weighing value of the abnormal feedback node corresponding to the schedulable pipeline is extracted, the target feeding mass of the component raw material in the current catalyst mixture material proportioning task is called, the corresponding relationship between the alternative weighing value of each schedulable pipeline and the target feeding mass is established, and a target feeding proportion mapping table is obtained; Sorting factor calculation submodule: based on the target feeding proportion mapping table, the conveying distance, flow rate fluctuation standard deviation and three idle period numbers of each schedulable pipeline are extracted, and a scheduling sorting parameter set is constructed; Priority queue generation submodule: according to the scheduling sorting parameter set, the relative priority relationship between pipelines is formed by aggregation processing according to the sorting parameters, and a pipeline scheduling queue is obtained.

[0012] As a further scheme of the present application, the activation instruction generation module comprises: Queue reading submodule: the first path combination content of the pipeline scheduling queue is extracted, a path execution preparation sequence is established, and a scheduling path reading result is obtained; The target matching sub-module obtains the target feeding quality of each component raw material in the current catalyst mixture material proportioning task, matches the raw material type in the scheduling path reading result with the corresponding target value, and generates a path target matching data set; The instruction construction sub-module constructs the path activation instruction for the proportioning device to execute according to the path target matching data set.

[0013] Compared with the prior art, the advantages and positive effects of the present application are that: In the present application, by obtaining the physical distance of each material conveying pipeline and the opening and closing delay of the control valve, and combining the task issuing time point, the effective response time interval corresponding to different pipelines is established, so that the effectiveness judgment of the feedback of the weighing sensor is more accurate. Further, through dynamic trend fitting between historical proportioning data and current adjacent pipeline data, accurate replacement weighing information can still be obtained in the case of feedback abnormality, effectively avoiding the proportioning accuracy deviation caused by feedback abnormality. The actual emptying state of the residual material in the pipeline is judged by combining the tail section flow rate and the change of the end storage capacity, so that the scheduling resource configuration has real-time effectiveness. The mapping relationship constructed by the target feeding quality and the current replacement value is used to calculate the pipeline priority by considering multiple factors such as conveying distance, flow rate fluctuation and historical idle period, so as to improve the rationality and accuracy of the scheduling path. Finally, the control instruction is generated according to the matching result to ensure the coordination and continuity of the overall operation of the system. The system is improved in terms of multi-node cooperation, feedback abnormality compensation, feeding accuracy and intelligent path selection. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 The system flowchart of the present application is shown in the figure; Figure 2 The flowchart of the pipeline response analysis module of the present application is shown in the figure; Figure 3 The flowchart of the replacement weighing value calculation module of the present application is shown in the figure; Figure 4 The flowchart of the emptying state judgment module of the present application is shown in the figure; Figure 5 The flowchart of the scheduling sequence adjustment module of the present application is shown in the figure; Figure 6 The flowchart of the activation instruction generation module of the present application is shown in the figure. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0016] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0017] Please refer to Figure 1 An automatic proportioning control system for catalyst mixture materials comprises: A pipeline response analysis module: obtains the length of each pipeline conveying catalyst mixture materials and the opening and closing delay time of the corresponding electric control valve, calculates the shortest response time and the maximum allowable lag time difference of each pipeline, compares with the raw material weighing sensor feedback time, and screens abnormal feedback nodes; A substitute weighing value calculation module: obtains the weighing trend of the pipeline where the abnormal feedback node is located in the last batch of catalyst mixture material proportioning process and the real-time weighing trend of the adjacent pipeline, and calculates the substitute weighing value by fitting the two trends; An empty state judgment module: obtains the instantaneous flow rate of the tail section of the pipeline where the abnormal feedback node is located and the capacity change data of the end storage section, judges the empty state of the residual catalyst material in the pipeline, and marks the schedulable pipeline; A scheduling sequence adjustment module: takes the substitute weighing value as the basis for comparison, calculates the scheduling priority order of the schedulable pipeline according to the target feeding quality of each component raw material in the current material proportioning task, and obtains the pipeline scheduling queue; An activation instruction generation module: generates path activation instructions according to the pipeline scheduling queue and the feeding target of the catalyst mixture material proportioning task; The abnormal feedback node includes a feedback time abnormality mark, a node belonging pipeline number, and a node weighing value. The substitute weighing value is a value calculated by trend fitting and adjacent pipeline data. The schedulable pipeline includes a pipeline state label, an end flow rate determination result, and a storage capacity change determination result. The pipeline scheduling queue includes a path sorting position, a target raw material corresponding relationship, and a priority reference index. The path activation instruction includes a control valve opening and closing instruction, a feeding start instruction, and a feeding amount adjustment instruction.

[0018] Please refer to Figure 2 The pipeline response analysis module comprises: Distance detection sub-module: obtain the conveying distance of each raw material pipeline, detect the opening and closing response delay time of the connected electric control valve, and calculate the shortest conveying response time of each pipeline under the task condition according to the task issuing time of the current catalyst mixture material proportioning task as the reference starting point; The conveying distance of each raw material pipeline is obtained, first, the spatial topology structure database of the raw material pipeline is established, the physical length of each pipeline is obtained by using the pipeline layout diagram or industrial surveying means, for example, the starting point of pipeline A is the raw material warehouse, the ending point is the reaction kettle interface, the length is 120 meters measured by on-site laser ranging or three-dimensional modeling tool SolidWorks, the length of pipeline B is 75 meters, then the inner diameter and material of each pipeline are registered, A is a stainless steel pipe with an inner diameter of 50 mm, B is a PVC pipe with an inner diameter of 40 mm, the recommended flow rate of different materials under standard pressure is obtained by referring to the process manual, the recommended flow rate of stainless steel pipe is 1.5 m / s under 0.3 MPa, the recommended flow rate of PVC pipe is 1.2 m / s under the same pressure, based on the formula , wherein, : the flow time of the material in the pipeline, in seconds (s), : the pipeline length, in meters (m), : the flow rate, in meters per second (m / s). The flow time of A pipeline is , and the flow time of B pipeline is , then the action response log of the control valve in the PLC system is called, the valve response delay is obtained by the time difference between the control signal issuing time and the initial flow rate detected by the flowmeter, for example, the delay of electric valve A is 2.4 seconds, and the total response time of A is , wherein, : the total response time of pipeline A, in seconds (s).

[0019] All delay data are calculated by the following unified formula: ; , wherein, : the total conveying response time of a certain pipeline, in seconds (s), : the flow time of the material, in seconds (s), : the electric valve response delay time, in seconds (s).

[0020] Then, the task instruction issuing time is taken as the reference time point , for example, if the task is issued at 10:05:20, the shortest feedback of A pipeline should appear at ; wherein, : the shortest feedback appearance time of pipeline A under the task, : the task issuing reference time, : The delivery response time of the pipeline, in seconds (s). The corresponding The minimum response time of each pipeline is calculated in sequence, providing the original basis for the next response time difference processing.

[0021] Response time difference construction submodule: Obtain the response time sequence of each pipeline in the current task, calculate the shortest delivery response time and the maximum response lag time difference, construct the effective time range of the weighing feedback of each pipeline under the current task, and generate the feedback time effective interval; Obtain the response time sequence of each pipeline in the current task, and the system calls the continuous feedback time point sequence from the task start time from the weighing sensor feedback database, for example, the feedback time points of pipeline C recorded in a task are 10:06:43, 10:06:51, and 10:07:00, which correspond to the time sequence The system calculates the shortest response time of each pipeline and the maximum time value in the response time sequence , such as the C pipeline in the task , , the lag time difference is calculated , where : The maximum response lag time difference, in seconds (s), : The maximum response time of the pipeline, in seconds (s), : The shortest response time of the pipeline, in seconds (s). According to this, the effective time interval of the weighing feedback is constructed , for example, in , the feedback effective interval is , the system records this interval in the pipeline task parameter configuration table, which is used for boundary judgment by the subsequent feedback screening submodule, and all feedback records exceeding this range will be marked, and then the feedback time window mapping relationship table under the task of each pipeline is established, and the task response analysis database is formed by structuring the feedback records.

[0022] Feedback screening submodule: Obtain the feedback time of each raw material weighing sensor, calculate the relative position of the feedback time and the corresponding feedback time effective interval of the pipeline, judge whether there is a situation earlier than the shortest response time or later than the maximum lag time difference, and screen abnormal feedback nodes; Obtain the feedback time of each raw material weighing sensor, extract the record of pipeline D feeding the weighing signal at 10:06:30 in a certain proportioning task by reading the timestamp field in the industrial weighing controller output data frame, and the system calls the effective feedback time interval of the pipeline in the task , assuming that the interval is , compare the feedback time with the upper and lower bounds of the interval, if the feedback time If the feedback time is earlier than the expected time, it is marked as "early feedback exception", if If the feedback time is later than the expected time, it is marked as "delayed feedback exception", for example, if the D pipeline feedback time 10:06:30 is earlier than 10:06:42.4, it is classified as early exception feedback, and another feedback time is 10:07:10, which belongs to delayed feedback exception, the system will map the exception feedback nodes according to the task and pipeline number The offset is calculated by , wherein, : offset time difference, unit: seconds (s), : actual weighing feedback time point, : minimum response reference time of feedback, : maximum response reference time of feedback, recorded in the exception feedback log, all exception records will participate in subsequent exception frequency analysis, feedback efficiency calculation and other statistical processes, forming an exception feedback summary table under each task.

[0023] Please refer to Figure 3 , the alternative weighing value calculation module includes: Number identification submodule: extract the raw material pipeline number corresponding to the exception feedback node, call the task identification in the last batch of catalyst mixed material proportioning task, determine the adjacent raw material pipeline number of parallel arrangement, and establish a mapping table of target and adjacent pipeline number; Extract the raw material pipeline number corresponding to the exception feedback node, first parse its binding feedback channel ID from the exception feedback log, and then establish a mapping relationship through the control system parameter configuration table to obtain its actual physical pipeline number, for example, if the feedback node binds the channel ID CH-04, the corresponding physical number is Pipe-A4, the system retrieves Pipe-A4 in the last batch of proportioning task in the historical task record, and obtains its task ID as MB2025061607 through the proportioning task scheduling record, all pipeline information configured under the same reaction unit in the task are also retrieved, and through the task proportioning structure table analysis, it is identified that the adjacent pipelines Pipe-A2 and Pipe-A5 connected with Pipe-A4 in the same module have synchronous operation conditions, and the system establishes a one-to-many pipeline number mapping table based on this, taking Pipe-A4 as the target number, Pipe-A2 and Pipe-A5 as its adjacent numbers, forming a complete record item and writing it into the number mapping cache area.

[0024] Trend feature extraction submodule: based on the target and adjacent pipeline number mapping table, obtain the weighing trend curve of the line where the exception node is located in the last batch of catalyst mixed material proportioning process, and the real-time weighing trend curve in the current period, generate a multi-source weighing trend data set; Based on the target and adjacent pipeline number mapping table, the system performs data grabbing operations on the three pipelines of Pipe-A4, Pipe-A2 and Pipe-A5. First, the historical data file stored in the weighing data server is called to read the weighing trend curve of Pipe-A4 in the task period of MB2025061607. This curve records the weighing change trend every second from the start to the end of the task. The system organizes these data in chronological order and segments them by percentile nodes, marking them as the reference curve of the last batch. At the same time, the real-time weighing data of Pipe-A2 and Pipe-A5 in the current running period is extracted from the real-time weighing cache database. The same time slice length and segmentation method is used to reconstruct the trend lines. After the weighing data is collected, it is aligned in time index mode and divided by second-level granularity. The three trend curves are then stored in the multi-source trend data set object, with Pipe-A4 data marked as the reference target source and Pipe-A2 and Pipe-A5 data marked as adjacent observation sources.

[0025] Fitting calculation submodule: according to the multi-source weighing trend data set, the weighing trend of the pipeline where the abnormal feedback node is located in the last batch and the real-time weighing trend of its adjacent pipelines in the current period are taken as joint input variables and input into the Bayesian dynamic regression model for curve fitting to calculate the alternative weighing value of the abnormal feedback node in the current period; According to the multi-source weighing trend data set, the weighing trend of the pipeline where the abnormal feedback node is located in the last batch and the real-time weighing trend of its adjacent pipelines in the current period are taken as joint input variables. For ease of description, let Pipe-A4 be the first pipeline (i.e. the pipeline where the abnormal feedback node is located), Pipe-A2 be the second pipeline (adjacent pipeline 1), and Pipe-A5 be the third pipeline (adjacent pipeline 2). First, the system takes the first pipeline as the target, and its complete weighing trend in the last batch task MB2025061607 is a sequence of weighing values recorded at equally spaced time points, collected once every second, with a sampling length of 300 seconds. The system standardizes this to form an input sequence , then obtains the real-time weighing trend data of the second and third pipelines in the current period, with a sampling frequency of once every second. After synchronization calibration and interpolation processing, we get and , which are combined into a time series input matrix of length 300 and dimension 3 . The system imports this multi-source input variable into the Bayesian dynamic linear regression model. The model constructs a prediction function with the trend of the first pipeline in the last period as the prior structure, and the regression equation is as follows: ; where, : time point The alternative weighing value of the first pipeline, in kilograms (kg). The standardized weighing value of the first pipeline in the previous cycle. The standardized weighing value of the second pipeline in the current cycle. The standardized weighing value of the third pipeline in the current cycle. : Dynamic bias term, : Dynamic regression weights of input variables.

[0026] The regression weights are set based on the following criteria: : Represents the bias term, whose value is derived from the average residual of the previous cycle on the first pipeline. It is used to compensate for systematic errors in linear fitting. If the predicted value has been consistently low in the historical trend, the bias term increases, and vice versa. This value is dynamically adjusted over time during model fitting to ensure the stability of the predicted residual.

[0027] This indicates the degree of influence of the historical trend of the first pipeline on the prediction result. It is set based on the goodness of fit between the trend and the current substitution value. The higher the goodness of fit, the greater its reference value and the greater its weight. The system judges this by its correlation coefficient with the target fitting trend.

[0028] This indicates the degree to which the weighing trend of the second pipeline is relevant to the current forecast value. It is set based on the frequency of coordinated changes between the second pipeline and the historical trend of the first pipeline. The more synchronized the two fluctuate, the stronger the trend similarity and the higher the weight.

[0029] :and Similarly, the stronger the coordination between the reference weights of the third pipeline, the greater the regression weights assigned by the system.

[0030] Suppose at a certain moment seconds, the standardized input variables are as follows: (History of the first pipeline) (Second pipeline currently) (Currently, the third pipeline). The corresponding weights are as follows: , , , .

[0031] Substitute into the regression formula: ; To restore the standardized forecast values ​​to actual values, let's assume the mean of the first pipeline in the previous cycle was 12.5 kg and the standard deviation was 3.1 kg: ; in: : The first pipeline replaces the weighing value at 120 seconds, in kilograms (kg).

[0032] The system will repeat this regression prediction process throughout the time series, eventually outputting a complete 300-point replacement weighing curve for the first pipeline.

[0033] Please refer to Figure 4 , the empty state judgment module includes: The flow rate collection submodule: obtains the raw material pipeline number corresponding to the abnormal feedback node, collects the instantaneous flow rate data of the tail section of each abnormal pipeline in the continuous three periods, judges whether the flow rate in each period is below the minimum instantaneous flow rate threshold, and generates a flow rate period identification result; After obtaining the raw material pipeline number corresponding to the abnormal feedback node, the system automatically calls the pipeline number marked as having a replacement value calculation requirement in the aforementioned abnormal feedback identification result, and locates to the flow rate sensor channel deployed at the tail section of each pipeline according to the number. The system traces back from the current task to the past in time dimension, extracts the instantaneous flow rate record data of the tail section of the pipeline in the continuous three proportioning periods, and constructs a time series of three periods in the form of sampling once per second. Then the system loads the minimum instantaneous flow rate threshold set in the task, for example, set to 0.18 m / s, and performs interval minimum value retrieval on the flow rate sequence of each period. If the flow rate of all time points in a period does not exceed the threshold, the system marks the period as a "low flow rate period". In this way, the judgment and statistical identification of each period are completed one by one, and finally the flow rate period identification result for each abnormal pipeline is generated, which contains the flow rate state judgment label of each period and whether the pipeline has a continuous low flow rate history.

[0034] The capacity change extraction submodule: obtains the capacity change of the raw material pipeline end storage section in the current period, compares the change value, determines whether the capacity fluctuation is within the stable range, and generates a capacity change determination result; Obtain the storage section number connected to the end of each abnormal feedback pipeline. The system retrieves the total capacity change data from the task start time to the end time period in the current period through the material storage tank capacity sensor data channel. The system calculates the capacity change value based on the difference between the storage values at the start and end time nodes, and compares the value with the stable fluctuation range set by the device, for example, the system sets the allowed fluctuation range to ±0.8 liters. If the capacity increase or decrease amplitude of the storage section in the current period falls within this range, it is recorded as "fluctuation stable". If it exceeds the range, it is recorded as "fluctuation abnormal". In addition, the system will also make a segmented trend judgment on the change curve. If the capacity is always rising or falling in the period, the system will mark the trend direction and one-way abnormal characteristics, and generate a capacity change determination result, which includes stability label, trend direction label and whether it is a continuous change state.

[0035] State determination submodule: screening schedulable pipelines that meet both flow rate cycle identification results and capacity change determination results; Cross-screening pipeline results that have completed flow rate cycle identification and capacity change determination in pipelines corresponding to all abnormal feedback nodes. The system uses "simultaneous satisfaction" as the logical condition to screen pipelines that have continuous low flow rate records in the last three cycles and whose current cycle capacity change is within the set stable fluctuation interval as candidate targets. The system establishes a state matrix, with each pipeline as a row and two determination results as column items. Only when both column items are "compliant" can the pipeline be marked as "schedulable", and its number will be recorded in the scheduling candidate list for subsequent scheduling module to call preferentially. If there is a pipeline that only meets one of the items, it will be marked as "partially schedulable" for manual review; if neither item is met, it will be considered "not schedulable" and automatically shielded from the execution instruction of the current round of matching task.

[0036] Please refer to Figure 5 , the scheduling sequence adjustment module includes: Target parameter extraction submodule: extracting the replacement weighing value of the abnormal feedback node corresponding to the schedulable pipeline, calling the target feeding mass of the component raw material in the current catalyst mixture material matching task, establishing the correspondence between the replacement weighing value of each schedulable pipeline and the target feeding mass, and obtaining the target feeding matching mapping table; Extracting all schedulable pipelines screened, the system first extracts the replacement weighing final value of each pipeline in the current task cycle from the previous replacement weighing trend data according to the identification of the abnormal feedback node bound to it, and then the system calls the component raw material matching requirement issued in the current catalyst mixture material matching task. This task information includes the target feeding mass of each raw material, for example, setting the target feeding value of the first pipeline corresponding raw material in the current task to 14.0 kg, and the second pipeline target value to 10.5 kg. The system establishes a mapping according to the raw material type and pipeline number to form a directional relationship between the replacement weighing value of each pipeline and its corresponding task feeding target value, and constructs a complete target feeding matching mapping table. This table records the corresponding items between the fitting weighing result of each schedulable pipeline in the current task and the should-be-feeding mass.

[0037] Sorting factor calculation submodule: based on the target feeding matching mapping table, extracting the conveying distance, flow rate fluctuation standard deviation, and number of three-round idle cycles of each schedulable pipeline to construct a scheduling sorting parameter set; After establishing the target feed ratio mapping table, three key parameters are extracted for each schedulable pipeline: conveying distance, standard deviation of flow velocity fluctuation, and number of idle cycles over three rounds. The system uses these three parameters as the sorting criteria to form a scheduling ranking parameter set, constructs a ranking factor scoring function, and performs quantitative calculations accordingly. Each schedulable pipeline will obtain its corresponding scheduling factor score value through the following calculation formula: ; in, : No. The ranking score of each pipeline : No. The transport distance of the pipeline, in meters (m). : The maximum transport distance (m) among the pipelines currently participating in the sorting. : No. The standard deviation (m / s) of the flow velocity fluctuation in the pipeline over the past three cycles. The maximum standard deviation of the flow velocity (m / s) in the pipeline involved in the sequencing. : No. The number of times the pipeline has been idle in the last three cycles. The maximum number of idle cycles across all pipelines involved in the sorting process. The system-defined weighting coefficients are used to adjust the importance of the three factors; the system settings are as follows: , , . This setting controls the degree of influence of conveying distance on the sequencing process. The system is based on the principle that longer conveying distances result in longer material response times in the pipeline and greater overall scheduling resource consumption; therefore, its weight is set to the highest, with a default value of 0.4. This is used to control the degree of influence of the standard deviation of flow velocity fluctuations. The system setting is based on the following: the greater the flow velocity fluctuation, the lower the stability of the pipeline. In order to ensure the continuity of scheduling, pipelines with smaller fluctuations are given priority. Therefore, the system gives this item a medium to high weight, with a default value of 0.35. This parameter controls the role of the number of idle cycles in the sorting process. The setting is based on the principle that more idle cycles indicate a lower workload for the pipeline, and the system tends to prioritize scheduling pipelines with higher usage frequency to balance resources. Therefore, fewer idle cycles result in higher priority. This parameter has the lowest weight and is set to 0.25.

[0038] There are three schedulable pipelines, numbered as pipeline 1, pipeline 2, and pipeline 3, with the following characteristic parameters: Delivery distance: The first pipeline is 120 meters, the second pipeline is 100 meters, and the third pipeline is 80 meters. Therefore... ; Flow rate standard deviation: 0.12 for the first pipeline, 0.15 for the second pipeline, and 0.09 for the third pipeline, so ; Idle cycle number: 1 for the first pipeline, 2 for the second pipeline, and 3 for the third pipeline, so .

[0039] Substitute the above formula to calculate the ranking factor score of the three pipelines respectively: First pipeline: ; Second pipeline: ; Third pipeline: .

[0040] The calculation result shows that the first pipeline has the highest score of 0.8467, the second pipeline has the second highest score of 0.7666, and the third pipeline has the lowest score of 0.4767. Subsequently, the scheduling priority relationship will be established according to the score value and the scheduling queue will be generated.

[0041] Priority queue generation submodule: according to the scheduling sorting parameter set, aggregate processing is performed according to the sorting parameters to form the relative priority relationship between pipelines and obtain the pipeline scheduling queue; After completing the calculation of the scheduling sorting parameter set, the system takes the sorting score value of each schedulable pipeline as the basis, calls the sorting function module to aggregate process all pipelines, first sorts the pipelines from high to low according to the score value, places the pipeline with the highest score at the head of the scheduling list, and arranges the remaining pipelines in order. If the score values are the same, the system determines the priority according to the standby rules, such as preferentially scheduling pipelines with shorter delivery distance, smaller flow rate fluctuation, or lower frequency of participation in the last two cycles, to form the final scheduling priority relationship. The system encodes this sorting structure as a linear list, numbers the pipelines in the scheduling order, and marks it as the pipeline scheduling priority queue under the current task, and pushes this queue to the control scheduling unit for actual feeding execution call. The system records this sorting structure to the scheduling log.

[0042] Please refer to Figure 6 , the activation instruction generation module includes: Queue reading submodule: extracts the path combination content at the head of the pipeline scheduling queue, establishes a path execution preparation sequence, and obtains the scheduling path reading result; According to the pipeline scheduling priority queue, the pipeline number corresponding to the head of the queue and its scheduling path configuration item are read, and the system loads the complete path structure from the scheduling path database according to the information such as the raw material conveying path, valve control node, target tank number associated with the pipeline, calls the control section information of the pipeline starting point, middle section diversion, tail section confluence and the like in stages, and combines to form a complete path execution preparation sequence, which includes all physical control point state sets required for path activation, such as the number of electric valves to be opened, the sequence of check valves to be unlocked, the identification of pipe sections to be preheated, etc. The system encodes the combination result as a scheduling path reading result.

[0043] Target matching sub-module: obtain the target feeding quality of each component raw material in the current catalyst mixture material proportioning task, match the raw material type in the scheduling path reading result with the corresponding target value, and generate a path target matching data set; After the system completes the path reading, the proportioning parameters under the current catalyst mixture material proportioning task are called, the target feeding quality of each component raw material is extracted from the task setting database, and the system compares the raw material type identifier in the scheduling path reading result with the raw material name in the proportioning task setting in pieces, such as identifying the raw material C1 in the scheduling path, the system searches and reads the corresponding target value of C1 in the proportioning table, for example, 15.0 kg, and records the matching relationship into the path target matching data set, which is composed of three items: “path section number-raw material name-target quality”.

[0044] Instruction construction sub-module: constructing path activation instructions for the execution of the proportioning device according to the path target matching data set; After obtaining the path target matching data set, the system constructs a path activation instruction set for driving the proportioning device to execute according to the path structure and target quality parameters contained in each matching record. Each instruction contains the control node number to be activated, the target feeding amount setting value, the duration of continuous activation or feedback trigger condition, etc. The system judges whether to adopt timing control or feedback trigger control mode according to the control strategy, such as selecting feedback trigger mode, embedding weighing feedback identification condition at the same time, for example, stopping when the cumulative feeding value reaches 14.9 kg. After the instruction construction is completed, it is packaged as a scheduling control package, and the instruction sequence is pushed to the proportioning control system for standby execution, and all instruction contents are recorded into the control log backup at the same time.

[0045] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. An automatic proportioning control system for a catalyst mixture, characterized in that, The system includes: Pipeline response analysis module: Obtain the length of each pipeline transporting catalyst mixture and the opening and closing delay time of the corresponding electric control valve, calculate the difference between the shortest response time and the maximum allowable lag time for each pipeline, compare it with the feedback time of the raw material weighing sensor, and screen abnormal feedback nodes. Alternative weighing value calculation module: Obtain the weighing trend of the pipeline where the abnormal feedback node is located in the previous batch of catalyst mixture proportioning process and the real-time weighing trend of adjacent pipelines, and calculate the alternative weighing value by fitting the trends of the two. Emptying status judgment module: acquires the instantaneous flow rate and capacity change data of the end storage section of the pipeline tail section of the abnormal feedback node, judges the emptying status of the residual catalyst material in the pipeline, and marks the schedulable pipeline. Scheduling order adjustment module: Using the alternative weighing value as the comparison basis, and based on the target feeding mass of each component raw material in the current material proportioning task, calculates the scheduling priority order of the schedulable pipelines to obtain the pipeline scheduling queue; Activation command generation module: Generates path activation commands based on the feeding target of the pipeline scheduling queue and the catalyst mixture proportioning task.

2. The automatic proportioning control system for catalyst mixtures according to claim 1, characterized in that, The abnormal feedback node includes an abnormal feedback time marker, the pipeline number to which the node belongs, and the node weighing value. The alternative weighing value is specifically a value calculated through trend fitting and adjacent pipeline data. The schedulable pipeline includes a pipeline status label, end flow rate determination result, and storage capacity change determination result. The pipeline scheduling queue includes path sorting position, target raw material correspondence, and priority reference index. The path activation command includes a control valve opening / closing command, a feeding start command, and a feeding amount adjustment command.

3. The automatic proportioning control system for catalyst mixtures according to claim 1, characterized in that, The pipeline response analysis module includes: Distance detection submodule: acquires the conveying distance of each raw material pipeline, detects the opening and closing response delay time of the connected electric control valve, and calculates the shortest conveying response time of each pipeline under the task conditions based on the issuance time of the current catalyst mixture ratio task as the baseline. Response time difference construction submodule: Obtain the response time sequence of each pipeline in the current task, calculate the difference between the shortest delivery response time and the maximum response lag time, construct the effective weighing feedback time range of each pipeline in the current task, and generate the effective feedback time interval; Feedback filtering submodule: Obtain the feedback time of each raw material weighing sensor, calculate the relative position of the feedback time with the effective interval of the feedback time corresponding to the pipeline, determine whether there is a situation that is earlier than the shortest response time or later than the maximum lag time difference, and filter abnormal feedback nodes.

4. The automatic proportioning control system for catalyst mixtures according to claim 3, characterized in that, The alternative weighing value calculation module includes: Numbering identification submodule: Extract the raw material pipeline number corresponding to the abnormal feedback node, call the task identifier corresponding to the previous batch of catalyst mixture proportioning task, determine the adjacent raw material pipeline numbers deployed in parallel, and establish a target-adjacent pipeline number mapping table; Trend feature extraction submodule: Based on the target and adjacent pipeline number mapping table, obtain the weighing trend curve of the line where the abnormal node is located in the previous batch of catalyst mixture proportioning process, as well as the real-time weighing trend curve in the current cycle, and generate a multi-source weighing trend dataset. Fitting Calculation Submodule: Based on the multi-source weighing trend dataset, the weighing trend of the line where the abnormal feedback node is located in the previous batch and the real-time weighing trend of the adjacent pipeline in the current cycle are used as joint input variables and input into the Bayesian dynamic regression model for curve fitting to calculate the alternative weighing value of the abnormal feedback node in the current cycle.

5. The automatic proportioning control system for catalyst mixtures according to claim 4, characterized in that, The clear status determination module includes: Flow rate acquisition submodule: Obtain the raw material pipeline number corresponding to the abnormal feedback node, collect the instantaneous flow rate data of the tail section of each abnormal pipeline in three consecutive cycles, determine whether the flow rate in each cycle is lower than the minimum instantaneous flow rate threshold, and generate flow rate cycle identification results. Capacity Change Extraction Submodule: Obtains the capacity change of the storage section at the end of the corresponding raw material pipeline in the current cycle, compares and judges the change values, determines whether the capacity fluctuation is within a stable range, and generates a capacity change judgment result. Status determination submodule: Filters schedulable pipelines that simultaneously satisfy the flow rate cycle identification result and the capacity change determination result.

6. The automatic proportioning control system for catalyst mixtures according to claim 5, characterized in that, The scheduling order adjustment module includes: Target parameter extraction submodule: Extracts the alternative weighing value of the abnormal feedback node corresponding to the schedulable pipeline, calls the target feed mass of the component raw materials in the current catalyst mixture proportioning task, establishes the correspondence between the alternative weighing value of each schedulable pipeline and the target feed mass, and obtains the target feed proportioning mapping table; The sorting factor calculation submodule extracts the conveying distance, flow rate fluctuation standard deviation, and number of three idle cycles for each schedulable pipeline based on the target feed ratio mapping table, and constructs a set of scheduling sorting parameters. Priority queue generation submodule: Based on the scheduling sorting parameter set, it performs aggregation processing according to the sorting parameters to form a relative priority relationship between pipelines and obtain the pipeline scheduling queue.

7. The automatic proportioning control system for catalyst mixtures according to claim 6, characterized in that, The activation instruction generation module includes: Queue Reading Submodule: Extracts the first and second path combination content of the pipeline scheduling queue, establishes the path execution preparation sequence, and obtains the scheduling path reading result; Target matching submodule: Obtains the target feed mass of each component raw material in the current catalyst mixture proportioning task, matches the raw material type and corresponding target value in the scheduling path reading result scheduling path, and generates a path target matching dataset; Instruction Construction Submodule: Based on the path target matching dataset, constructs path activation instructions for execution by the batching device.

Images