A limestone calcination intelligent control method and system based on the Internet of Things

Abstract

The application discloses a limestone calcination intelligent control method and system based on the Internet of Things and relates to the field of limestone calcination. The method comprises the following steps: collecting limestone calcination data and associated quality data through the Internet of Things to generate standardized data; identifying a temperature normal working condition segment and generating a temperature constraint identifier; extracting heat response, air-oxygen matching and material residence records to form a calcination process record; inputting an improved TTMs model to generate a branch calcination record; determining under-burning and over-burning deviation based on the temperature constraint and the branch record, and generating a deviation record; identifying implicit under-burning risks, local over-burning risks and risk source identifiers; generating a control parameter correction amount and issuing the same to an execution device; and based on feedback data, calibrating risks and correction amounts to generate next-period control parameters to realize closed-loop control. The application realizes multi-source collaborative control and closed-loop optimization adjustment of a calcination process, improves calcination stability and can be applied to an industrial kiln intelligent control scene.

Description

Technical Field

[0001] This invention relates to the field of limestone calcination technology, and in particular to an intelligent control method and system for limestone calcination based on the Internet of Things. Background Technology

[0002] Existing limestone calcination control technologies mainly rely on distributed sensing devices to collect data on kiln temperature, kiln pressure, air volume, and material conveying status. These data are then combined with empirical thresholds or fixed control strategies to regulate fuel supply and combustion air volume. Some existing solutions introduce data-driven models to analyze operational data for estimating calcination status and adjusting control parameters. However, these models typically use single time-series features or local operating condition data as input sources, and they lack sufficient modeling of the coupling relationship between the material residence process, heat response process, and air-oxygen matching process within the kiln.

[0003] In practical applications, the calcination process is affected by multiple factors, including batch-to-batch material differences, kiln temperature fluctuations, and disturbances in air and material supply, resulting in distinct phased and delayed characteristics in the calcination state. Existing technologies for analyzing the calcination process often fail to distinguish between normal and abnormal temperature fluctuation conditions, lack structured extraction of stable operating segments, and struggle to form a process description suitable for precise control. Furthermore, the relationships between heat response, air-oxygen matching, and material residence are often treated as independent variables, lacking a unified path-based expression method, making it difficult to identify deviations from the multi-dimensional process coordination perspective.

[0004] Existing control strategies, after deviation identification, mostly use fixed rules or a single model to output control corrections, failing to combine historical execution feedback to perform closed-loop calibration of control parameters and actual execution deviations. This leads to the accumulation of deviations between control commands and execution results, which in turn affects the stability and consistency of subsequent control cycles. These problems limit the ability of limestone calcination processes to achieve refined control and stable operation under complex conditions.

[0005] Therefore, how to provide an intelligent control method and system for limestone calcination based on the Internet of Things is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose an intelligent control method and system for limestone calcination based on the Internet of Things. This invention standardizes calcination operation data and finished product quality data, constructs normal temperature operating condition segments and branch calcination processing paths, and achieves closed-loop control by combining deviation direction identification and feedback calibration. It has the advantages of clear process modeling, accurate deviation identification, and adaptive correction of control parameters.

[0007] An intelligent control method for limestone calcination based on the Internet of Things according to an embodiment of the present invention includes the following steps:

[0008] The limestone calcination operation data and finished product quality feedback data are collected by the Internet of Things, pre-processed and associated by material batch to generate standardized calcination status data.

[0009] Based on standardized calcination state data, determine the operating conditions of the calcination section temperature, and generate normal temperature operating condition segments and normal temperature constraint identifiers.

[0010] Extract heat response records, air-oxygen matching records, and material residence records from normal temperature operating conditions, and generate calcination process records according to record categories;

[0011] The calcination process records are input into the improved TTMs model, and heat response processing path, air-oxygen matching processing path, and material residence processing path are established according to the record category to generate branch calcination records;

[0012] The improved TTMs model compares the consistency of changes between the branch calcination records based on the normal temperature constraint marker, determines the under-burning deviation direction and the over-burning deviation direction, and generates under-burning deviation records and over-burning deviation records.

[0013] Identify hidden under-burning risks, localized over-burning risks, and risk source markers based on under-burning deviation records and over-burning deviation records;

[0014] Based on the hidden underburning risk, local overburning risk, and risk source identification, control parameter correction values ​​are generated and sent to the calcination execution equipment.

[0015] Collect calcination execution feedback data, calibrate risk source identification and control parameter correction based on the calcination execution feedback data, and generate limestone calcination control parameters for the next control cycle.

[0016] Optionally, the process of preprocessing separately and then associating by material batch includes:

[0017] The limestone calcination operation data is parsed, timestamps are unified, equipment numbers are matched, duplicates are removed, anomalies are corrected, and missing data is filled to obtain calcination operation preprocessing data.

[0018] The finished product quality feedback data is analyzed for indicators, formatted, results standardized, batch matched and time aligned to obtain preprocessed quality feedback data.

[0019] Based on the material batch information, by establishing a unified batch identifier, the calcination operation pretreatment data and the quality feedback pretreatment data are matched and associated, summarized and organized, and standardized calcination status data is generated.

[0020] Optionally, the process of generating the normal temperature operating condition segment and the normal temperature constraint identifier includes:

[0021] Extract the calcination temperature records of the same batch of materials, sort them according to the time of collection to form a temperature sequence, compare the temperature sequence with the preset target temperature range, and mark the low temperature, normal and high temperature operating condition records.

[0022] Continuous normal temperature records are bound together, and records shorter than a preset threshold are removed to form normal temperature operating condition segments.

[0023] Write the material batch, start and end times, and temperature matching results to the temperature normal operating condition chip, and generate a temperature normal constraint identifier.

[0024] Optionally, the process of generating the calcination process record includes:

[0025] The standardized calcination state data within the normal temperature operating condition segment are aligned according to material batch and acquisition time, and the changes in operating state at adjacent time points are extracted.

[0026] Based on changes in fuel supply, the thermal response relationship is established by correlating changes in calcination section temperature and flue gas temperature within the lag interval.

[0027] Based on changes in combustion air volume, and by correlating changes in kiln pressure and flue gas oxygen content, an air-oxygen matching relationship is established.

[0028] Based on the feeding and discharging status, the material residence area is divided and the material residence relationship is established by combining the finished product quality feedback;

[0029] The three relationships are classified, identified, and merged to generate a calcination process record.

[0030] Optionally, the improved TTMs model includes a record category encoding unit, a three-path input construction unit, a TTMs time-series hybrid backbone, and a branched calcination record output unit;

[0031] Optionally, the process of generating the branch calcination record includes:

[0032] The timing of the fuel supply change in the heat response relationship is taken as the starting point of the path, the lag interval is taken as the path interval, and the temperature changes of the calcination section and the flue gas temperature within the lag interval are taken as path response items. The heat response processing path is generated in the order of the path starting points.

[0033] The moment when the combustion air volume changes in the air-oxygen matching relationship occurs is taken as the starting point of the path, the air-oxygen matching observation interval is taken as the path interval, and the changes in kiln pressure and flue gas oxygen content within the air-oxygen matching observation interval are taken as path response items. The air-oxygen matching processing path is generated in the order of the path starting points.

[0034] The material dwell time reference time in the material dwell relationship is used as the starting point of the path, the discharge reference time is used as the ending point of the path, the material dwell interval and the finished product quality feedback status are used as the path result items, and the material dwell processing path is generated according to the order of the path starting points.

[0035] The heat response processing path, the air-oxygen matching processing path, and the material residence processing path are grouped together according to the same material batch and the same normal temperature constraint identifier to generate branch calcination records.

[0036] Optionally, the improved TTMs model includes a temperature normal constraint control unit and a deviation direction output unit, and the process of generating underheating deviation records and overheating deviation records includes:

[0037] Determine the heat response path record and the wind-oxygen matching path record that have a time correspondence with the material residence interval;

[0038] Compare the fuel supply changes in the heat response path record with the path response items to mark the insufficient heat response state;

[0039] Compare the changes in combustion air volume in the wind-oxygen matching path record with the path response items to mark abnormal wind-oxygen matching status.

[0040] Compare the material dwell area and finished product quality feedback status with the preset material dwell range and preset quality target, and mark the material dwell status and quality feedback status;

[0041] When insufficient heat response, insufficient material residence and quality deviation exist simultaneously in the same material residence range, determine the direction of underburning deviation and generate an underburning deviation record;

[0042] When the same material residence area simultaneously exhibits abnormal air-oxygen matching, excessively long material residence time, and quality deviation, the direction of overburning deviation is determined, and an overburning deviation record is generated.

[0043] The improved TTMs model uses the TinyTime Mixers model as its basic time-series mixing backbone. The TinyTime Mixers model is a lightweight pre-trained model for multivariate time-series processing, used to perform time segmentation, time dimension mixing, and variable channel mixing on continuously acquired multivariate time-series data. The improved TTMs model includes a record category encoding unit, a three-path input construction unit, a TTMs time-series mixing backbone, a branched calcination record output unit, a temperature normal constraint control unit, and an off-direction output unit.

[0044] The record category coding unit is used to identify the relationship category identifier in the calcination process record; the three-path input construction unit is used to convert the heat response relationship, air-oxygen matching relationship, and material residence relationship into three types of input sequences; the TTMs time-series mixing backbone is used to perform time-series mixing processing on the three types of input sequences; the branch calcination record output unit is used to generate heat response processing paths, air-oxygen matching processing paths, and material residence processing paths; the temperature normal constraint comparison unit is used to compare the time correspondence and change consistency of the three types of processing paths under the temperature normal constraint identifier; the deviation direction output unit is used to output the under-burning deviation direction and the over-burning deviation direction, and generate under-burning deviation records and over-burning deviation records.

[0045] Optionally, the process of identifying latent underburning risk, localized overburning risk, and risk source identification includes:

[0046] A continuity check is performed on the underburning deviation records, and underburning deviation records that are collected continuously and correspond to the same material residence interval are merged into hidden underburning risks.

[0047] Based on the insufficient heat response state and insufficient material retention state corresponding to the hidden underburning risk, determine the source identifier of insufficient heat response and the source identifier of insufficient material retention;

[0048] The overheating deviation records are checked for continuity. Overheating deviation records that are collected continuously and correspond to the same material residence range are merged into local overheating risk.

[0049] Based on the abnormal state of air-oxygen matching and the state of excessive material residence time corresponding to the risk of local overburning, determine the source identifier of abnormal air-oxygen matching and the source identifier of excessive material residence time;

[0050] When both latent underburning risk and local overburning risk are generated within the same material residence area, the latent underburning risk, local overburning risk and corresponding risk source identifiers are retained respectively.

[0051] Optionally, the process of generating the control parameter correction amount includes:

[0052] Determine the control cycle to be corrected corresponding to the material residence range;

[0053] The changes in fuel supply, heat response, material residence time, combustion air volume, kiln pressure, and flue gas oxygen content are read from the control cycle to be corrected.

[0054] When there is an indication of insufficient heat response, the fuel supply correction amount is generated based on the difference between the preset heat response amplitude and the actual heat response amplitude.

[0055] When there are sources indicating insufficient or excessive material dwell time, the feed rate correction amount or discharge rate correction amount is generated based on the difference between the preset material dwell range and the actual dwell length.

[0056] When there is an abnormal source of air-oxygen matching, the combustion air volume correction and kiln pressure correction are generated based on the kiln pressure and flue gas oxygen content deviation.

[0057] Each control correction is limited and merged to generate control parameter correction values.

[0058] Optionally, the process of calibrating the risk source identification and control parameter correction includes:

[0059] By comparing the control parameter correction with the actual execution value, the equipment execution deviation is obtained, the control parameter correction is calibrated, and the execution calibration correction is obtained.

[0060] The heat response, kiln pressure change, and finished product quality status are compared with the corresponding preset targets to generate calcination state convergence results.

[0061] When the convergence result of the calcination state meets the preset convergence condition, the corresponding risk source identifier is retained, and the calibration correction amount is processed according to the first amplitude coefficient.

[0062] When the convergence result of the calcination state does not meet the preset convergence conditions, the corresponding risk source identifier is retained, and the calibration correction amount is performed according to the second amplitude coefficient.

[0063] When the convergence result of the calcination state does not meet the preset consistency condition with the risk source identifier, the risk source identifier is corrected, and the calibration correction amount is re-determined based on the corrected risk source identifier.

[0064] The current limestone calcination control parameters, the processed execution calibration correction, and the corrected risk source identifier are combined to generate the limestone calcination control parameters for the next control cycle.

[0065] An intelligent control system for limestone calcination based on the Internet of Things according to an embodiment of the present invention includes the following modules:

[0066] The data acquisition and standardization module is used to collect calcination operation data and finished product quality data, preprocess and associate them by batch to generate standardized calcination status data;

[0067] The temperature condition determination module is used to determine the temperature condition of the calcination section and generate normal temperature segments and constraint indicators.

[0068] The calcination process record generation module is used to extract heat response, air-oxygen matching and material residence records to generate calcination process records;

[0069] The branch calcination record generation module is used to input the calcination process record into the improved TTMs model, establish processing paths according to categories, and generate branch calcination records.

[0070] The deviation direction determination module is used to compare the branch records with the temperature constraint identifier to determine the deviation direction of under-burning and over-burning, and generate corresponding records.

[0071] The risk identification module is used to identify hidden under-burning risks, local over-burning risks, and the sources of these risks.

[0072] The control correction module is used to generate control parameter correction values ​​and send them to the execution device;

[0073] The feedback calibration module is used to calibrate the sources of risk and the correction amount of control parameters based on the execution feedback data, and generate the control parameters for the next cycle.

[0074] The beneficial effects of this invention are:

[0075] (1) By standardizing the calcination operation data and finished product quality feedback data, and introducing normal temperature operating condition segments and constraint labels, the structured extraction of the stable calcination range is realized, so that the calcination process is transformed from discrete data expression into a continuous and comparable process record, thereby improving the integrity and analyzability of process information.

[0076] (2) By constructing heat response processing path, wind-oxygen matching processing path and material residence processing path, and realizing the consistency comparison and deviation direction identification in the improved TTMs model, the underburning and overburning states can be determined based on the multi-process coupling relationship, thereby improving the accuracy and stability of abnormal state identification.

[0077] (3) By introducing a closed-loop feedback calibration mechanism for risk source identification and control parameter correction, the execution feedback data and control correction process are corrected accordingly, so that the control parameters can be dynamically adjusted with the execution deviation, reducing the accumulation of control error and improving the stability and consistency of the long-term operation of the calcination process. Attached Figure Description

[0078] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0079] Figure 1 This is a flowchart of an intelligent control method for limestone calcination based on the Internet of Things proposed in this invention;

[0080] Figure 2 This is a module connection diagram of an intelligent control system for limestone calcination based on the Internet of Things proposed in this invention;

[0081] Figure 3 This diagram illustrates the construction of a branched calcination record using an improved TTMs model for an IoT-based intelligent control method for limestone calcination proposed in this invention.

[0082] Figure 4 (a) is a schematic diagram of the physical structure of the calcining kiln and the Internet of Things acquisition and control in Embodiment 2 of the intelligent control method for limestone calcination based on the Internet of Things proposed in this invention;

[0083] Figure 4 (b) is a schematic diagram of the intelligent control logic closed loop of Embodiment 2 of the intelligent control method for limestone calcination based on the Internet of Things proposed in this invention. Detailed Implementation

[0084] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0085] Example 1: Reference Figures 1-3 A smart control method for limestone calcination based on the Internet of Things includes the following steps:

[0086] The limestone calcination operation data and finished product quality feedback data are collected by the Internet of Things, pre-processed and associated by material batch to generate standardized calcination status data.

[0087] Based on standardized calcination state data, determine the operating conditions of the calcination section temperature, and generate normal temperature operating condition segments and normal temperature constraint identifiers.

[0088] Extract heat response records, air-oxygen matching records, and material residence records from normal temperature operating conditions, and generate calcination process records according to record categories;

[0089] The calcination process records are input into the improved TTMs model, and heat response processing path, air-oxygen matching processing path, and material residence processing path are established according to the record category to generate branch calcination records;

[0090] The improved TTMs model compares the consistency of changes between the branch calcination records based on the normal temperature constraint marker, determines the under-burning deviation direction and the over-burning deviation direction, and generates under-burning deviation records and over-burning deviation records.

[0091] Identify hidden under-burning risks, localized over-burning risks, and risk source markers based on under-burning deviation records and over-burning deviation records;

[0092] Based on the hidden underburning risk, local overburning risk, and risk source identification, control parameter correction values ​​are generated and sent to the calcination execution equipment.

[0093] Collect calcination execution feedback data, calibrate risk source identification and control parameter correction based on the calcination execution feedback data, and generate limestone calcination control parameters for the next control cycle.

[0094] In this embodiment, the process of preprocessing and then associating materials by batch includes:

[0095] The limestone calcination operation data is parsed, timestamps are unified, equipment numbers are matched, duplicates are removed, anomalies are corrected, and missing data is filled to obtain calcination operation preprocessing data.

[0096] Among these features, field parsing converts raw fields uploaded by different acquisition devices into standardized fields containing acquisition time, device number, measurement point type, operating value, unit, operating status, and material batch information; timestamp unification converts time information from different acquisition frequencies and formats to a unified time base and sorts it; device number matching binds acquisition devices with corresponding kiln sections and measurement point locations; deduplication preserves the most complete data for duplicate records of the same device, at the same time, and at the same measurement point; anomaly correction corrects or marks abnormal or faulty data based on preset operating ranges and variation ranges; and missing data completion imputes short-term missing data and marks long-term missing data, restricting its participation in subsequent analysis.

[0097] The finished product quality feedback data is analyzed for indicators, formatted, results standardized, batch matched and time aligned to obtain preprocessed quality feedback data.

[0098] Index analysis involves extracting and categorizing the test indicators reflecting the calcination effect into fields such as activity, residual components, uniformity, and quality grade; format standardization involves standardizing the indicator names, numerical formats, units, time, and batch formats; result standardization involves converting test results into a uniform quality status according to quality evaluation rules; batch matching involves matching the test batch with the output batch and tracing back to the material batch; and time alignment involves mapping quality data to the corresponding calcination time period based on the material flow process.

[0099] Based on the material batch information, by establishing a unified batch identifier, the calcination operation pretreatment data and the quality feedback pretreatment data are matched and associated, summarized and organized, and standardized calcination status data is generated.

[0100] Specifically, a unified batch identifier is established based on the feeding, calcination time period, discharge, and testing batches; operational data and quality data are collected into the corresponding batches respectively; for one-to-many or many-to-one relationships, the corresponding range is determined according to the discharge sequence and sampling records; batch records containing operational process and quality results are integrated in chronological order, and standardized calcination status data are generated.

[0101] In this embodiment, the process of generating the normal temperature operating condition segment and the normal temperature constraint identifier includes:

[0102] Extract the calcination temperature records of the same batch of materials, sort them according to the time of collection to form a temperature sequence, compare the temperature sequence with the preset target temperature range, and mark the low temperature, normal and high temperature operating condition records.

[0103] The calcination section temperature record refers to the records of temperature measuring points belonging to the calcination section in the standardized calcination state data. The temperature sequence is arranged according to the residence order of the same material batch in the calcination section, so that the temperature sequence can reflect the continuous thermal process experienced by the same material batch. The preset target temperature range refers to the temperature control interval pre-set for the limestone calcination section. This temperature control interval consists of a lower target temperature limit and a higher target temperature limit. The lower target temperature limit is used to characterize the minimum control temperature at which limestone can continuously undergo an effective calcination reaction, and the higher target temperature limit is used to characterize the maximum control temperature to avoid over-calcination of limestone. The preset target temperature range is pre-written into the control parameter table according to the kiln type, limestone particle size, fuel characteristics, material residence requirements, and finished product quality requirements, and is bound before the same material batch enters the calcination section.

[0104] When the temperature of the calcination section is lower than the lower limit of the target temperature, the record at the corresponding acquisition time is marked as a low-temperature operating condition record; when the temperature of the calcination section is between the lower limit of the target temperature and the upper limit of the target temperature, the record at the corresponding acquisition time is marked as a normal temperature operating condition record; when the temperature of the calcination section is higher than the upper limit of the target temperature, the record at the corresponding acquisition time is marked as a high-temperature operating condition record.

[0105] Low-temperature operating condition records indicate that the calcination section temperature at the current acquisition time has not reached the preset target temperature range; high-temperature operating condition records indicate that the calcination section temperature at the current acquisition time exceeds the preset target temperature range; and normal-temperature operating condition records indicate that the calcination section temperature at the current acquisition time is within the preset target temperature range. A normal-temperature operating condition record only indicates that the calcination section temperature meets the target temperature range and does not indicate that the limestone material has reached a fully calcined state.

[0106] Continuous normal temperature records are bound together, and records shorter than a preset threshold are removed to form normal temperature operating condition segments.

[0107] Continuous normal temperature records refer to a set of records within the same material batch whose acquisition times are adjacent and are all marked as normal temperature operating condition records. When binding continuous normal temperature records, the start acquisition time of the continuous record is taken as the segment start time, and the end acquisition time of the continuous record is taken as the segment end time. The standardized calcination state data corresponding to the continuous record set are grouped into the same candidate normal temperature segment.

[0108] The preset threshold is the minimum duration of a candidate normal temperature segment that can participate in the subsequent identification of hidden underburning risk and local overburning risk. The preset threshold is set in advance based on the IoT sampling cycle, the minimum effective residence time of limestone in the calcination section, and the length of the control cycle. When the duration of a candidate normal temperature segment is lower than the preset threshold, the candidate normal temperature segment is determined to be a short-term temperature fluctuation segment and is removed. When the duration of a candidate normal temperature segment reaches the preset threshold, the candidate normal temperature segment is determined to be a normal temperature operating condition segment.

[0109] Write the material batch, start and end times, and temperature matching results to the temperature normal operating condition chip, and generate a temperature normal constraint identifier.

[0110] The temperature matching result records the matching status between the calcination section temperature value and the preset target temperature range within the normal operating condition segment. The temperature matching result includes the matching status of the lower limit of the target temperature, the matching status of the upper limit of the target temperature, and the matching status of the segment duration. The normal temperature constraint identifier consists of the material batch, the segment start time, the segment end time, and the temperature matching result.

[0111] In this embodiment, the process of generating the calcination process record includes:

[0112] The standardized calcination state data within the normal temperature operating condition segment are aligned according to material batch and acquisition time, and the changes in operating state at adjacent time points are extracted.

[0113] Based on changes in fuel supply, the thermal response relationship is established by correlating changes in calcination section temperature and flue gas temperature within the lag interval.

[0114] When establishing the heat response relationship, the time when the fuel supply status change reaches a preset change threshold is extracted from the basic change records as the fuel supply change occurrence time. Based on the preset heat response lag time, a corresponding lag interval is extracted after the fuel supply change occurrence time. Within the lag interval, the calcination section temperature change and flue gas temperature change are extracted. The fuel supply change, calcination section temperature change, and flue gas temperature change are sequentially bound according to the fuel supply change occurrence time, temperature response start time, and temperature response end time. When there is a corresponding change relationship between the fuel supply change direction and the calcination section temperature change direction or the flue gas temperature change direction, the corresponding fuel supply change, calcination section temperature change, and flue gas temperature change are written into the same heat response relationship. The heat response relationship must at least record the material batch, the fuel supply change occurrence time, the lag interval, the calcination section temperature response change, the flue gas temperature response change, and the heat response direction.

[0115] Based on changes in combustion air volume, and by correlating changes in kiln pressure and flue gas oxygen content, an air-oxygen matching relationship is established.

[0116] When establishing the air-oxygen matching relationship, the collection time when the change in combustion air volume reaches a preset change threshold is extracted from the basic change records as the time when the combustion air volume change occurs. Based on the preset air-oxygen matching observation interval, a corresponding collection interval is extracted from the time when the combustion air volume change occurs. Within the collection interval, kiln pressure changes and flue gas oxygen content changes are extracted. The changes in combustion air volume, kiln pressure, and flue gas oxygen content are bound according to the times when the combustion air volume change occurs, kiln pressure change occurs, and flue gas oxygen content change occurs. When the changes in kiln pressure and flue gas oxygen content after a change in combustion air volume can form continuous change records within the same collection interval, the corresponding changes in combustion air volume, kiln pressure, and flue gas oxygen content are written into the same air-oxygen matching relationship. The air-oxygen matching relationship must at least record the material batch, the time when the combustion air volume change occurs, the air-oxygen matching observation interval, the kiln pressure change, the flue gas oxygen content change, and the air-oxygen matching direction.

[0117] Based on the feeding and discharging status, the material residence area is divided and the material residence relationship is established by combining the finished product quality feedback;

[0118] When establishing material dwell relationships, the feed status changes and discharge status changes of the same material batch are extracted from the basic change records. The acquisition time corresponding to the feed status change is used as the starting reference time for the material entering the normal temperature operating condition segment, and the acquisition time corresponding to the discharge status change is used as the ending reference time for the material leaving the normal temperature operating condition segment. Material dwell intervals are divided based on the starting and ending reference times. When the same material batch has multiple feed status changes and multiple discharge status changes within the normal temperature operating condition segment, they are paired one by one according to the order of acquisition time to form multiple material dwell intervals. The material dwell intervals are bound to the finished product quality feedback status to form a material dwell relationship that reflects the correspondence between the material dwell process and the finished product quality feedback. The material dwell relationship records at least the material batch, feed reference time, discharge reference time, material dwell interval, finished product quality feedback status, and dwell matching result.

[0119] The three relationships are classified, identified, and merged to generate a calcination process record.

[0120] The classification and identification process involves writing heat response category identifiers, air-oxygen matching category identifiers, and material residence category identifiers to the heat response relationship, air-oxygen matching category, and material residence category, respectively, according to the relationship type. Simultaneously, material batch, segment start time, segment end time, and temperature normal constraint identifiers are written, ensuring that each relationship record can be traced back to the corresponding temperature normal operating condition segment. The merging process involves grouping relationship records according to relationship type within the same material batch and the same temperature normal operating condition segment; merging relationship records with continuous acquisition times, consistent change directions, and overlapping corresponding intervals within the same relationship type; retaining relationship records with continuous acquisition times but opposite change directions within the same relationship type; and summarizing the merged heat response relationship, air-oxygen matching relationship, and material residence relationship in the order of acquisition time to generate a calcination process record. The calcination process record includes heat response relationship, air-oxygen matching relationship, material residence relationship, relationship category identifier, material batch, and temperature normal constraint identifier.

[0121] In this embodiment, the improved TTMs model includes a record category encoding unit, a three-path input construction unit, a TTMs time-series hybrid backbone, and a branched calcination record output unit. The process of generating branched calcination records includes:

[0122] The timing of the fuel supply change in the heat response relationship is taken as the starting point of the path, the lag interval is taken as the path interval, and the temperature changes of the calcination section and the flue gas temperature within the lag interval are taken as path response items. The heat response processing path is generated in the order of the path starting points.

[0123] The moment of fuel supply change is used as the path starting point, and the moment when the fuel supply status undergoes an effective change is used as the starting position of the heat response processing path. The lag interval, as the path interval, is a time interval formed by continuously collected records starting after the moment of fuel supply change, according to a preset heat response lag duration. The calcination section temperature change and flue gas temperature change are used as path response items. Within the lag interval, the direction and magnitude of the calcination section temperature change, the direction and magnitude of the flue gas temperature change are extracted, and the change data are bound to the same path starting point, so that the path response items can reflect the corresponding heat response results after the fuel supply change. When generating the heat response processing path, each path starting point, corresponding path interval, and corresponding path response item are combined into a heat response path record, arranged in chronological order of the path starting points.

[0124] The moment when the combustion air volume changes in the air-oxygen matching relationship occurs is taken as the starting point of the path, the air-oxygen matching observation interval is taken as the path interval, and the changes in kiln pressure and flue gas oxygen content within the air-oxygen matching observation interval are taken as path response items. The air-oxygen matching processing path is generated in the order of the path starting points.

[0125] The moment of change in combustion air volume is used as the path starting point, meaning the starting position of the air-oxygen matching processing path is the moment when the combustion air volume status undergoes an effective change. The air-oxygen matching observation interval, as the path interval, refers to the time interval formed by continuously collecting data from the moment the combustion air volume change occurs, according to a preset air-oxygen matching observation duration. Kiln pressure change and flue gas oxygen content change are used as path response items. Within the air-oxygen matching observation interval, the direction and magnitude of kiln pressure change, and the direction and magnitude of flue gas oxygen content change are extracted. These change data are then bound to the same path starting point, enabling the path response item to reflect the kiln pressure response and flue gas oxygen content response corresponding to the change in combustion air volume. When generating the air-oxygen matching processing path, each path starting point, corresponding path interval, and corresponding path response item are combined into an air-oxygen matching path record, arranged in chronological order of the path starting points.

[0126] The material dwell time reference time in the material dwell relationship is used as the starting point of the path, the discharge reference time is used as the ending point of the path, the material dwell interval and the finished product quality feedback status are used as the path result items, and the material dwell processing path is generated according to the order of the path starting points.

[0127] The feeding reference time serves as the path start point, with the collection time of the same material batch entering the normal temperature operating segment as the starting position of the material dwell time processing path. The discharging reference time serves as the path end point, with the collection time of the same material batch leaving the normal temperature operating segment as the ending position of the material dwell time processing path. The material dwell time interval, as a path result item, is the continuous dwell time period formed from the feeding reference time to the discharging reference time. The finished product quality feedback status, as a path result item, refers to writing the quality feedback preprocessing data into this continuous dwell time period, so that the material dwell time processing path can reflect the correspondence between the material dwell process and the finished product quality feedback result. When generating the material dwell time processing path, the path start point, path end point, material dwell time interval, and finished product quality feedback status are combined into a material dwell time path record, arranged in the order of the path start point.

[0128] The heat response processing path, the air-oxygen matching processing path, and the material residence processing path are grouped together according to the same material batch and the same temperature normal constraint identifier to generate a separate calcination record;

[0129] The data collection process includes: using material batch and normal temperature constraint identifier as collection conditions, filtering heat response path records, air-oxygen matching path records, and material residence path records belonging to the same normal temperature operating condition segment; writing heat response path records into heat response sub-paths, air-oxygen matching path records into air-oxygen matching sub-paths, and material residence path records into material residence sub-paths; sorting the path records within the three sub-paths according to the order of collection time, retaining the path start point, path interval, path response item, path end point, and path result item in each path record; binding the three sub-paths with the same material batch and the same normal temperature constraint identifier to generate sub-path calcination records.

[0130] In this embodiment, the improved TTMs model includes a temperature normal constraint control unit and a deviation direction output unit. The process of generating underheating deviation records and overheating deviation records includes:

[0131] Using the material residence interval as a reference, the infeed reference time to the discharge reference time of the same material batch within the normal temperature operating segment is used as the reference interval. This ensures that the heat response processing path, the air-oxygen matching processing path, and the material residence processing path are all compared within the same material residence interval. This process can avoid incorrectly associating operational changes between different material batches or different normal temperature operating segments.

[0132] Based on the material batch, normal temperature constraint identifier and collection time in the branch calcination record, determine the heat response path record and wind-oxygen matching path record that have a time correspondence with the material residence interval;

[0133] When determining the time correspondence, first filter the heat response path records and wind-oxygen matching path records where the material batch and temperature normal constraint identifier are the same. Then, read the path start point and path interval of each heat response path record to determine whether the path start point or path interval is located within the material residence interval. When the path start point is located within the material residence interval, or the path interval overlaps with the material residence interval in time, the heat response path record is determined to be a heat response path record with a time correspondence with the material residence interval. Read the path start point and path interval of each wind-oxygen matching path record to determine whether the path start point or path interval is located within the material residence interval. When the path start point is located within the material residence interval, or the path interval overlaps with the material residence interval in time, the wind-oxygen matching path record is determined to be a wind-oxygen matching path record with a time correspondence with the material residence interval.

[0134] When the same material residence area corresponds to more than two heat response path records, all heat response path records are retained in the order of their starting points and used for subsequent comparisons; when the same material residence area corresponds to more than two wind-oxygen matching path records, all wind-oxygen matching path records are retained in the order of their starting points and used for subsequent comparisons; when there is no time correspondence between the heat response path record or the wind-oxygen matching path record and the material residence area, the corresponding path record is excluded from the deviation direction determination for that material residence area.

[0135] Compare the fuel supply changes in the heat response path record with the path response items to mark the insufficient heat response state;

[0136] Read the direction and magnitude of fuel supply changes, and read the direction, magnitude, and flue gas temperature changes in the calcination section, as well as the direction and magnitude of flue gas temperature changes in the path response item. When the fuel supply increases, if the magnitude of the calcination section temperature change in the corresponding path interval does not reach the preset thermal response magnitude, or if the direction of the flue gas temperature change does not form a corresponding temperature rise response with the fuel supply change, mark the state of insufficient thermal response. When the fuel supply decreases, if the calcination section temperature and flue gas temperature in the corresponding path interval continue to rise and exceed the preset magnitude, do not mark the state of insufficient thermal response, and retain the control result as the thermal response control result for the non-underburning direction.

[0137] Compare the changes in combustion air volume in the wind-oxygen matching path record with the path response items to mark abnormal wind-oxygen matching status.

[0138] Read the direction and magnitude of changes in combustion air volume, and read the direction and magnitude of changes in kiln pressure, flue gas oxygen content, and flue gas oxygen content in the path response item. When the combustion air volume changes, if the changes in kiln pressure and flue gas oxygen content do not form a continuous response matching the change in combustion air volume within the corresponding path interval, or if the magnitude of the change in flue gas oxygen content exceeds the preset air-oxygen matching range, mark the air-oxygen matching as abnormal. When the changes in kiln pressure and flue gas oxygen content are within the preset air-oxygen matching range after the change in combustion air volume, do not mark the air-oxygen matching as abnormal, and retain the result as the normal air-oxygen matching control result.

[0139] Compare the material dwell area and finished product quality feedback status with the preset material dwell range and preset quality target, and mark the material dwell status and quality feedback status;

[0140] The process involves reading the start and end times of the material dwell time interval to determine the duration of material dwell time. This duration is then compared to a preset material dwell time range. If the duration is below the lower limit of the preset range, the material dwell time is marked as insufficient. If it is within the preset range, it is marked as normal. If it is above the upper limit, it is marked as excessive. The quality feedback status marking process includes: reading the finished product quality feedback status and comparing it to a preset quality target. If the quality feedback status meets the target, it is marked as normal. If it does not meet the target, it is marked as deviating from the target.

[0141] When insufficient heat response, insufficient material residence and quality deviation exist simultaneously in the same material residence range, determine the direction of underburning deviation and generate an underburning deviation record;

[0142] Within the same material residence interval, the states of insufficient heat response, insufficient material residence, and quality deviation are combined and judged in the same interval. When all three exist simultaneously, the corresponding material residence interval is judged as the under-burning deviation direction jointly pointed to by insufficient heat response and insufficient material residence. The under-burning deviation record is written into the material batch, temperature normal constraint mark, material residence interval, corresponding heat response path record, insufficient material residence state, and quality deviation state.

[0143] When the same material residence area simultaneously exhibits abnormal air-oxygen matching, excessively long material residence time, and quality deviation, the direction of overburning deviation is determined, and an overburning deviation record is generated.

[0144] Within the same material residence area, the abnormal air-oxygen matching status, the excessively long material residence status, and the quality deviation status are combined and judged in the same period. When all three exist simultaneously, the corresponding material residence area is judged as the overburning deviation direction jointly pointed to by the abnormal air-oxygen matching and the excessively long material residence. The overburning deviation record is written into the material batch, the normal temperature constraint mark, the material residence area, the corresponding air-oxygen matching path record, the excessively long material residence status, and the quality deviation status.

[0145] The "same interval state combination judgment" refers to: using the same material batch, the same normal temperature constraint indicator, and the same material residence interval as the judgment range, reading the following states within this judgment range: insufficient heat response, abnormal airflow-oxygen matching, insufficient material residence, normal material residence, excessively long material residence, normal quality, and quality deviation. These states are then written into the same interval state judgment group. Within the same interval state judgment group, if three states—insufficient heat response, insufficient material residence, and quality deviation—exist, the interval state judgment group is determined as an underburning deviation combination, and the direction of the underburning deviation is determined. If three states—abnormal airflow-oxygen matching, excessively long material residence, and quality deviation—exist, the interval state judgment group is determined as an overburning deviation combination, and the direction of the overburning deviation is determined.

[0146] When there is no simultaneous combination of insufficient heat response, insufficient material residence, and quality deviation within the same material residence interval, and there is no simultaneous combination of abnormal air-oxygen matching, excessively long material residence, and quality deviation, no under-burning deviation record or over-burning deviation record is generated; the corresponding material residence interval is marked as having no deviation direction, and the heat response comparison result, air-oxygen matching comparison result, material residence status, and quality feedback status corresponding to that material residence interval are retained in the branch calcination record for continued comparison in the next control cycle.

[0147] When the same material residence area simultaneously meets the judgment conditions for both under-burning deviation and over-burning deviation, under-burning deviation records and over-burning deviation records are generated respectively, and further distinctions are made based on the risk source identifier during subsequent risk identification.

[0148] Under-burning deviation records formed under the same material batch, the same temperature normal constraint mark, and the same material residence range are grouped into the under-burning deviation record set, and over-burning deviation records formed under the same material batch, the same temperature normal constraint mark, and the same material residence range are grouped into the over-burning deviation record set. The under-burning deviation record set and the over-burning deviation record set are jointly output to the risk identification step to identify hidden under-burning risks, local over-burning risks, and risk source marks.

[0149] For example, within a normal operating temperature segment corresponding to the same material batch A, the calcination section temperature is always within the preset target temperature range, therefore this segment is marked with a normal temperature constraint. Material batch A forms a material residence interval within this segment. The system uses this material residence interval as a reference to search for heat response path records and air-oxygen matching path records under the same material batch and the same normal temperature constraint.

[0150] Within the material residence interval, the fuel supply was increased. The system read the corresponding heat response path record and found that after the fuel supply change, the temperature change in the calcination section was insufficient within the preset heat response lag interval, and the flue gas temperature did not show a corresponding temperature rise response. Therefore, it was marked as a state of insufficient heat response. The system then read the material residence interval and found that the residence duration of this batch of material was lower than the lower limit of the preset material residence range. Therefore, it was marked as a state of insufficient material residence. Subsequently, the corresponding finished product quality feedback status was read and it was found that the activity of the finished product did not meet the preset quality target. Therefore, it was marked as a state of quality deviation.

[0151] Because the same material batch A, the same normal temperature constraint identifier, and the same material residence range simultaneously exhibit insufficient heat response, insufficient material residence, and quality deviation states, the system identifies this range's state judgment group as an under-burning deviation combination and generates an under-burning deviation record. This under-burning deviation record includes material batch A, normal temperature constraint identifier, material residence range, corresponding heat response path record, insufficient material residence state, and quality deviation state.

[0152] For example, in the normal operating condition segment corresponding to another material batch B, the calcination section temperature is also within the preset target temperature range. Using the material residence interval of material batch B as a reference, the system reads the corresponding air-oxygen matching path record and finds that after the combustion air volume is increased, the changes in kiln pressure and flue gas oxygen content are not within the preset air-oxygen matching range, thus marking it as an abnormal air-oxygen matching state. After reading the material residence interval, the system finds that the material residence duration is higher than the upper limit of the preset material residence range, thus marking it as a state of excessively long material residence. The finished product quality feedback status shows a decrease in finished product activity, which is also marked as a quality deviation state.

[0153] Because the same batch of material B, the same normal temperature constraint, and the same material residence range all have abnormal air-oxygen matching, excessively long material residence, and quality deviation states, the system determines the state judgment group of this range as an overburning deviation combination and generates an overburning deviation record.

[0154] In this embodiment, the improved TTMs model uses the Tiny Time Mixers model as the basic time series mixing backbone. The Tiny Time Mixers model is a lightweight pre-trained model for multivariate time series processing, suitable for time series segment processing, time dimension mixing, and variable channel mixing of continuously acquired industrial process data. The improvement in this invention does not change the basic time series mixing capability of the Tiny Time Mixers model, but adds a record category encoding unit and a three-path input construction unit at the front end of the basic time series mixing backbone, and adds a branched calcination record output unit, a temperature normal constraint control unit, and a deviation direction output unit at the back end of the basic time series mixing backbone.

[0155] The record category coding unit is used to read the relationship category identifier in the calcination process record and split the calcination process record into heat response relationship, air-oxygen matching relationship, and material residence relationship. The heat response relationship is used to characterize the response process of calcination section temperature and flue gas temperature after changes in fuel supply; the air-oxygen matching relationship is used to characterize the response process of kiln pressure and flue gas oxygen content after changes in combustion air volume; the material residence relationship is used to characterize the residence process of material from the feed reference time to the discharge reference time and the finished product quality feedback status.

[0156] The three-path input construction unit is used to convert heat response relationships into heat response input sequences, air-oxygen matching relationships into air-oxygen matching input sequences, and material residence relationships into material residence input sequences. The heat response input sequence includes the time of fuel supply change, the direction of fuel supply change, the magnitude of fuel supply change, the lag interval, the direction of calcination section temperature change, the magnitude of calcination section temperature change, the direction of flue gas temperature change, and the magnitude of flue gas temperature change. The air-oxygen matching input sequence includes the time of combustion air volume change, the direction of combustion air volume change, the magnitude of combustion air volume change, the air-oxygen matching observation interval, the direction of kiln pressure change, the magnitude of kiln pressure change, the direction of flue gas oxygen content change, and the magnitude of flue gas oxygen content change. The material residence input sequence includes the feed reference time, the discharge reference time, the material residence interval, the material residence duration, and the finished product quality feedback status.

[0157] The TTMs temporal hybrid backbone is used to process the heat response input sequence, wind-oxygen matching input sequence, and material residence input sequence separately. The processing includes: dividing each type of input sequence into fixed-length time segments according to the acquisition time; mixing the time dimension of the change relationship between adjacent acquisition times within the same time segment; mixing the variable channel of the correspondence between different variables within the same time segment; and outputting the heat response time sequence, wind-oxygen matching time sequence, and material residence time sequence.

[0158] The branch calcination record output unit generates heat response processing paths based on heat response timing, air-oxygen matching processing paths based on air-oxygen matching timing, and material residence processing paths based on material residence timing. The heat response processing path uses the moment of fuel supply change as the starting point, the lag interval as the path interval, and the changes in calcination section temperature and flue gas temperature as path response items. The air-oxygen matching processing path uses the moment of combustion air volume change as the starting point, the air-oxygen matching observation interval as the path interval, and the changes in kiln pressure and flue gas oxygen content as path response items. The material residence processing path uses the feed reference time as the starting point, the discharge reference time as the ending point, and the material residence interval and finished product quality feedback status as path result items. The three types of processing paths are grouped according to the same material batch and the same temperature normal constraint identifier to generate branch calcination records.

[0159] The temperature normal constraint comparison unit is used to read the temperature normal constraint identifier and uses the material residence interval in the material residence processing path as the comparison benchmark. The temperature normal constraint comparison unit first filters heat response path records and air-oxygen matching path records under the same material batch and the same temperature normal constraint identifier. Then, it determines whether the path start point or path interval of the heat response path record has a time correspondence with the material residence interval, and similarly determines whether the path start point or path interval of the air-oxygen matching path record has a time correspondence with the material residence interval. Path records with a time correspondence participate in the consistency comparison of changes within the same material residence interval; path records without a time correspondence do not participate in the deviation direction determination for that material residence interval.

[0160] The deviation direction output unit is used to read the states of insufficient heat response, abnormal air-oxygen matching, insufficient material residence, excessive material residence, and quality deviation within the same material residence area. When these states exist within the same material residence area, the deviation direction output unit outputs the underburning deviation direction, generating an underburning deviation record. When these states exist, the deviation direction output unit outputs the overburning deviation direction, generating an overburning deviation record. When no such combination exists within the same material residence area, the deviation direction output unit does not output underburning or overburning deviation directions, but retains the corresponding path record for the next control cycle for continued comparison.

[0161] In this embodiment, the process of identifying hidden underburning risks, localized overburning risks, and risk source identifiers includes:

[0162] A continuity check is performed on the underburning deviation records, and underburning deviation records that are collected continuously and correspond to the same material residence interval are merged into hidden underburning risks.

[0163] Read the material batch, temperature normal constraint indicator, material residence range, and collection time from each underburning deviation record. Filter underburning deviation records belonging to the same material batch, the same temperature normal constraint indicator, and the same material residence range, and arrange them in order of collection time from first to last. When the collection time interval between two adjacent underburning deviation records does not exceed a preset continuous interval, and both adjacent underburning deviation records point to insufficient heat response, insufficient material residence, and quality deviation, the adjacent two underburning deviation records are determined to be continuous. When the number of consecutive underburning deviation records reaches a preset consecutive number, or the duration of consecutive underburning deviation records reaches a preset duration, the corresponding underburning deviation records are merged into a latent underburning risk.

[0164] When merging into latent underburning risks, the material batch, temperature normal constraint mark, and material residence interval corresponding to the continuous underburning deviation record are used as risk attribution information, and the start and end collection times of the continuous underburning deviation record are used as the risk duration interval. The corresponding insufficient heat response state, insufficient material residence state, and quality deviation state are retained. A single underburning deviation record that does not meet the continuous inspection conditions is not merged into a latent underburning risk, but is retained as an underburning deviation record to be observed.

[0165] Based on the insufficient heat response state and insufficient material retention state corresponding to the hidden underburning risk, determine the source identifier of insufficient heat response and the source identifier of insufficient material retention;

[0166] The overheating deviation records are checked for continuity. Overheating deviation records that are collected continuously and correspond to the same material residence range are merged into local overheating risk.

[0167] Read the material batch, normal temperature constraint identifier, material residence range, and collection time from each overheating deviation record. Filter overheating deviation records belonging to the same material batch, the same normal temperature constraint identifier, and the same material residence range, and arrange them in order of collection time from first to last. When the collection time interval between two adjacent overheating deviation records does not exceed a preset continuous interval, and both adjacent overheating deviation records point to abnormal air-oxygen matching, excessively long material residence, and quality deviation, the adjacent two overheating deviation records are determined to be continuous. When the number of consecutive overheating deviation records reaches a preset consecutive number, or the duration of consecutive overheating deviation records reaches a preset duration, the corresponding overheating deviation records are merged into a local overheating risk.

[0168] When merging into local overheating risks, the material batch, temperature normal constraint mark, and material residence interval corresponding to the continuous overheating deviation record are used as risk attribution information. The start and end collection times of the continuous overheating deviation record are used as the risk duration interval. The corresponding abnormal air-oxygen matching state, excessive material residence time state, and quality deviation state are retained. Single overheating deviation records that do not meet the continuous inspection conditions are not merged into local overheating risks, but are retained as overheating deviation records to be observed.

[0169] Based on the abnormal state of air-oxygen matching and the state of excessive material residence time corresponding to the risk of local overburning, determine the source identifier of abnormal air-oxygen matching and the source identifier of excessive material residence time;

[0170] When both latent underburning risk and local overburning risk are generated within the same material residence area, the latent underburning risk, local overburning risk and corresponding risk source identifiers are retained respectively.

[0171] In this embodiment, the process of generating the control parameter correction amount includes:

[0172] Determine the control cycle to be corrected corresponding to the material residence range;

[0173] The changes in fuel supply, heat response, material residence time, combustion air volume, kiln pressure, and flue gas oxygen content are read from the control cycle to be corrected.

[0174] When there is an indication of insufficient heat response, the fuel supply correction amount is generated based on the difference between the preset heat response amplitude and the actual heat response amplitude.

[0175] When there are sources indicating insufficient or excessive material dwell time, the feed rate correction amount or discharge rate correction amount is generated based on the difference between the preset material dwell range and the actual dwell length.

[0176] When there is an abnormal source of air-oxygen matching, the combustion air volume correction and kiln pressure correction are generated based on the kiln pressure and flue gas oxygen content deviation.

[0177] Each control correction is limited and merged to generate control parameter correction values.

[0178] Specifically, the correction amount is generated based on the difference. Instead of directly using the difference as the control quantity for the calcination equipment, the difference is treated as a deviation and combined with preset control response conversion relationships, correction directions, and single correction limits to generate the control parameter correction amount. The control response conversion relationships are pre-generated from historical calcination control records, equipment calibration records, and finished product quality feedback records. Specifically, historical control parameter changes are batch-linked with the calcination state response after the corresponding control cycle. The magnitude of calcination state change caused by a unit change in control parameters is calculated, forming conversion relationships between fuel supply and heat response magnitude, between feed rate and discharge rate and material residence duration, and between combustion air volume and kiln pressure and flue gas oxygen content and kiln pressure changes. These relationships are then written into the control parameter table.

[0179] When a source of insufficient heat response is identified, the preset heat response amplitude and the actual heat response amplitude are read. The heat response gap is obtained by subtracting the actual heat response amplitude from the preset heat response amplitude. The heat response gap indicates that the temperature changes in the calcination section and the flue gas temperature after a change in fuel supply have not reached the expected response level. Based on the heat response gap, the fuel supply correction direction is determined to be an increasing direction. According to the control response conversion relationship between fuel supply and heat response amplitude, the heat response gap is converted into an initial fuel supply correction amount. The initial fuel supply correction amount is compared with the single fuel supply correction limit. If the limit is exceeded, the fuel supply correction amount is generated according to the limit value; otherwise, the fuel supply correction amount is generated according to the initial fuel supply correction amount.

[0180] When a source of insufficient material dwell time is identified, the preset lower limit of the material dwell time range and the material dwell time duration are read. The insufficient dwell time is obtained by subtracting the material dwell time duration from the preset lower limit of the material dwell time range. The insufficient dwell time indicates that the material does not receive heat for a sufficient duration within the normal temperature operating segment. Based on the insufficient dwell time, the correction direction for the material throughput speed is determined to be a decreasing direction. Then, according to the control response conversion relationship between the feed rate, discharge rate, and material dwell time duration, the insufficient dwell time is converted into the initial feed rate correction amount and the initial discharge rate correction amount. After single-correction limiting processing, the feed rate correction amount and the discharge rate correction amount are generated.

[0181] When a source of excessive material dwell time is detected, the system reads the material dwell duration and the preset upper limit of the material dwell range. The excessive dwell time is then calculated by subtracting the preset upper limit from the material dwell duration. The excessive dwell time indicates that the material has been subjected to heat for an excessively long period within the normal operating temperature range. Based on the excessive dwell time, the correction direction for the material throughput speed is determined to be increasing. According to the control response conversion relationship between the discharge speed and the material dwell duration, the excessive dwell time is converted into an initial discharge speed correction amount. After a single correction and limiting process, the discharge speed correction amount is generated.

[0182] When an abnormal source of air-oxygen matching is detected, the kiln pressure change, flue gas oxygen content change, preset kiln pressure matching range, and preset flue gas oxygen content matching range are read to determine the kiln pressure deviation and oxygen content deviation. The kiln pressure deviation indicates the degree to which the current kiln pressure response deviates from the target kiln pressure response, and the oxygen content deviation indicates the degree to which the current flue gas oxygen content deviates from the target oxygen content state. The kiln pressure correction direction is determined based on the kiln pressure deviation, and the combustion air volume correction direction is determined based on the oxygen content deviation. According to the control response conversion relationship between combustion air volume and kiln pressure and flue gas oxygen content and kiln pressure changes, the kiln pressure deviation is converted into the initial kiln pressure correction, and the oxygen content deviation is converted into the initial combustion air volume correction. After single-correction limiting processing, the kiln pressure correction and combustion air volume correction are generated.

[0183] The preset control response conversion relationship refers to the pre-established correspondence between changes in control parameters and changes in calcination state. It represents the degree of impact of unit changes in fuel supply, feed rate, discharge rate, combustion air volume, and kiln pressure on the magnitude of heat response, material residence duration, flue gas oxygen content, and kiln pressure. This control response conversion relationship is obtained by statistically analyzing historical calcination control records, equipment calibration records, and finished product quality feedback records. It is written into the control parameter table and used to convert heat response gaps, insufficient residence time, excessive residence time, kiln pressure deviations, and oxygen content deviations into corresponding control parameter corrections.

[0184] For example, for every 1% increase in fuel supply, the corresponding increase in the kiln's thermal response is approximately 2% to 3%, while the oxygen content in the flue gas decreases by approximately 0.2% to 0.5%.

[0185] For every 1% increase in feed rate, the corresponding material residence time decreases by approximately 1% to 2%, and the heat demand in the kiln increases by approximately 1% to 2%.

[0186] For every 1% increase in discharge speed, the corresponding residence time of the material in the kiln is shortened by about 1% to 1.5%, which may lead to incomplete calcination.

[0187] For every 1% increase in combustion air volume, the corresponding oxygen content in the flue gas increases by approximately 0.3% to 0.6%, while the combustion intensity inside the kiln also increases.

[0188] For every 10Pa increase in kiln pressure, the corresponding increase in gas flow resistance inside the kiln may lead to a decrease in flue gas discharge efficiency of approximately 1% to 2%.

[0189] The above conversion relationships can be corrected and calibrated according to different equipment specifications, raw material characteristics and operating conditions, and used to guide the dynamic adjustment of control parameters.

[0190] A batch of material was identified as having a latent underburning risk during a normal operating temperature period. The control parameter table pre-records that for every 1% increase in fuel supply, the thermal response amplitude increases by approximately 2%. The preset thermal response amplitude for this batch is 10%, the actual thermal response amplitude is 6%, and the thermal response shortfall is 4%. Based on the control response conversion relationship between fuel supply and thermal response amplitude, the initial fuel supply correction is determined to be an increase of 2%. If the single fuel supply correction limit is 3%, then this initial correction does not exceed the limit, and the generated fuel supply correction is an increase of 2%.

[0191] The same batch of materials also had an insufficient material residence time indicator. The control parameter table pre-recorded that for every 1% decrease in feed rate, the material residence time increases by approximately 1.5%. The preset lower limit of the material residence range for this batch was 50, the actual material residence time was 47, and the insufficient residence time was 3. Based on the control response conversion relationship between feed rate and material residence time, a feed rate correction of 2% was generated, allowing subsequent materials to obtain a longer thermal treatment time in the calcination section.

[0192] In another batch of materials, the system identified localized overburning risk and sources of abnormal air-oxygen matching. The control parameter table pre-recorded that for every 1% decrease in combustion air volume, the flue gas oxygen content changes by approximately 0.4%. In this batch, the flue gas oxygen content was higher than the preset flue gas oxygen content matching range, with an oxygen content deviation of 0.8%. Based on the control response conversion relationship between combustion air volume and flue gas oxygen content, a combustion air volume correction of 2% was generated. If the kiln pressure also deviates from the preset kiln pressure matching range, a kiln pressure correction was simultaneously generated based on the kiln pressure control response conversion relationship. These corrections, after being limited, were written into the control command of the control cycle to be corrected and sent to the corresponding calcination execution equipment.

[0193] In this embodiment, the process of calibrating the risk source identification and control parameter correction includes:

[0194] By comparing the control parameter correction with the actual execution value, the equipment execution deviation is obtained, the control parameter correction is calibrated, and the execution calibration correction is obtained.

[0195] Specifically, the system reads the control parameter correction values ​​issued within the control cycle to be corrected, and reads the actual execution values ​​fed back by the calcination equipment within the same control cycle. The system then subtracts the target correction value from the actual correction value from the actual execution value for each controlled object to obtain the fuel supply execution deviation, feed rate execution deviation, discharge rate execution deviation, combustion air volume execution deviation, and kiln pressure execution deviation. If the absolute value of the execution deviation for a certain controlled object is greater than the preset execution deviation threshold, it is determined that the corresponding controlled object has an execution deviation, and this execution deviation is written into the execution calibration record.

[0196] The correction amount based on the equipment execution deviation calibration control parameter refers to retaining the portion of the original control parameter correction amount that was not actually completed by the calcining execution equipment in the next control cycle, and recalculating it in combination with the upper limit, lower limit and single correction limit of the calcining execution equipment to form the execution calibration correction amount.

[0197] The heat response, kiln pressure change, and finished product quality status are compared with the corresponding preset targets to generate calcination state convergence results.

[0198] The corresponding preset target refers to the target control requirements that match the various feedback parameters in the calcination execution feedback data; the preset target corresponding to the heat response is the preset heat response amplitude, which determines whether the target heat response is formed after the fuel supply is corrected; the preset target corresponding to the kiln pressure change is the preset kiln pressure matching range, which determines whether the kiln pressure control requirements are met after the kiln pressure is adjusted; the preset target corresponding to the finished product quality status is the preset quality target, which determines whether the calcination quality of the corresponding material batch reaches the target quality status.

[0199] When the convergence result of the calcination state meets the preset convergence condition, the corresponding risk source identifier is retained, and the calibration correction amount is processed according to the first amplitude coefficient.

[0200] Preset consistency conditions refer to the judgment rules used to determine whether the risk source identifier corresponds to the calcination state response after execution. Preset consistency conditions include: when the risk source identifier is an insufficient heat response source identifier, after performing fuel supply-related corrections, the absolute value of the deviation between the heat response and the preset heat response amplitude should be less than the absolute value of the corresponding deviation in the previous control cycle; when the risk source identifier is an abnormal air-oxygen matching source identifier, after performing combustion air volume and kiln pressure-related corrections, the kiln pressure change should be within the preset kiln pressure matching range, or the absolute value of the deviation between the kiln pressure change and the preset kiln pressure matching range should be less than the absolute value of the corresponding deviation in the previous control cycle; when the risk source identifier is an insufficient material residence source identifier or an excessively long material residence source identifier, after performing feed rate or discharge rate-related corrections, the finished product quality status should meet the preset quality target, or the quality deviation level corresponding to the finished product quality status should be better than the quality deviation level corresponding to the previous control cycle.

[0201] If the calcination state response after execution does not conform to the above correspondence, it is determined that the calcination state convergence result and the risk source identifier do not meet the preset consistency condition. At this time, the deviation types of the heat response, kiln pressure change and finished product quality status after execution are read, and the heat response insufficient source identifier, the air-oxygen matching abnormal source identifier, the material residence insufficient source identifier or the material residence excessively long source identifier are rematched. Based on the rematched risk source identifier, the execution calibration correction amount is re-determined.

[0202] The first amplitude coefficient is a convergence amplitude processing coefficient pre-written into the control parameter table. It is used to process and execute calibration corrections when the convergence result in the calcination state meets the preset convergence conditions. The value of the first amplitude coefficient is greater than 0 and less than or equal to 1. During processing, the calibration correction is multiplied by the first amplitude coefficient to obtain the correction amount to be synthesized in the next control cycle. This setting indicates that the current risk source identification is still valid, but the control correction has already generated a response in the target direction. The next control cycle will continue to execute closed-loop calibration according to the convergence amplitude processing coefficient.

[0203] When the convergence result of the calcination state does not meet the preset convergence conditions, the corresponding risk source identifier is retained, and the calibration correction amount is performed according to the second amplitude coefficient.

[0204] The second amplitude coefficient is a non-convergence amplitude processing coefficient pre-written into the control parameter table. It is used to process the calibration correction amount when the convergence result in the calcination state does not meet the preset convergence conditions. The value of the second amplitude coefficient is greater than or equal to 1, and the second amplitude coefficient is not less than the first amplitude coefficient. During processing, the calibration correction amount is multiplied by the second amplitude coefficient, and then compared with the single correction limit of the corresponding controlled object. If it exceeds the single correction limit, the value of the single correction limit is used; if it does not exceed the single correction limit, the value of the product is used to obtain the correction amount to be synthesized in the next control cycle. This setting indicates that the current risk source identification is still valid, but the correction in the previous control cycle did not make the corresponding calcination state meet the convergence judgment, and closed-loop calibration needs to continue according to the non-convergence amplitude processing coefficient.

[0205] When the convergence result of the calcination state does not meet the preset consistency condition with the risk source identifier, the risk source identifier is corrected, and the calibration correction amount is re-determined based on the corrected risk source identifier.

[0206] Preset consistency conditions refer to the judgment rules used to determine whether the risk source identifier corresponds to the calcination state response after execution. Preset consistency conditions include: when the risk source identifier is an insufficient heat response source identifier, after performing fuel supply-related corrections, the absolute value of the deviation between the heat response and the preset heat response amplitude should be less than the absolute value of the corresponding deviation in the previous control cycle; when the risk source identifier is an abnormal air-oxygen matching source identifier, after performing combustion air volume and kiln pressure-related corrections, the kiln pressure change should be within the preset kiln pressure matching range, or the absolute value of the deviation between the kiln pressure change and the preset kiln pressure matching range should be less than the absolute value of the corresponding deviation in the previous control cycle; when the risk source identifier is an insufficient material residence source identifier or an excessively long material residence source identifier, after performing feed rate or discharge rate-related corrections, the finished product quality status should meet the preset quality target, or the quality deviation level corresponding to the finished product quality status should be better than the quality deviation level corresponding to the previous control cycle.

[0207] When the calcination state response after execution does not conform to the above correspondence, it is determined that the calcination state convergence result and the risk source identifier do not meet the preset consistency condition. At this time, the deviation types of the heat response, kiln pressure change and finished product quality status after execution are read, and the heat response insufficient source identifier, the air-oxygen matching abnormal source identifier, the material residence insufficient source identifier or the material residence excessively long source identifier are rematched. The execution calibration correction amount is re-determined based on the rematched risk source identifier.

[0208] The current limestone calcination control parameters, the processed execution calibration correction, and the corrected risk source identifier are combined to generate the limestone calcination control parameters for the next control cycle.

[0209] An intelligent control system for limestone calcination based on the Internet of Things according to an embodiment of the present invention includes the following modules:

[0210] The data acquisition and standardization module is used to collect calcination operation data and finished product quality data, preprocess and associate them by batch to generate standardized calcination status data;

[0211] The temperature condition determination module is used to determine the temperature condition of the calcination section and generate normal temperature segments and constraint indicators.

[0212] The calcination process record generation module is used to extract heat response, air-oxygen matching and material residence records to generate calcination process records;

[0213] The branch calcination record generation module is used to input the calcination process record into the improved TTMs model, establish processing paths according to categories, and generate branch calcination records.

[0214] The deviation direction determination module is used to compare the branch records with the temperature constraint identifier to determine the deviation direction of under-burning and over-burning, and generate corresponding records.

[0215] The risk identification module is used to identify hidden under-burning risks, local over-burning risks, and the sources of these risks.

[0216] The control correction module is used to generate control parameter correction values ​​and send them to the execution device;

[0217] The feedback calibration module is used to calibrate the sources of risk and the correction amount of control parameters based on the execution feedback data, and generate the control parameters for the next cycle.

[0218] Example 2: Reference Figure 4 To verify the feasibility of this invention in practice, it was applied to a rotary kiln calcination environment with continuous operation characteristics. This scenario presents problems such as fluctuations in fuel calorific value, uneven material particle size, lagging thermal inertia within the kiln, and delayed air-oxygen regulation, leading to temperature drift and the coexistence of localized under-burning and localized over-burning during long-term operation of the calcination process.

[0219] In this implementation scenario, the calcination system structure reference Figure 4 As shown in (a), the system includes the calcining kiln body, a temperature sensing unit, a kiln pressure sensing unit, an oxygen content sensing unit, and an IoT data acquisition unit. The temperature sensor, kiln pressure sensor, and oxygen content sensor are respectively located in the stable reaction zone in the middle of the calcining kiln. Figure 4 (a) shows the location of the sensor nodes. Fuel, combustion air, and feed are input variables into the calcining kiln system, while flue gas and lime output are output variables discharged from the system. The IoT acquisition unit synchronously collects the above multi-source data with a sampling period of 1 second, and the continuous operation cycle covers approximately 20,000 sampling points.

[0220] After the collected data enters the control link, it is processed according to... Figure 4(b) shows the standardized processing of the control flow. After aligning the operational data and finished product quality feedback data with timestamps, standardized calcination state data is generated. The finished product quality feedback data includes three categories of indicators: calcium oxide content, loss on ignition, and activity index. The average calcium oxide content is set at 88% as the target control benchmark.

[0221] During the temperature condition identification phase, a sliding window analysis is performed on the calcination section temperature sequence. The window length is set to 120 seconds. When the temperature standard deviation within the window is less than 18°C, it is determined to be a normal temperature condition segment, and a normal temperature constraint marker is generated. This marker is used to limit subsequent TTMs models to only perform path comparison under stable thermal conditions.

[0222] During the path construction phase, the normal temperature operating condition segment is divided into three types of process records: heat response records, air-oxygen matching records, and material residence records, which are then input into the improved TTMs model. The model constructs heat response processing paths, air-oxygen matching processing paths, and material residence processing paths respectively, thereby forming a branched calcination record structure.

[0223] During the deviation identification phase, the improved TTMs model performs a consistency comparison analysis on the three types of paths based on the normal temperature constraint identifier. When the heat response path lags behind the fuel change by more than 6 seconds and the kiln temperature response amplitude is lower than 92% of the target response amplitude, it is determined to be an underburning deviation. When the oxygen content is higher than the upper limit of the target range by 3% and the material residence time exceeds the upper limit of the target range by 8%, it is determined to be an overburning deviation, and a corresponding deviation record is generated.

[0224] During the risk identification phase, continuous aggregation analysis is performed on the deviation records. When underheating deviations occur stably for more than 180 consecutive seconds, they are identified as latent underheating risks. When overheating deviations occur stably for more than 160 consecutive seconds, they are identified as localized overheating risks, and a risk source identifier is generated.

[0225] During the control correction phase, control parameter correction values ​​are generated based on the risk source identification and applied to the fuel regulation unit, airflow regulation unit, feed regulation unit, and discharge regulation unit. Fuel regulation is used to correct the heat input level, airflow regulation is used to correct the oxygen supply ratio, and feed and discharge regulation are used to adjust the material residence time distribution, thereby achieving multi-variable coordinated control.

[0226] During the feedback calibration phase, calcination execution feedback data is collected and compared with the control parameter correction amount for consistency. After correcting the execution deviation, the parameters for the next control cycle are generated to achieve closed-loop iterative optimization.

[0227] Table 1: Comparison and Verification Table of Calcination Control Effect of the Invention

[0228] Based on the comparison results shown in Table 1, we can analyze the results from four aspects: stability of the calcination process, ability to suppress abnormal occurrences, ability to control fluctuations in key process variables, and consistency of finished product quality.

[0229] Regarding the stability of the calcination process, the method of this invention increases the proportion of normal temperature conditions from 61.8% to 86.4%, indicating that by extracting normal temperature condition segments and introducing normal temperature constraint markers, the calcination control analysis object shifts from full-time domain data to the stable condition subspace, thereby reducing the interference of unstable disturbances on model judgment and improving the ability to identify stable operating ranges.

[0230] Regarding the ability to suppress anomalies, the proportion of underburning decreased from 13.9% to 4.1%, and the proportion of overburning decreased from 12.6% to 3.5%. This indicates that the improved TTMs model, based on the consistency comparison of the branch calcination records, has a stronger ability to distinguish between the deviation directions of underburning and overburning, enabling abnormal states to be identified and incorporated into the control and correction link at an early stage of deviation, thereby reducing the probability of deviations accumulating and expanding into overt quality defects.

[0231] Regarding the control of key process variables, the kiln temperature fluctuation range converged from 95℃ to 34℃, the kiln pressure fluctuation amplitude decreased from 19.2% to 7.1%, the oxygen content fluctuation amplitude decreased from 6.8% to 2.3%, the material residence time deviation converged from ±17.5% to ±5.8%, and the heat response lag time decreased from 11.2 seconds to 3.9 seconds. This indicates that the heat response path, the air-oxygen matching path, and the material residence path are synchronously constrained in the TTMs model, so that multivariate control is no longer adjusted independently, but is collaboratively corrected based on path consistency.

[0232] Regarding the consistency of finished product quality, the standard deviation of calcium oxide content in the finished product decreased from 0.23 to 0.09, the qualified rate of the finished product increased from 85.7% to 97.6%, and the control convergence cycle was shortened from 6–8 cycles to 2–3 cycles. This indicates that the closed-loop feedback calibration mechanism based on risk source identification can effectively eliminate the accumulation of control parameter execution deviations, enabling the parameters generated in the next control cycle to converge to the stable control range more quickly, thereby improving the overall quality consistency and control efficiency of the calcination process.

[0233] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A smart control method for limestone calcination based on the Internet of Things, characterized in that, Includes the following steps: The limestone calcination operation data and finished product quality feedback data are collected by the Internet of Things, pre-processed and associated by material batch to generate standardized calcination status data. Based on standardized calcination state data, determine the operating conditions of the calcination section temperature, and generate normal temperature operating condition segments and normal temperature constraint identifiers. Extract heat response records, air-oxygen matching records, and material residence records from normal temperature operating conditions, and generate calcination process records according to record categories; The calcination process records are input into the improved TTMs model. Based on the record category, heat response processing paths, air-oxygen matching processing paths, and material residence processing paths are established to generate branched calcination records. The improved TTMs model includes a record category encoding unit, a three-path input construction unit, a TTMs time-series hybrid backbone, and a branched calcination record output unit. The process of generating branched calcination records includes: The timing of the fuel supply change in the heat response relationship is taken as the starting point of the path, the lag interval is taken as the path interval, and the temperature changes of the calcination section and the flue gas temperature within the lag interval are taken as path response items. The heat response processing path is generated in the order of the path starting points. The moment when the combustion air volume changes in the air-oxygen matching relationship occurs is taken as the starting point of the path, the air-oxygen matching observation interval is taken as the path interval, and the changes in kiln pressure and flue gas oxygen content within the air-oxygen matching observation interval are taken as path response items. The air-oxygen matching processing path is generated in the order of the path starting points. The material dwell time reference time in the material dwell relationship is used as the starting point of the path, the discharge reference time is used as the ending point of the path, the material dwell interval and the finished product quality feedback status are used as the path result items, and the material dwell processing path is generated according to the order of the path starting points. The heat response processing path, the air-oxygen matching processing path, and the material residence processing path are grouped together according to the same material batch and the same temperature normal constraint identifier to generate a separate calcination record; The improved TTMs model compares the consistency of changes between the branch calcination records based on the normal temperature constraint marker, determines the under-burning deviation direction and the over-burning deviation direction, and generates under-burning deviation records and over-burning deviation records. Identify hidden under-burning risks, localized over-burning risks, and risk source markers based on under-burning deviation records and over-burning deviation records; Based on the hidden underburning risk, local overburning risk, and risk source identification, control parameter correction values ​​are generated and sent to the calcination execution equipment. Collect calcination execution feedback data, calibrate risk source identification and control parameter correction based on the calcination execution feedback data, and generate limestone calcination control parameters for the next control cycle.

2. The intelligent control method for limestone calcination based on the Internet of Things according to claim 1, characterized in that, The process of preprocessing separately and then associating materials by batch includes: The limestone calcination operation data is parsed, timestamps are unified, equipment numbers are matched, duplicates are removed, anomalies are corrected, and missing data is filled to obtain calcination operation preprocessing data. The finished product quality feedback data is analyzed for indicators, formatted, results standardized, batch matched and time aligned to obtain preprocessed quality feedback data. Based on the material batch information, by establishing a unified batch identifier, the calcination operation pretreatment data and the quality feedback pretreatment data are matched and associated, summarized and organized, and standardized calcination status data is generated.

3. The intelligent control method for limestone calcination based on the Internet of Things according to claim 2, characterized in that, The process of generating the normal operating temperature segment and the normal temperature constraint identifier includes: Extract the calcination temperature records of the same batch of materials, sort them according to the time of collection to form a temperature sequence, compare the temperature sequence with the preset target temperature range, and mark the low temperature, normal and high temperature operating condition records. Continuous normal temperature records are bound together, and records shorter than a preset threshold are removed to form normal temperature operating condition segments. Write the material batch, start and end times, and temperature matching results to the temperature normal operating condition chip, and generate a temperature normal constraint identifier.

4. The intelligent control method for limestone calcination based on the Internet of Things according to claim 3, characterized in that, The process of generating the calcination process record includes: The standardized calcination state data within the normal temperature operating condition segment are aligned according to material batch and acquisition time, and the changes in operating state at adjacent time points are extracted. Based on changes in fuel supply, the thermal response relationship is established by correlating changes in calcination section temperature and flue gas temperature within the lag interval. Based on changes in combustion air volume, and by correlating changes in kiln pressure and flue gas oxygen content, an air-oxygen matching relationship is established. Based on the feeding and discharging status, the material residence area is divided and the material residence relationship is established by combining the finished product quality feedback; The three relationships are classified, identified, and merged to generate a calcination process record.

5. The intelligent control method for limestone calcination based on the Internet of Things according to claim 4, characterized in that, The improved TTMs model also includes a temperature normal constraint control unit and a deviation direction output unit. The process of generating underburned deviation records and overburned deviation records includes: Determine the heat response path record and the wind-oxygen matching path record that have a time correspondence with the material residence interval; Compare the fuel supply changes in the heat response path record with the path response items to mark the insufficient heat response state; Compare the changes in combustion air volume in the wind-oxygen matching path record with the path response items to mark abnormal wind-oxygen matching status. Compare the material dwell area and finished product quality feedback status with the preset material dwell range and preset quality target, and mark the material dwell status and quality feedback status; When insufficient heat response, insufficient material residence and quality deviation exist simultaneously in the same material residence range, determine the direction of underburning deviation and generate an underburning deviation record; When the same material residence area simultaneously exhibits abnormal air-oxygen matching, excessively long material residence time, and quality deviation, the direction of overburning deviation is determined, and an overburning deviation record is generated.

6. The intelligent control method for limestone calcination based on the Internet of Things according to claim 5, characterized in that, The process of identifying latent underburning risks, localized overburning risks, and risk source identifiers includes: A continuity check is performed on the underburning deviation records, and underburning deviation records that are collected continuously and correspond to the same material residence interval are merged into hidden underburning risks. Based on the insufficient heat response state and insufficient material retention state corresponding to the hidden underburning risk, determine the source identifier of insufficient heat response and the source identifier of insufficient material retention; The overheating deviation records are checked for continuity. Overheating deviation records that are collected continuously and correspond to the same material residence range are merged into local overheating risk. Based on the abnormal state of air-oxygen matching and the state of excessive material residence time corresponding to the risk of local overburning, determine the source identifier of abnormal air-oxygen matching and the source identifier of excessive material residence time; When both latent underburning risk and local overburning risk are generated within the same material residence area, the latent underburning risk, local overburning risk and corresponding risk source identifiers are retained respectively.

7. The intelligent control method for limestone calcination based on the Internet of Things according to claim 6, characterized in that, The process of generating the control parameter correction amount includes: Determine the control cycle to be corrected corresponding to the material residence range; The changes in fuel supply, heat response, material residence time, combustion air volume, kiln pressure, and flue gas oxygen content are read from the control cycle to be corrected. When there is an indication of insufficient heat response, the fuel supply correction amount is generated based on the difference between the preset heat response amplitude and the actual heat response amplitude. When there are sources indicating insufficient or excessive material dwell time, the feed rate correction amount or discharge rate correction amount is generated based on the difference between the preset material dwell range and the actual dwell length. When there is an abnormal source of air-oxygen matching, the combustion air volume correction and kiln pressure correction are generated based on the kiln pressure and flue gas oxygen content deviation. Each control correction is limited and merged to generate control parameter correction values.

8. The intelligent control method for limestone calcination based on the Internet of Things according to claim 7, characterized in that, The process of identifying the source of calibration risk and correcting control parameters includes: By comparing the control parameter correction with the actual execution value, the equipment execution deviation is obtained, the control parameter correction is calibrated, and the execution calibration correction is obtained. The heat response, kiln pressure change, and finished product quality status are compared with the corresponding preset targets to generate calcination state convergence results. When the convergence result of the calcination state meets the preset convergence condition, the corresponding risk source identifier is retained, and the calibration correction amount is processed according to the first amplitude coefficient. When the convergence result of the calcination state does not meet the preset convergence conditions, the corresponding risk source identifier is retained, and the calibration correction amount is performed according to the second amplitude coefficient. When the convergence result of the calcination state does not meet the preset consistency condition with the risk source identifier, the risk source identifier is corrected, and the calibration correction amount is re-determined based on the corrected risk source identifier. The current limestone calcination control parameters, the processed execution calibration correction, and the corrected risk source identifier are combined to generate the limestone calcination control parameters for the next control cycle.

9. An IoT-based intelligent control system for limestone calcination, applied to the IoT-based intelligent control method for limestone calcination as described in any one of claims 1 to 8, characterized in that, Includes the following modules: The data acquisition and standardization module is used to collect calcination operation data and finished product quality data, preprocess and associate them by batch to generate standardized calcination status data; The temperature condition determination module is used to determine the temperature condition of the calcination section and generate normal temperature segments and constraint indicators. The calcination process record generation module is used to extract heat response, air-oxygen matching and material residence records to generate calcination process records; The branch calcination record generation module is used to input the calcination process record into the improved TTMs model, establish processing paths according to categories, and generate branch calcination records. The deviation direction determination module is used to compare the branch records with the temperature constraint identifier to determine the deviation direction of under-burning and over-burning, and generate corresponding records. The risk identification module is used to identify hidden under-burning risks, local over-burning risks, and the sources of these risks. The control correction module is used to generate control parameter correction values ​​and send them to the execution device; The feedback calibration module is used to calibrate the sources of risk and the correction amount of control parameters based on the execution feedback data, and generate the control parameters for the next cycle.

Images