Underwater pellet cutter circulating water tank liquid level temperature linkage control system and method

Abstract

The application discloses an underwater pellet cutting machine circulating water tank liquid level temperature linkage control system and method, and belongs to the technical field of temperature control. The method specifically comprises the following steps: collecting liquid level parameters and water temperature parameters in a circulating water tank in real time, constructing a current operation state of the circulating water tank based on a liquid level change rate, a temperature change rate and a coupling deviation value of the liquid level and the temperature, performing hierarchical judgment on the circulating water tank according to the current operation state, and generating corresponding linkage control targets for different partitions. When it is judged that the liquid level is a dominant intervention area or the temperature is a dominant intervention area, linkage control of priority of a main variable and compensation of a secondary variable is performed. When it is judged that the liquid level and the temperature are in a conflict area, combined control of pre-inhibition and correction based on a change trend is performed. The application reduces mutual interference between different adjustment processes, and improves the overall coordination degree and adaptability in the operation process of the circulating water tank.

Description

Technical Field

[0001] This invention belongs to the field of temperature control technology, specifically the underwater pelletizer circulating water tank liquid level and temperature linkage control system and method. Background Technology

[0002] Underwater pelletizers are widely used in polymer granulation processes, cutting and cooling molten material in an aqueous medium to form granular products. The circulating water tank, a crucial component of this system, performs multiple functions including cooling, conveying, and particle separation. Its internal water temperature and level directly affect pelletizing stability and subsequent process connections. During actual operation, the circulating water continuously absorbs heat and participates in material exchange, resulting in dynamic changes in both liquid level and temperature, with a certain degree of coupling between the two.

[0003] In existing technologies, the control of circulating water tanks typically employs independent methods for level control and temperature control. For example, level maintenance is achieved through a level sensor in conjunction with water replenishment or drainage, while temperature control is achieved through heat exchange devices or cooling water regulation. However, such control methods often rely on a single parameter threshold and lack a unified mechanism for handling the correlation between level and temperature changes. When level changes affect the temperature distribution or temperature fluctuations, in turn, affect the level state, problems such as asynchronous control responses or mutual interference in the adjustment process can easily occur. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a liquid level and temperature linkage control system and method for the circulating water tank of an underwater pelletizer. This method combines the characteristics of coordinated changes in liquid level and temperature to uniformly determine the operating status and implement linkage control, thereby improving the consistency and coordination of temperature regulation logic during the operation of the circulating water tank.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for linking the liquid level and temperature of the circulating water tank in an underwater pelletizer includes:

[0007] Real-time acquisition of liquid level and water temperature parameters in the circulating water tank, and construction of the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range.

[0008] The circulating water tank is divided into four zones based on the current operating status: steady-state coordination zone, liquid level-dominated intervention zone, temperature-dominated intervention zone, and liquid level-temperature conflict zone. Corresponding linkage control targets are generated for each zone. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level and water temperature parameters.

[0009] When the area is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, a linkage control with priority given to the main variable and compensation for the secondary variable is implemented. When the area is determined to be a liquid level-temperature conflict zone, a joint control based on the trend of change is implemented with pre-suppression and correction.

[0010] During the liquid level and temperature regression process, the coupling deviation value is continuously updated, and the water replenishment, drainage and heat exchange intensity are adaptively redistributed according to the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, the circulating water tank is kept in a disturbance collaborative control state within the threshold range.

[0011] Specifically, the real-time acquisition of liquid level and temperature parameters within the circulating water tank, and the construction of the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and the coupling deviation between liquid level and temperature, includes:

[0012] Within a preset acquisition period, the liquid level and water temperature parameters of the circulating water tank at corresponding times are acquired, and liquid level time series and temperature time series are formed in chronological order.

[0013] The liquid level time series and temperature time series are aligned according to the same decision window, isolated collected values ​​that do not fall into the decision window are removed, and the aligned liquid level time series and temperature time series are paired as homogeneous state segments.

[0014] Based on each homogeneous state segment, the direction of change, continuity of change, and stage transition position of the liquid level parameter, as well as the direction of change, continuity of change, and stage transition position of the water temperature parameter are identified in sequence, and the liquid level evolution segment corresponding to the liquid level change rate and the temperature evolution segment corresponding to the temperature change rate are determined accordingly.

[0015] The liquid level evolution segment and the temperature evolution segment are sequentially correlated. Liquid level-temperature correlation segments are generated according to the sequential triggering relationship, synchronous change relationship and reverse restraint relationship. Based on the deviation sequence, deviation duration and deviation overlap of each correlation segment relative to the preset cooperative working interval, the coupling deviation value of liquid level and temperature is generated.

[0016] The current operating state of the circulating water tank is calibrated based on the liquid level evolution segment, the temperature evolution segment, and the coupling deviation value. The current operating state includes one of the following: liquid level deviation state, temperature deviation state, liquid level and temperature synchronous deviation state, and liquid level and temperature alternating deviation state.

[0017] Specifically, the liquid level evolution segment and the temperature evolution segment are sequentially correlated to generate a liquid level-temperature correlation segment. Based on the order of deviation, duration of deviation, and overlap of deviations of each correlation segment relative to a preset collaborative working interval, a coupling deviation value between liquid level and temperature is generated, including:

[0018] The liquid level evolution segment and the temperature evolution segment are rearranged on a unified time axis, and continuous time slice units are formed with the start time and end time of each evolution segment as the boundary. A candidate association set is established for liquid level evolution segments and temperature evolution segments that fall into the same time slice unit.

[0019] For each candidate association set, based on the order of triggering of the liquid level evolution segment and the temperature evolution segment, the overlap of the start and end of the change, and the consistency of the change direction, it is determined that it belongs to one of the following: sequential triggering relationship, synchronous change relationship, or reverse restraint relationship, and each candidate association set is assigned a corresponding relationship identifier.

[0020] Based on the relationship identifier, continuous time slice units are merged or divided to construct liquid level-temperature associated segments. Adjacent time slice units with consistent relationship identifiers are merged into the same associated segment, and time slice units with switched relationship identifiers are used as segment boundaries for segmentation.

[0021] Each associated segment is serialized according to its entry order, dwell time, and overlapping range relative to the preset collaborative working area, forming a coupling deviation value that includes deviations from the order, deviations from the duration, and deviations from the overlapping segments.

[0022] Specifically, based on the liquid level evolution segment, the temperature evolution segment, and the coupling deviation value, the current operating state of the circulating water tank is calibrated, including:

[0023] The liquid level evolution segment and the temperature evolution segment are cross-arranged according to the corresponding time slice units to form a state sequence unit composed of the interwoven liquid level evolution trajectory and temperature evolution trajectory, and the time continuity between each state sequence unit is maintained.

[0024] The coupling deviation value is introduced into each state sequence unit, and a deviation attribute is assigned to each state sequence unit. The deviation attribute includes a deviation start identifier, a deviation continuation identifier, and a deviation overlap identifier, forming a composite state segment with deviation attributes.

[0025] The composite state segments are merged according to a preset state mapping rule. Composite state segments with the same combination of deviation attributes and continuous in time are merged into state candidate segments, and each state candidate segment is assigned a corresponding running state identifier category.

[0026] Priority screening and boundary determination are performed among the candidate state segments. The candidate state segments that meet the priority determination conditions are determined as the current operating state of the circulating water tank, and the unselected candidate state segments are retained as reference segments for subsequent state evolution.

[0027] Specifically, based on the current operating state, the circulating water tank is divided into four zones: a steady-state coordination zone, a level-dominated intervention zone, a temperature-dominated intervention zone, and a level-temperature conflict zone. Corresponding linkage control targets are generated for each zone, including:

[0028] The current operating state is deconstructed into liquid level state components and temperature state components, and the corresponding change direction identifier, change stage identifier, and deviation attribute identifier are extracted to form a basic set for state determination.

[0029] Based on the aforementioned state determination base set, the dominant attributes of liquid level and temperature are determined according to the prior relationship, continuity relationship, and alternation relationship between the liquid level state component and the temperature state component, resulting in one of the following: liquid level dominant identifier, temperature dominant identifier, or no dominant identifier.

[0030] Based on the dominant attribute identifier and the corresponding deviation attribute identifier, the operating status of the circulating water tank is divided into zones. The state that satisfies both the liquid level and temperature within the coordinated range is divided into the steady-state coordinated zone. The state with a single dominant attribute is divided into the corresponding dominant intervention zone. The state with no dominant identifier and intersecting deviation attributes is divided into the liquid level and temperature conflict zone.

[0031] Based on each zone, a linkage control target is constructed by combining liquid level parameters and temperature parameters. In this process, the changing direction pairing, deviation sequence pairing, and deviation superposition pairing in the combination relationship are matched to form the target constraint sequence of the corresponding zone.

[0032] During the control execution process, the linkage control targets of each partition are dynamically revised according to the target constraint sequence. When the change direction indicator or deviation attribute indicator is switched, the corresponding partition is switched and the linkage control targets are updated synchronously.

[0033] Specifically, based on the dominant attribute identifier and the corresponding deviation attribute identifier, the operating status of the circulating water tank is divided into zones, including a steady-state coordination zone, a dominant intervention zone, and a liquid level / temperature conflict zone, including:

[0034] The dominant attribute identifier and the corresponding deviation attribute identifier are aligned and combined in chronological order to form a judgment sequence consisting of dominant type and deviation type pairing;

[0035] The determination sequence is segmented, and the pairing set that maintains the same dominant attribute identifier and extends continuously from the deviation attribute identifier is divided into the same determination segment, and the segment boundary is formed at the position where the dominant attribute identifier or the deviation attribute identifier switches.

[0036] For each judgment segment, it is partitioned and classified according to the combination type of its dominant attribute identifier and deviation attribute identifier to obtain the initial partitioning and classification results;

[0037] Based on the initial partition classification results, the partition boundaries between adjacent judgment segments are calibrated. When there are judgment segments whose boundaries cross a preset time span, they are split or merged. The range of each partition is redefined based on the processed judgment segments to form the partition generation results, including the steady-state coordination zone, the dominant intervention zone, and the liquid level and temperature conflict zone.

[0038] Specifically, based on each zone, a linkage control target formed by a combination of liquid level parameters and temperature parameters is constructed, forming a target constraint sequence for the corresponding zone, including:

[0039] Extract the liquid level parameter sequence and temperature parameter sequence corresponding to each partition, and then split the liquid level parameter sequence and temperature parameter sequence synchronously according to the time segment within the partition to form multiple parameter alignment units;

[0040] The liquid level parameter and temperature parameter in each parameter alignment unit are combined to construct the relationship, and the pairing of change direction, deviation sequence, and deviation superposition are generated respectively. The different types of pairing results are classified and labeled.

[0041] Based on the classification identifier, the parameter alignment units are sequentially connected to form the target generation sequence by connecting parameter alignment units with continuous pairing relationships within the same partition.

[0042] Based on the target generation sequence, the change direction pairing, deviation sequence pairing, and deviation superposition pairing are sequentially embedded into the preset constraint expression structure to form the target constraint sequence of the corresponding partition.

[0043] Specifically, during the control execution process, the linkage control targets of each partition are dynamically revised according to the target constraint sequence. When the change direction indicator or deviation attribute indicator changes, the corresponding partition is switched and the linkage control targets are updated synchronously, including:

[0044] During control execution, the target constraint sequence is read cyclically, and the sequence segment corresponding to the current control moment is used as the execution reference unit, while maintaining the correspondence between the execution reference unit and the current partition;

[0045] Based on the execution reference unit, the change direction identifier and deviation attribute identifier of the liquid level parameter and temperature parameter are continuously acquired and compared with the pairing relationship recorded in the execution reference unit to identify whether there is an identifier switching event.

[0046] When the identifier switching event is detected, the current partition is re-determined based on the changed direction identifier and deviation attribute identifier after the switch, and the original execution reference unit is terminated at the switching time. At the same time, the sequence start unit corresponding to the new partition is generated.

[0047] For the sequence starting unit of the new partition, reconstruct the target constraint sequence corresponding to it.

[0048] Specifically, when the system is determined to be in a liquid level-dominated intervention zone or a temperature-dominated intervention zone, a linkage control system prioritizing the main variable and compensating for the secondary variable is implemented. When the system is determined to be in a liquid level-temperature conflict zone, a joint control system based on trend-based pre-suppression and correction is implemented, including:

[0049] Read the current partition identifier. When it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, determine the corresponding main and secondary variables, and extract the change direction identifier and deviation attribute identifier of the main and secondary variables at the current moment.

[0050] Within the dominant intervention zone, a main variable control sequence is generated based on the change direction identifier and deviation attribute identifier of the main variable. During the advancement of the main variable control sequence, a compensation sequence corresponding to the secondary variable is generated synchronously, so that the change process of the secondary variable maintains a pairing relationship with the main variable control sequence.

[0051] When a liquid level and temperature conflict zone is identified, a change trend sequence is constructed based on the change order of liquid level parameters and temperature parameters in historical time slices and the current change direction identifier. A pre-suppression sequence is then generated based on the change trend sequence to intervene in the parameter change path that has entered the zone.

[0052] After the pre-suppression sequence is executed, the trend sequence is updated by combining the currently acquired liquid level and temperature parameters, and a correction sequence is generated based on the updated trend sequence.

[0053] The underwater pelletizer circulating water tank liquid level and temperature linkage control system is used to realize the underwater pelletizer circulating water tank liquid level and temperature linkage control method, including: a data acquisition module, a control target generation module, a linkage control module and an update control module;

[0054] The data acquisition module is used to collect the liquid level parameters and water temperature parameters in the circulating water tank in real time, and to construct the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range.

[0055] The control target generation module is used to classify the circulating water tank into layers based on the current operating state, dividing it into a steady-state coordination zone, a liquid level-dominated intervention zone, a temperature-dominated intervention zone, and a liquid level-temperature conflict zone, and generating corresponding linkage control targets for different zones. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level parameters and water temperature parameters.

[0056] The linkage control module is used to perform linkage control with priority to main variables and compensation for secondary variables when it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, and to perform joint control based on the trend of change for pre-suppression and correction when it is determined to be a liquid level-temperature conflict zone.

[0057] The update control module is used to continuously update the coupling deviation value during the liquid level and temperature regression process, and adaptively redistribute the water replenishment, drainage and heat exchange intensity according to the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, the circulating water tank is kept in a disturbance collaborative control state within the threshold range.

[0058] Compared with the prior art, the beneficial effects of the present invention are:

[0059] This invention proposes a liquid level and temperature linkage control system and method for an underwater pelletizer circulating water tank. By constructing a coupled relationship between liquid level and temperature parameters, it introduces an operating state calibration mechanism based on evolutionary segments and deviation representations, and performs zone-based judgment to form a zone-based linkage control framework. Simultaneously, it combines target constraint sequences to achieve dynamic revision during the control process. In different zones, it executes joint control with priority given to primary variables, compensation for secondary variables, and pre-suppression and correction based on changing trends. This ensures that liquid level and temperature are no longer independent control objects but participate in unified regulation in a collaborative relationship. This method can maintain the consistency and continuity of control logic under various operating conditions, reduce mutual interference between different adjustment processes, and enable the control strategy to switch according to the evolution of the operating state, thereby improving the overall coordination and adaptability of the circulating water tank during operation. Attached Figure Description

[0060] Figure 1 Flowchart of the underwater pelletizer circulating water tank level and temperature linkage control method provided by the present invention;

[0061] Figure 2 This is a schematic diagram of temporal correlation and segmentation provided by the present invention;

[0062] Figure 3 This is a schematic diagram of the zoning determination of the operating status of the circulating water tank provided by the present invention;

[0063] Figure 4 The architecture diagram of the underwater pelletizer circulating water tank liquid level and temperature linkage control system provided by the present invention. Detailed Implementation

[0064] The present application will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present application, but do not limit the present application in any way. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present application. These all fall within the protection scope of the present application.

[0065] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0066] It should be noted that, unless there is a conflict, the various features in the embodiments of this application can be combined with each other, all of which are within the protection scope of this application. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. In addition, the terms "first," "second," and "third" used in this application do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0067] Unless otherwise defined, all technical and scientific terms used in this specification have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The term "and / or" as used in this specification includes any and all combinations of one or more of the associated listed items.

[0068] Example 1

[0069] Please see Figures 1-3 The present invention provides an embodiment of a method for linkage control of liquid level and temperature in the circulating water tank of an underwater pelletizer, comprising the following specific steps:

[0070] Step S1: Real-time acquisition of liquid level and water temperature parameters in the circulating water tank, and construction of the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range.

[0071] like Figure 2 As shown, the specific steps of step S1 are as follows:

[0072] Step S101: Within the preset acquisition period, acquire the liquid level parameters and water temperature parameters of the circulating water tank at the corresponding time, and form the liquid level time series and temperature time series in chronological order.

[0073] In this embodiment, the preset acquisition period is a continuous acquisition time window containing several equally spaced sampling moments. Based on the operating fluctuation characteristics of the circulating water tank and the control accuracy requirements of the liquid level and water temperature parameters, the acquisition period is set to a time interval of 50ms-200ms for each sampling moment, with a total acquisition duration of no less than 30s. Specifically, based on the static pressure level transmitter installed at the static pressure measuring port at the bottom of the circulating water tank and the Pt100 platinum resistance temperature sensor inserted into the effective heat exchange area of ​​the circulating water, at each sampling trigger moment within the preset acquisition period, the real-time analog liquid level signal and the real-time analog water temperature signal in the circulating water tank are synchronously acquired. The two analog signals are synchronously converted into corresponding digital format liquid level and water temperature parameters, which are then simultaneously added to the liquid level and water temperature parameters acquired at the same sampling trigger moment, uniquely paired with the parameters acquired at that trigger moment. The system uses a timestamp format of year-month-day-hour-minute-second-millisecond to collect and calibrate the time dimension of single-sampling point data. After collecting and calibrating all sampling points within the preset collection period, the liquid level and water temperature parameters of all sampling points are preprocessed using a first-order hysteresis filtering algorithm to filter out sampling spike noise and random interference caused by water flow fluctuations in the circulating water tank and vibrations from the operation of the supporting circulating water pump. Then, the preprocessed liquid level parameters are sequentially spliced ​​together according to the ascending time order of the timestamps to form a continuous and complete liquid level time series segment with a continuous time axis. Simultaneously, the preprocessed water temperature parameters, which correspond one-to-one with the timestamps of each liquid level parameter in the liquid level time series segment, are sequentially spliced ​​together according to the time axis order completely consistent with the liquid level time series segment to form a temperature time series segment that is completely aligned with the time dimension of the liquid level time series segment and corresponds one-to-one with the sampling points.

[0074] Step S102: The liquid level time series and the temperature time series are aligned according to the same judgment time window, isolated collected values ​​that do not fall into the judgment time window are removed, and the aligned liquid level time series and temperature time series are paired as homogeneous state segments.

[0075] In this embodiment, the determination window is a continuous time interval with fixed start and end boundaries and a threshold for the number of effective sampling points, preset based on the operating condition response characteristics of the coupled changes in the liquid level and water temperature of the circulating water tank. Specifically, based on the unified timestamp reference of the liquid level time series and the temperature time series obtained in step S101, the earliest sampling time that is common to both time series and whose timestamps completely overlap is taken as the starting boundary of the determination window, and the latest sampling time that is common to both time series and whose timestamps completely overlap is taken as the ending boundary of the determination window. At the same time, the minimum threshold for the number of effective sampling points within the determination window is preset to be no less than 20, thereby determining the same determination window that is completely consistent with the time reference of the two time series.

[0076] For feasible implementation, firstly, iterate through all collected values ​​with unique timestamps within the liquid level and temperature time segments, comparing each collected value's timestamp against the defined start and end boundaries of the judgment window. Collected values ​​whose timestamps do not fall within the judgment window's start and end boundaries, as well as single-channel collected values ​​without corresponding paired sampling points due to abnormal acquisition triggering, momentary sensor disconnection, or analog-to-digital conversion errors, are uniformly marked as isolated collected values. Simultaneously, a removal operation is performed to completely remove all marked isolated collected values ​​from their corresponding time segments. Then, the removed values ​​are processed... After isolated data collection, the liquid level and temperature time series segments are rearranged in ascending order according to the timestamp rules to complete the standardization and straightening of the two time series segments. Then, based on the same time axis reference of the judgment window, each unique timestamp is used as a pairing identifier to bind and pair the straightened liquid level parameters and water temperature parameters corresponding to the same timestamp one by one. Finally, all liquid level parameter sequences and water temperature parameter sequences are combined to form a homogeneous state segment of the circulating water tank operating status at the same data collection time.

[0077] Step S103: Based on each homogeneous state segment, sequentially identify the direction of change, continuity of change, and stage transition position of the liquid level parameter, as well as the direction of change, continuity of change, and stage transition position of the water temperature parameter, and determine the liquid level evolution segment corresponding to the liquid level change rate and the temperature evolution segment corresponding to the temperature change rate accordingly.

[0078] In this embodiment, the homogeneous state segments obtained in step S102, with perfectly aligned time axes and paired sampling points, are used as the processing objects. First, the liquid level parameter sequence and water temperature parameter sequence bound and paired within the homogeneous state segments are extracted. For the two parameter sequences, the same temporal feature recognition logic is used to sequentially complete the full recognition of the direction of change, continuity of change, and stage transition position. Then, the corresponding evolution segments are divided based on the recognition results. Specifically, for a single parameter sequence, a sliding difference window with a width of 3 consecutive sampling points and a step size of 1 sampling point is preset. The window is traversed point by point along the time axis of the sequence. The ratio of the parameter difference between the first and last sampling points in each sliding window to the corresponding time interval is calculated to obtain the instantaneous rate of change corresponding to the window. At the same time, the direction of change of the parameter in the window is determined according to the sign of the parameter difference and the preset dead zone threshold. The dead zone threshold of the liquid level parameter is preset to 0 of the rated full-scale liquid level of the circulating water tank. The dead zone threshold for the water temperature parameter is preset to 0.1℃. When the parameter difference is positive and its absolute value is greater than the dead zone threshold, it is determined to be an upward direction; when the parameter difference is negative and its absolute value is greater than the dead zone threshold, it is determined to be a downward direction; when the absolute value of the parameter difference is less than or equal to the dead zone threshold, it is determined to be a stable direction. After completing the identification of the change direction of the entire sequence, for adjacent continuous sliding windows with the same change direction, the continuity of parameter change is identified by continuity verification logic. The allowable fluctuation threshold for the instantaneous change rate of the liquid level parameter is preset to ±15%, and the allowable fluctuation threshold for the instantaneous change rate of the water temperature parameter is ±10%. The relative difference of the instantaneous change rate of adjacent windows in the same direction is calculated one by one. If the relative difference falls within the corresponding allowable fluctuation threshold range, it is determined to be a continuous change; if it exceeds the threshold, it is marked as a continuity interruption point. Subsequently, based on the instantaneous change rate sequence of the entire sequence, the second-order difference method combined with the Pettitt mutation point test method is used to identify the stage transition position of the parameter.

[0079] In a feasible manner, after fully identifying the change direction, continuity, and stage transition positions of the liquid level parameter sequence and the water temperature parameter sequence, the liquid level parameter sequence and the water temperature parameter sequence within the same source state segment are divided into intervals using the stage transition positions determined by each of the two parameter sequences as segmentation nodes. Each continuous subsequence obtained after segmentation must meet the following constraints: the number of sampling points is not less than 5, the change direction of all sampling points in the subsequence is consistent, and the continuity of change between adjacent sampling points meets the preset threshold requirements. For example, the instantaneous change rate of adjacent sampling points must be within the allowable fluctuation range (liquid level ±15%, water temperature ±10%), and the change amplitude must exceed the dead zone threshold to be determined as rising or falling. Then, the least squares method is used to perform linear fitting on each subsequence that meets the constraints to obtain the fitting slope of the parameter change over time in the subsequence, which is the stage change rate corresponding to the subsequence. Finally, each continuous subsequence of liquid level parameter is determined as the liquid level evolution segment corresponding to the liquid level change rate, and each continuous subsequence of water temperature parameter is simultaneously determined as the temperature evolution segment corresponding to the temperature change rate.

[0080] like Figure 2 As shown, the evolution of the liquid level (blue curve) and temperature (red curve) in the circulating water tank over a period of time is illustrated, along with the different related segments divided by analyzing the evolution trends of liquid level and temperature. The horizontal axis represents time, and the vertical axis is divided into two parts: the upper blue part represents the liquid level, and the lower red part represents the temperature. The dashed lines in the figure divide the entire time series into 6 segments: A, B, C, D, E, and F, each segment corresponding to a different liquid level-temperature evolution relationship.

[0081] A: First, the liquid level changes (corresponding to time period t0-t1), the liquid level drops, the temperature changes slowly, the liquid level deviates from the normal operating condition first, and the temperature has not yet responded; B: First, the temperature changes (t1-t2), the liquid level is stable, the temperature shows a clear downward trend, the temperature deviates from the normal operating condition first, and the liquid level is relatively stable; C: First, the temperature changes (t2-t3), the liquid level rises, the temperature rises, the liquid level and temperature rise simultaneously, the time intervals highly overlap and the direction of change is consistent, which is determined to be a synchronous change relationship; D: First, the temperature changes (t3-t4), the liquid level remains high, the temperature remains high, the liquid level and temperature are stable during the plateau period; E: First, the temperature changes (t5-t6), the liquid level drops, the temperature rises, the direction of change of liquid level and temperature is opposite; F: First, the temperature changes (t6-t7), the liquid level slowly recovers, and the temperature remains balanced.

[0082] Step S104: The liquid level evolution segment and the temperature evolution segment are sequentially associated. Liquid level-temperature associated segments are generated according to the sequential triggering relationship, synchronous change relationship and reverse restraint relationship. Based on the order of deviation, duration of deviation and overlap of deviation of each associated segment relative to the preset cooperative working interval, the coupling deviation value of liquid level and temperature is generated.

[0083] In this embodiment, all liquid level evolution segments and temperature evolution segments obtained in step S103, which have unique time intervals, determined change directions, and corresponding stage change rates, are taken as processing objects. First, a unified time axis rearrangement operation is performed on the two evolution segments. Specifically, high-precision timestamps of the start and end times of each of the liquid level evolution segments and temperature evolution segments are extracted. All extracted timestamps are sorted according to ascending order to remove duplicates. Using the two adjacent timestamps after sorting as the start and end boundaries, several continuous time slice units with continuous time length, no overlap, and no time gap are formed. Each continuous time slice unit has a unique and indivisible time interval, and there are no start and end time nodes of any liquid level evolution segment or temperature evolution segment within the time interval of any continuous time slice unit. This completes the time axis standardization segmentation of the two evolution segments under the same time base.

[0084] In practice, after completing the division of continuous time slice units, each continuous time slice unit is traversed, and the inclusion relationship between the time intervals of all liquid level evolution segments and temperature evolution segments and the time interval of the time slice unit is compared one by one. The liquid level evolution segments and temperature evolution segments of the time slice unit are completely covered by the time interval, and all of them are collected into the set corresponding to the time slice unit, forming the candidate association set corresponding to the time slice unit. This ensures that the liquid level evolution segments and temperature evolution segments in each candidate association set completely fall within the time interval of the same continuous time slice unit and have a completely homogeneous time coverage range.

[0085] For each candidate association set, the trigger start time, end time, direction of change, and stage change rate of all liquid level evolution segments within the set are extracted, along with the trigger start time, end time, direction of change, and stage change rate of the corresponding temperature evolution segment. Existing time series overlap calculation methods are used to calculate the time interval overlap between the liquid level and temperature evolution segments within the set. The overlap is calculated as the ratio of the intersection duration to the union duration of the two segments' time intervals. Simultaneously, the absolute time difference between the start and end times of the two segments, as well as the consistency of their direction of change, are calculated. The preset synchronous change judgment thresholds are: overlap ≥ 95%, absolute time difference between the start and end times ≤ 2 sampling intervals, consistent direction of change, and the rate of change fluctuation within a preset coordinated fluctuation range. The intervals are trained or calibrated based on the rated operating conditions and historical normal operation data of the circulating water tank. Candidate association sets that meet the threshold conditions are determined to be synchronous change relationships. The preset conditions for determining sequential triggering relationships are that the two segments change in the same direction, the overlap of the time intervals is within the range of 30%-95%, and the absolute time difference between the start times of the two segments is greater than 2 sampling intervals. In addition, the start time and end time of one segment are both earlier than the corresponding time of the other segment. Candidate association sets that meet the conditions are determined to be sequential triggering relationships. The preset conditions for determining reverse restraint relationships are that the two segments change in completely opposite directions and the overlap of the time intervals is ≥30%. Candidate association sets that meet the conditions are determined to be reverse restraint relationships. After the relationship determination is completed, each candidate association set is assigned a relationship identifier that uniquely corresponds to the determination result.

[0086] After assigning relation identifiers to all candidate association sets, merge or segment operations are performed on all continuous time slice units based on the relation identifiers to construct liquid level-temperature association segments. Specifically, all continuous time slice units are traversed in ascending order of the time axis, and the relation identifiers corresponding to two adjacent continuous time slice units are compared. If the relation identifiers of two adjacent continuous time slice units are completely identical, the time intervals of the two continuous time slice units are merged. The merged time interval uses the earliest start time of the two continuous time slice units as the start time of the new segment and the latest end time of the new segment as the end time of the new segment. At the same time, the candidate association sets corresponding to the two continuous time slice units are aggregated to form a merged segment. If the relationship identifiers of two adjacent consecutive time slice units are inconsistent, the adjacent boundary of the two consecutive time slice units is marked as the segmentation boundary of the liquid level-temperature associated segment. The adjacent time slice units are divided using this boundary as a node to ensure that the relationship identifiers of all consecutive time slice units in each segment are completely unified. The final liquid level-temperature associated segment has a preset collaborative working interval that is a working condition interval that is pre-trained using an existing Gaussian mixture model based on the rated working conditions and historical normal operating data of the circulating water tank. This interval includes the allowable range of liquid level change rate, the allowable range of temperature change rate, and the synchronous change relationship identifiers corresponding to normal operation.

[0087] After constructing the liquid level-temperature correlation segment, all liquid level-temperature correlation segments are traversed. For each segment, the relationship identifier, the liquid level change rate of the liquid level evolution segment, and the temperature change rate of the temperature evolution segment are determined to see if they fall within a preset collaborative working interval. Segments that do not fall within the preset collaborative working interval are marked as deviation segments. Then, all deviation segments are serialized and numbered according to their ascending entry order on the time axis, with the numbering order representing the order of deviation. For each deviation segment, its continuous time interval is extracted, and the duration of that time interval, the deviation value of the liquid level change rate within the segment, and the deviation value of the temperature change rate are recorded. A unified relation identifier is used to integrate and form the deviation duration segment corresponding to the deviation section. The deviation duration segment contains full information such as the deviation start time, deviation end time, total deviation duration, parameter deviation amplitude, and corresponding relation identifier. For multiple deviation duration segments with overlapping time intervals, the intersection of their time intervals is extracted. The time range corresponding to the intersection, the number of overlapping deviation duration segments, the parameter deviation characteristics of each overlapping segment, and the relation identifier type are integrated to form the deviation overlap segment. Finally, the sequence of all deviations is structurally integrated to form a quantifiable, traceable, and fully time-series-characteristic coupled deviation value of liquid level and temperature.

[0088] like Figure 3As shown, step S105: Based on the liquid level evolution segment, the temperature evolution segment, and the coupling deviation value, the current operating state of the circulating water tank is calibrated. The current operating state includes one of the following: liquid level deviation state, temperature deviation state, liquid level and temperature synchronous deviation state, and liquid level and temperature alternating deviation state.

[0089] In this embodiment, the liquid level evolution segment and temperature evolution segment obtained in step S103, which have unique time intervals, determined change directions, and corresponding stage change rates, and the coupling deviation value generated in step S104, which includes deviations from the order of events, deviations from the duration of events, and deviations from overlapping events, are processed as the processing objects. First, the continuous time slice unit that has completed the unified time axis rearrangement in step S104 is taken as the smallest time unit, and the cross-arrangement operation of the liquid level evolution segment and temperature evolution segment is performed. Specifically, the full trajectory data of the liquid level evolution segment and temperature evolution segment covered by each continuous time slice unit is extracted first. The liquid level evolution trajectory and temperature evolution trajectory are both characterized by unique timestamps, measured parameter values, fitting stage change rates, and change directions. The identified discrete trajectory points are arranged in ascending order of time. Based on the unified time axis of the continuous time slice unit, and with the timestamp of each sampling moment as the unique index, the liquid level trajectory points and temperature trajectory points corresponding to the same sampling moment are bound one by one to form the working condition state tuple of a single sampling moment. Then, all working condition state tuples in the time slice unit are arranged continuously in ascending order of timestamp to complete the cross arrangement of liquid level evolution trajectory and temperature evolution trajectory, forming a state sequence unit that completely matches the time interval of the time slice unit. At the same time, it is ensured that the end time of the previous unit of the adjacent state sequence unit and the start time of the next unit completely coincide, with no time gaps and no interval overlap, and the time continuity between each state sequence unit is maintained throughout the process.

[0090] In practice, after completing the construction of all state sequence units, the aforementioned coupling deviation values ​​are introduced into each state sequence unit according to the time axis matching rules. First, based on the existing time interval topology comparison algorithm, the topological relationship between the time interval of each state sequence unit and the time intervals of each deviation duration segment and deviation overlap segment in the coupling deviation value is verified one by one. The preset deviation attribute assignment rules are as follows: when the start time of the time interval of the state sequence unit completely coincides with the start time of a certain deviation duration segment, and the preceding adjacent state sequence units of the unit have no deviation-type attribute assignment, the state sequence unit is assigned a deviation start mark; when the time interval of the state sequence unit completely falls within the time interval of a certain deviation duration segment, and the unit neither meets the deviation start mark assignment condition nor has 100% intersection coverage with the time interval of any deviation overlap segment, the state sequence unit is assigned a deviation continuation mark.

[0091] When the time interval of a state sequence unit has 100% overlap with the time interval of a certain deviation overlap segment, the state sequence unit is assigned a deviation overlap identifier. A single state sequence unit can superimpose multiple deviation attributes based on the topological coverage relationship of its time interval. After the deviation attributes of all state sequence units are assigned, each state sequence unit with complete evolution trajectory data and corresponding deviation attributes forms a composite state segment. Each composite state segment retains a unique time interval, full working condition state tuple data, and corresponding deviation attribute combination.

[0092] Specifically, a pre-defined state mapping rule is constructed based on the rated operating conditions of the circulating water tank and a historical fault case library. This rule adopts the logic of an existing rule-based expert system and predefines a one-to-one mapping relationship between different combinations of deviation attributes and the categories of circulating water tank operating status indicators. This includes, but is not limited to: combinations with no deviation attributes are mapped to normal operating status indicators; combinations with only deviation start indicators are mapped to operating condition change start indicators; combinations with only deviation continuation indicators are mapped to single-parameter abnormality continuation status indicators; combinations with both deviation continuation and deviation overlap indicators are mapped to coupling abnormality aggravation status indicators; and combinations with both deviation start and deviation overlap indicators are mapped to sudden coupling fault status indicators. After generating composite state segments, all composite state segments are traversed in ascending order along the time axis. For each composite state segment, the deviation attribute combination of adjacent composite state segments is compared to see if they are completely consistent. At the same time, it is verified whether the time interval of adjacent composite state segments meets the requirement of complete time continuity. Adjacent composite state segments that simultaneously meet the requirements of completely consistent deviation attribute combination and time continuity are merged into the same state candidate segment. The time interval of the merged state candidate segment is defined by the earliest start time of the merged composite state segment as the segment start time and the latest end time as the segment end time. At the same time, the full trajectory data and deviation attribute information of all composite state segments within the merged range are completely preserved. Then, according to the above-mentioned preset state mapping rules, each state candidate segment is assigned a unique operating state identifier category corresponding to its deviation attribute combination, ensuring that each state candidate segment has a unique operating condition orientation.

[0093] The system is feasible and pre-defined with priority judgment conditions based on the risk level classification of circulating water tank operating conditions. The priority is sorted from high to low fault risk as follows: sudden coupling fault status identifier > coupling anomaly aggravation status identifier > single parameter anomaly persistence status identifier > operating condition change start status identifier > normal operation status identifier. The higher the risk level, the greater the corresponding priority weight value. After completing the construction of the status candidate segments, the priority weight values ​​corresponding to the operating status identifier categories of all status candidate segments are extracted. At the same time, the timestamp corresponding to the current sampling time of the circulating water tank is locked. All status candidate segments whose time interval completely covers the current time are selected as the candidate set to be screened. All status candidate segments in the candidate set to be screened are sorted in descending order of priority weight value from high to low. The status candidate segment with the highest priority weight value in the sorting result is determined. To determine the current operating state of the circulating water tank, boundary determination is performed simultaneously for multiple candidate state segments with overlapping time intervals. The time interval boundary of the candidate state segment with the higher priority weight value is taken as the legal priority boundary. The time intervals of the lower priority candidate state segments are pruned, and the time intervals that overlap with the higher priority candidate state segments are removed. Only the non-overlapping valid time intervals are retained. Finally, all candidate state segments that are not determined to be the current operating state, including the pruned low-priority candidate state segments and historical candidate state segments whose time intervals do not cover the current moment, are used as reference segments for subsequent state evolution. They are structured and stored together with their corresponding full trajectory data, deviation attribute combinations, and operating state identifier categories for subsequent working condition trend prediction, anomaly tracing, and fault early warning model iteration optimization.

[0094] Step S2: Based on the current operating state, the circulating water tank is divided into a steady-state coordination zone, a liquid level-dominated intervention zone, a temperature-dominated intervention zone, and a liquid level-temperature conflict zone. Corresponding linkage control targets are generated for different zones. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level parameters and water temperature parameters.

[0095] The specific steps of step S2 are as follows:

[0096] Step S201: Deconstruct the current operating state, split it into liquid level state component and temperature state component, and extract the corresponding change direction identifier, change stage identifier and deviation attribute identifier respectively to form a basic set for state determination.

[0097] In this embodiment, the candidate state segment corresponding to the current operating state of the circulating water tank determined by priority screening in step S105, which has a unique continuous time interval, full operating condition trajectory data, and a combination of corresponding operating state identifier category and deviation attribute, is taken as the only processing object. First, the parameter dimension deconstruction operation of the current operating state is performed. Specifically, the full structured data of the liquid level evolution trajectory and temperature evolution trajectory with completely aligned time axes and one-to-one pairing sampling points within the candidate state segment are extracted. Based on the decoupling and splitting rules of the parameter physical dimension, the current operating state is completely split into liquid level state component and temperature state component, which are independent of each other but have completely unified time references. The liquid level state component completely retains all timestamped liquid level measured values, liquid level stage change rate, and liquid level evolution segment attribution information within the time interval of the candidate state segment. The temperature state component completely retains all timestamped water temperature measured values, water temperature stage change rate, and temperature evolution segment attribution information within the same time interval, ensuring that the number of sampling points, timestamp sequence, and time interval of the two state components are completely consistent, with no time misalignment or data loss.

[0098] In practice, after splitting the two state components, the corresponding change direction identifier, change stage identifier, and deviation attribute identifier are extracted sequentially for the liquid level state component and the temperature state component according to a unified identifier extraction rule. For the extraction of the change direction identifier, the corresponding evolution segment that perfectly matches the time interval of the state component is first locked, and the change direction identifier bound to the evolution segment and stored in a standardized enumeration format is extracted, including three categories: rising, falling, and stable. The change direction determination of the liquid level state component is based on a preset dead zone threshold of 0.2% of the rated full scale liquid level, and the change direction determination of the temperature state component is based on a preset dead zone threshold of 0.1℃. 0.2% of the rated full scale liquid level is set as the dead zone threshold. If the liquid level change of two consecutive sampling points is less than 0.2% of the full scale change, the liquid level is considered to have no significant rise or fall and is judged as stable. Only when the change is greater than 0.2% is it judged as rising or falling according to the positive or negative direction, ensuring the consistency and traceability of the identifier extraction.

[0099] To extract the change stage identifier, the transition position and complete evolution duration of the corresponding evolution segment are first retrieved through second-order difference operation and Pettit's mutation point test with a significance level of 0.05. The relative position ratio of the time interval corresponding to the state component within the complete duration of the evolution segment is calculated. Combined with the change direction and change continuity verification results of the evolution segment, the state component is assigned a corresponding change stage identifier. The preset determination rule for the change stage identifier is as follows: when the time ratio of the starting point of the state component's time interval to the starting point of the evolution segment is ≤20%, it is determined as the starting segment identifier of the corresponding change direction; when the time ratio is within the range of 20%-80% and the change continuity meets the preset fluctuation threshold requirement, it is determined as the stable segment identifier of the corresponding change direction; when the time ratio is ≥80%, it is determined as the ending segment identifier of the corresponding change direction; if the evolution segment is in a stable direction, it is uniformly determined as a stable steady-state segment identifier. At the same time, for evolution segments with continuity interruption points, a stage switching sub-identifier is added synchronously at the interruption point position to ensure that the change stage identifier can accurately represent the evolution process of the parameter in the current operating state.

[0100] For the extraction of deviation attribute identifiers, the deviation amplitude and deviation duration percentage of the measured liquid level change rate of the liquid level state component relative to the allowable range of liquid level change rate within the preset collaborative working interval are calculated, as well as the deviation amplitude and deviation duration percentage of the measured water temperature change rate of the temperature state component relative to the allowable range of temperature change rate within the preset collaborative working interval. Combined with the deviation attributes pre-assigned in step S105, corresponding deviation attribute identifiers are assigned to the two state components, including four basic identifiers: no deviation, slight deviation, moderate deviation, and severe deviation. Simultaneously, three additional identifiers are added: deviation start, deviation continuation, and deviation overlap. The process sub-identifiers are categorized as follows: deviation amplitude within 20% of the allowable range is considered slight deviation, 20%-50% is moderate deviation, and above 50% is severe deviation. This ensures that the deviation attribute identifiers can quantify the degree and progress of parameter anomalies. After fully extracting the three types of identifiers corresponding to the two state components, the liquid level state component and its corresponding change direction identifier, change stage identifier, and deviation attribute identifier are bound and paired with the temperature state component and its corresponding change direction identifier, change stage identifier, and deviation attribute identifier, according to a unified timestamp benchmark, forming a structured set of state determination bases.

[0101] Step S202: Based on the aforementioned state determination base set, the dominant attributes of liquid level and temperature are determined according to the prior relationship, continuity relationship, and alternation relationship between the liquid level state component and the temperature state component, to obtain one of the following: liquid level dominant identifier, temperature dominant identifier, or no dominant identifier.

[0102] In this embodiment, the order of the liquid level state component and the temperature state component on the same time axis is first compared to extract their starting positions and duration intervals. This identifies whether the liquid level change precedes the temperature change, or vice versa, thus determining the prior relationship. Based on this, the overlapping intervals of the liquid level state component and the temperature state component in time are further identified. When both maintain continuous change or change in the same direction within the same time interval, this interval is determined to be a continuous relationship. When the liquid level state component and the temperature state component are detected to be adjacent... When changes occur alternately or the direction of change alternates within a given interval, the process is determined to be an alternating relationship. Subsequently, based on the proportion and order of occurrence of various relationships within a preset time window (consistent with the acquisition interval of 100ms in step S101), the dominant attributes of liquid level and temperature are determined. Specifically, when both the preceding relationship and the continuous relationship point to the liquid level state component, it is marked as a liquid level dominant identifier. When the corresponding relationship points to the temperature state component, it is marked as a temperature dominant identifier. When the preceding relationship and the alternating relationship are intertwined and do not form a single dominant trend, it is marked as a non-dominant identifier.

[0103] Step S203: Based on the dominant attribute identifier and the corresponding deviation attribute identifier, the operating status of the circulating water tank is divided into zones. The state that satisfies both the liquid level and temperature within the coordinated range is divided into the steady-state coordinated zone. The state with a single dominant attribute is divided into the dominant intervention zone. The state with no dominant identifier and overlapping deviation attributes is divided into the liquid level and temperature conflict zone.

[0104] In this embodiment, the dominant attribute identifier generated in step S202, which is completely identical to the time reference of the state determination base set, and the deviation attribute identifiers with unique timestamps bound to the liquid level state component and temperature state component in the state determination base set in step S201, are used as the only input data sources. First, the time axis alignment and combination operation of the dominant attribute identifier and the corresponding deviation attribute identifier is performed. Specifically, the high-precision timestamps of all continuous sampling points within the complete time interval corresponding to the current running state are extracted. These timestamps match the 100ms sampling interval preset in step S101, ensuring that each sampling point is uniquely bound to the dominant attribute identifier, liquid level state component deviation attribute identifier, and temperature state component deviation attribute identifier at the corresponding time. Then, according to the ascending time order of the timestamps, the dominant attribute identifier at the same sampling time is paired with the deviation attribute identifiers of the two components one by one to form a continuous determination sequence composed of dominant type and deviation type pairing with a single sampling point as the smallest unit, without time gaps or data loss. Each sequence element is accompanied by a unique and identical timestamp identifier.

[0105] Feasible, after completing the judgment sequence construction, the segment construction operation of the judgment sequence is executed. First, the core constraint rules and boundary judgment conditions for segment merging are preset. The segment merging constraint rules are that the dominant attribute identifiers of adjacent sampling points are completely consistent, and the deviation attribute identifiers of the two components are in a continuous extension state. The continuous extension state is defined as the deviation level change without a jump across levels, the deviation core attribute (no deviation / deviation exists) not being reversed, and the switching of the deviation process identifier conforming to the preset temporal evolution logic. The boundary judgment conditions are that the dominant attribute identifiers of adjacent sampling points change their enumeration values, or the core attribute of the deviation attribute identifiers of the two components is reversed, or the deviation... When a level jump occurs, the sampling points that meet the boundary judgment conditions are marked as the segment boundary. Then, the entire judgment sequence is traversed in ascending order of time. Using the marked segment boundary as the dividing node, all dominant type and deviation type pairing sets that meet the segment merging constraint rules between two adjacent segment boundaries are uniformly divided into the same judgment segment. At the same time, the minimum number of valid sampling points for a single judgment segment is preset to 5. Instantaneous jump segments with fewer than 5 sampling points are directly marked as unprocessed boundary segments to ensure that each valid judgment segment has a unique continuous time interval, a fixed dominant attribute identifier, and a continuously evolving deviation attribute pairing set.

[0106] After completing the construction of the full judgment section, the partitioning and classification operation of each judgment section is performed. First, a partitioning mapping rule based on the operating characteristics of the circulating water tank and fully matching the identification system of the previous steps is predefined. The rule is as follows: when the dominant attribute identifier of the judgment section is no dominant identifier, and the deviation attribute identifiers of the liquid level state component and the temperature state component are both without deviation, and the corresponding parameter change rate falls completely within the cooperative working range preset in step S104, it is classified as a steady-state cooperative zone; when the dominant attribute identifier of the judgment section is a liquid level dominant identifier or a temperature dominant identifier, and the deviation attribute identifier of the dominant component is slightly or above, the deviation level of the follower component is lower than that of the dominant component, and the change direction of both conforms to step S2 When the preceding-following or continuous dominant relationship is determined in 02, it is classified as a dominant intervention zone; when the dominant attribute identifier of the judgment segment is no dominant identifier, the deviation attribute identifiers of the two components are both slightly or above and the change directions are completely opposite, or the dominant attribute identifier switches frequently within the segment, or both components are heavily deviated and there is a reverse restraint relationship determined in step S104, it is classified as a liquid level and temperature conflict zone. Then, for each valid judgment segment, the combination type of its dominant attribute identifier and deviation attribute identifier is extracted and matched and verified one by one with the predefined partition mapping rules. Each judgment segment is assigned a unique corresponding partition type identifier to obtain the initial partition classification result covering the entire time interval.

[0107] Specifically, after completing the initial partitioning and classification, the partition boundary calibration operation of adjacent judgment segments is performed based on the initial partitioning and classification results. The minimum effective segment time span is preset to 10 sampling intervals, i.e., 1000ms. This time span matches the inertial characteristics of the working condition response of the circulating water tank liquid level and water temperature parameters. All adjacent judgment segments are traversed in ascending order of the time axis, and the time span of each individual judgment segment is checked one by one. When there is a boundary that crosses the judgment segment with a time span smaller than the preset minimum effective segment time span, the corresponding splitting or merging process is performed: if the initial partitioning types of the two adjacent judgment segments before and after the boundary crossing segment are completely consistent, the full data of the boundary crossing segment and the time interval are merged into the same partition judgment segment of the adjacent segments before and after, and the partition boundary and time interval of the merged segment are updated synchronously.

[0108] If the initial partition types of the adjacent segments spanned by the boundary are inconsistent, then based on the matching degree between the dominant attribute identifier, deviation attribute identifier and the adjacent segments, the segment spanned by the boundary is split and merged into the adjacent segments with higher matching degree. The matching degree is calculated by weighted summation of three dimensions: consistency of dominant attribute, continuity of deviation level, and convergence of change direction, with weights of 40%, 35%, and 25%, respectively. At the same time, adjacent segments with completely identical partition type identifiers and no other partition type segments in between are merged to form continuous and uninterrupted segments. After completing boundary calibration and segment processing, the partition type of all processed judgment segments is re-verified, ultimately forming a partition generation result that covers the complete time interval of the current operating state, with no time overlap, no interval gaps, and clear boundaries. This result includes three types of partitions: steady-state coordination zone, dominant intervention zone, and liquid level and temperature conflict zone. Each partition is accompanied by a unique continuous time interval, full data of the corresponding dominant attribute and deviation attribute combination, and operating condition trajectory data of all sampling points within the segment, which can be directly used for subsequent protection action triggering and operating condition control strategy generation of the circulating water tank.

[0109] Step S204: Based on each zone, construct a linkage control target formed by the combination of liquid level parameters and temperature parameters. In this step, match the pairing of change direction, deviation sequence, and deviation superposition in the combination relationship to form the target constraint sequence of the corresponding zone.

[0110] In this embodiment, the three types of partitions finally generated in step S203—steady-state collaborative zone, dominant intervention zone, and liquid level / temperature conflict zone—with unique continuous time intervals, standardized partition type identifiers, and full-volume homogeneous working condition trajectory data, are used as the sole processing objects. First, each partition is traversed sequentially according to its partition type. The original liquid level measured parameter sequence and water temperature measured parameter sequence, which perfectly match the 100ms sampling interval preset in step S101 and have one-to-one corresponding homogeneous timestamps, are extracted from the complete time interval of each partition. Simultaneously, the liquid level change rate sequence, temperature change rate sequence, deviation attribute sequence, and change direction identifier sequence, which are perfectly aligned with the time axes of the two parameter sequences, are extracted. Then, the internal parameters of each partition are processed through the steps... The start and end times of the effective judgment segment after boundary calibration in S203 are the splitting boundaries. At the same time, the time reference of the continuous time slice unit rearranged by the unified time axis in step S104 is matched. The liquid level parameter sequence and temperature parameter sequence in the same partition are synchronously split. During the splitting process, it is strictly ensured that the number of sampling points of the liquid level parameter sequence and temperature parameter sequence in each split sub-unit is completely consistent, the timestamps correspond one-to-one, there is no time misalignment, and there is no missing data. At the same time, the minimum number of effective sampling points of a single sub-unit is preset to 5. Sub-units with insufficient sampling points are merged with adjacent sub-units of the same attribute before the splitting is completed. Finally, multiple parameter alignment units with unified time reference and completely aligned parameters are formed.

[0111] Specifically, after completing the construction of the full parameter alignment unit, for the liquid level and temperature parameters within each parameter alignment unit, a standardized construction of three types of combination relationships is performed based on a unified timestamp index. The first type is change direction pairing, which extracts the liquid level change direction identifier and temperature change direction identifier corresponding to each sampling moment within the parameter alignment unit, binds the two change direction identifiers at the same moment to form a single-moment change direction pairing, and then uses a frequency statistics algorithm to count the dominant pairing types that account for more than 80% within the unit, assigning corresponding classification identifiers according to a standardized enumeration format, including four categories: stable-stable, synchronous rise, synchronous fall, and reverse change. The second type is deviation sequence pairing, which extracts the trigger moment when the liquid level and temperature parameters within the parameter alignment unit change from a non-deviation state to a deviation state, and the turning point moment of the deviation level change, and calculates the two-way parameter... The absolute time difference and sequential relationship of the deviation trigger time are combined with the consistency of the deviation attributes to form four types of pairing results: synchronous deviation, liquid level-first deviation, temperature-first deviation, and no deviation. Corresponding standardized enumeration classification labels are assigned synchronously. The third type is deviation superposition pairing. First, the liquid level deviation amplitude and temperature deviation amplitude in the parameter alignment unit are normalized to the 0-1 interval based on the allowable range of the preset collaborative working interval. Then, the vector superposition method is used to calculate the coupled superposition value of the two normalized deviation amplitudes at each sampling time. At the same time, the sign correlation of the two deviation amplitudes is calculated. Based on the magnitude and sign correlation of the coupled superposition value, five types of pairing results are formed: no superposition, slight superposition, moderate superposition, severe superposition, and reverse cancellation. Corresponding standardized enumeration classification labels are assigned synchronously, completing the full construction and classification label assignment of the three types of pairing results.

[0112] The feasible approach involves assigning classification identifiers to all parameter alignment units. Then, based on these identifiers, sequential concatenation along the timeline is performed on each parameter alignment unit. The concatenation process strictly adheres to the core rule of not concatenating across partitions. All parameter alignment units within the same partition are traversed in ascending order of timestamps. The consistency of the three-category pairing classification identifiers between adjacent parameter alignment units and the continuity of the time intervals between the two units are compared one by one. Specifically, the termination time of the previous unit must completely coincide with the start time of the next unit without any time gap. If the dominant classification identifiers of the three-category pairings of two adjacent parameter alignment units are completely identical and the time intervals meet the continuity requirement, the two units are directly concatenated. For parameter alignment units within the same partition with identical classification identifiers but with a small gap of no more than two sampling intervals in their time intervals, existing linear interpolation methods are used to fill in the gaps in the parameters and identifiers before concatenation. Finally, all parameter alignment units with continuous pairing relationships within the same partition are connected to form a target generation sequence that perfectly matches the partition's time interval, is temporally continuous, and has no data breakpoints. Each target generation sequence is uniquely bound to the type identifier of its respective partition.

[0113] Subsequently, based on the target generation sequence corresponding to each partition, the results of change direction pairing, deviation sequence pairing, and deviation superposition pairing within the sequence are sequentially embedded into a preset, time-dimensional quadruple constraint expression structure commonly used in the industrial process control field. This constraint expression structure fixedly contains four core fields: constraint object, constraint dimension, constraint threshold, and constraint effective time interval. Among them, the change direction pairing results are embedded into the change rate sign constraint and time-series coordination constraint fields in the constraint dimension, synchronously matching the change rate allowable threshold of the corresponding partition. The deviation sequence pairing results are embedded into the time-series priority constraint field in the constraint dimension, synchronously matching the time-series constraint rule of regression of the preceding parameter first and regression of the subsequent parameter. The deviation superposition pairing results are embedded into the coupling amplitude constraint and parameter upper and lower limit constraint fields in the constraint dimension, synchronously matching the coupling deviation allowable threshold of the corresponding superposition level. At the same time, the constraint priority is preset for different partition types, with the parameter upper and lower limit constraints of the steady-state coordination zone having the highest priority, the dominant parameter constraints of the dominant intervention zone having the highest priority, and the coupling deviation constraints of the liquid level and temperature conflict zone having the highest priority. After embedding, all constraint entries with effective time intervals are continuously sorted in ascending order of the time axis, finally forming a target constraint sequence that corresponds one-to-one with each partition.

[0114] Step S205: During the control execution process, the linkage control target of each partition is dynamically revised according to the target constraint sequence. When the change direction indicator or deviation attribute indicator is switched, the corresponding partition is switched and the linkage control target is updated synchronously.

[0115] In this embodiment, the control execution process is based on a fixed 100ms control scan cycle that is completely synchronized with the preset acquisition cycle in step S101. It adopts a cyclic scanning mechanism commonly used in the field of industrial control. In the program execution stage of each control scan cycle, the target constraint sequence generated in step S204, which has a unique partition affiliation, a precise effective time interval, and a standardized four-tuple constraint structure, is read cyclically. First, the timestamp corresponding to the current control scan cycle is locked by the existing timestamp matching algorithm. Then, a continuous sequence segment in the target constraint sequence whose effective time interval completely covers the current control time is extracted. This segment is determined as the execution reference unit of the current control scan cycle. The time span of the execution reference unit is preset to 5 consecutive control scan cycles. At the same time, the partition affiliation identifier is checked to ensure that the partition type identifier of the execution reference unit is completely consistent with the partition affiliation corresponding to the current operating state, so as to ensure that the constraint rules of the execution reference unit are completely matched with the characteristics of the current operating condition partition.

[0116] Specifically, within each control scan cycle, based on the time reference of the execution reference unit, the measured liquid level and temperature parameters at the current control moment are synchronously acquired in real time through the acquisition link in step S101. Simultaneously, the identifier extraction logic in step S201 is invoked to generate change direction identifiers and deviation attribute identifiers corresponding to the liquid level and temperature parameters at the current moment. These two types of identifiers are then compared and verified field-by-field with the pre-recorded reference identifiers in the execution reference unit that match the change direction and deviation attributes at the corresponding moment. The preset rule for determining identifier switching events is: when the change direction identifier occurs... Inconsistent switching of enumerated values ​​such as stable to rising, rising to falling, and falling to stable, or deviation of attribute identifiers, or leaps in deviation levels from mild to moderate to severe, or inconsistencies between the pairing relationship of two types of identifiers and the baseline pairing relationship of the execution reference unit, and if the above states continue for more than two consecutive control scan cycles, are determined to be valid identifier switching events. This avoids false triggering caused by instantaneous sampling noise and vibration of the circulating water pump. At the same time, for the identified identifier switching events, the precise switching trigger time, the identifier values ​​before and after the switching, and the type of pairing relationship deviation are recorded to complete the structured marking of the switching events.

[0117] In practice, when a valid identifier switching event is detected, the execution of the constraint rules of the original execution reference unit within the current control scan cycle is immediately terminated. The valid termination time of the original execution reference unit is locked as the trigger time of the identifier switching event. Based on the change direction identifier and deviation attribute identifier after the switch, combined with the liquid level and temperature parameter sequences continuously collected within three control scan cycles before and after the switch time, the partition type corresponding to the current operating condition is re-determined, and the switch verification of the current partition is completed. At the same time, with the switch trigger time as the starting boundary and five consecutive control scan cycles as the initial time span, the liquid level and temperature parameters continuously collected after the switch time, as well as the corresponding identifier after the switch, are extracted to construct a sequence starting unit that matches the newly determined partition type. The sequence starting unit completely contains the initial data of the change direction pairing, deviation sequence pairing, and deviation superposition pairing corresponding to the new partition, ensuring that the time base of the new unit is completely connected with the trigger time of the switch event, without time gaps or data breakpoints.

[0118] Specifically, after constructing the sequence starting unit of the new partition, the constraint expression structure embedding rules and pairing relationship construction logic of step S204 are called. For the pairing results of real-time change direction, deviation sequence, and deviation superposition within the sequence starting unit, combined with the constraint priority rules preset for the type of the new partition, the upper and lower limit constraints of parameters and the stability of change rate are executed first in the steady-state coordination zone, the regression constraints of the dominant parameter and the coordination constraints of the following parameter are executed first in the dominant intervention zone, and the amplitude limit constraint of coupling deviation and the decoupling constraint of reverse change are executed first in the liquid level and temperature conflict zone. The three types of pairing results are embedded into the preset four-tuple constraint expression structure in sequence to generate the initial target constraint sequence. At the same time, in the subsequent control scan cycle, the newly generated target constraint sequence is continuously extended and dynamically corrected based on the real-time acquired parameters and identifiers to ensure that the constraint sequence is completely synchronized with the changes in real-time operating conditions, so as to realize the switching and transition of the linkage control target during the partition switching process.

[0119] Step S3: When the liquid level or temperature is determined to be the main intervention zone or the temperature-dominant intervention zone, the main variable priority and secondary variable compensation linkage control is executed. When the liquid level and temperature conflict zone is determined to be the liquid level and temperature conflict zone, the joint control of pre-suppression and correction based on the change trend is executed.

[0120] The specific steps of step S3 are as follows:

[0121] Step S301: Read the current partition identifier. When it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, determine the corresponding main variable and secondary variable, and extract the change direction identifier and deviation attribute identifier of the main variable and secondary variable at the current moment.

[0122] In this embodiment, the partition identifier output by the preceding partition determination step is read, and variables are assigned according to a pre-set partition-variable mapping rule. Specifically, when the partition identifier is a level-dominant intervention zone, the level parameter corresponding to the level state component is determined as the primary variable, and the temperature parameter corresponding to the temperature state component is determined as the secondary variable. When the partition identifier is a temperature-dominant intervention zone, the temperature parameter is determined as the primary variable, and the level parameter as the secondary variable. After assigning the primary and secondary variables, the level state component and temperature state component corresponding to the current control moment are extracted from the state determination base set. By comparing the relative change relationship between the current sampled value and the sampled value at the previous moment, the change direction identifiers of the primary and secondary variables are determined. Furthermore, by segmenting the position of the current sampled value within a preset target interval, the corresponding deviation attribute identifier is obtained. The change direction identifier and deviation attribute identifier can be obtained using time-series data interpretation methods commonly used in existing industrial process control and recorded in discrete identifier form, thereby completing the identification of primary and secondary variables and the extraction of their state attributes.

[0123] Step S302: Within the dominant intervention zone, a main variable control sequence is generated based on the change direction identifier and deviation attribute identifier of the main variable. During the advancement of the main variable control sequence, a compensation sequence corresponding to the secondary variable is generated synchronously, so that the change process of the secondary variable maintains a pairing relationship with the main variable control sequence.

[0124] In this embodiment, based on the change direction identifier and deviation attribute identifier of the main variable obtained in step S301, the main variable is segmented within a preset time advancement window. The change direction and deviation attribute of consecutive moments are combined to form main variable control segments arranged in chronological order, and the main variable control sequence is constructed sequentially according to the segment order. During the generation of the main variable control sequence, for each main variable control segment, the change direction identifier and deviation attribute identifier of the secondary variable at the corresponding moment are extracted, and the secondary variable is responded and mapped according to the existing coupling relationship rules between the two. The coupling relationship rules are set using the master-slave variable coordination and adjustment strategy commonly used in existing industrial control. For example, the response direction of the secondary variable is determined according to the change direction of the main variable, and the response interval of the secondary variable is divided according to the deviation degree of the main variable, thereby forming secondary variable compensation segments that correspond one-to-one with the main variable control segments. Subsequently, each secondary variable compensation segment is arranged in the same chronological order as the main variable control segment to construct the secondary variable compensation sequence, and the synchronous correspondence with the main variable control sequence is maintained during the sequence advancement process, thereby realizing the pairing generation between the main variable control sequence and the secondary variable compensation sequence.

[0125] It should be noted that the main variable control sequence and the secondary variable compensation sequence correspond to specific equipment control execution quantities. Specifically, the adjustment of the liquid level parameter is achieved by controlling the opening of the water supply valve or the drain valve, and the adjustment of the temperature parameter is achieved by controlling the heat exchanger heat exchange intensity or the cooling water flow rate. During the control process, the change trend of the control command of the corresponding actuator is determined according to the main variable control sequence, and the execution quantity is continuously adjusted by proportional regulation or PID control. At the same time, the secondary variable compensation sequence is used to synchronously correct another control quantity, so that its change trend is coordinated with the main variable control, thereby realizing the linkage regulation of liquid level and temperature.

[0126] Step S303: When a liquid level and temperature conflict zone is identified, a change trend sequence is constructed based on the change order of liquid level parameters and temperature parameters in the historical time slice and the current change direction identifier. A pre-suppression sequence is generated based on the change trend sequence to intervene in the parameter change path before it enters.

[0127] In this embodiment, the change records of liquid level and temperature parameters are extracted from historical time slices. The starting points, directions of change, and alternating positions of the two types of parameters are marked according to time sequence to form a basic sequence reflecting the alternating evolution relationship between the two. Subsequently, the liquid level change direction identifier and temperature change direction identifier obtained at the current moment are embedded into the end position of the basic sequence, and the sequence is extended according to the continuity relationship between adjacent time slices to obtain a change trend sequence containing historical evolution and current state. On this basis, combined with the feedforward regulation commonly used in existing process control, the alternating segments, reverse segments, or continuous deviation segments appearing in the change trend sequence are identified, and corresponding preprocessing instruction identifiers are assigned to each type of segment according to preset rules. The preprocessing instruction identifiers are arranged in time sequence to generate a pre-suppression sequence. The pre-suppression sequence is used to determine the execution order of control instructions before the parameters actually enter the corresponding change path, thereby completing the pre-intervention preparation for the parameter change path.

[0128] Step S304: After the pre-suppression sequence is executed, the trend sequence is updated by combining the currently acquired liquid level parameters and temperature parameters, and a correction sequence is generated based on the updated trend sequence.

[0129] In this embodiment, at the end of the execution time window corresponding to the pre-suppression sequence, the liquid level and temperature parameters at the current moment are acquired and compared with the corresponding parameters before the execution of the pre-suppression sequence to determine the actual change direction and deviation position of the liquid level and temperature within the time window. Subsequently, the real-time acquired change information is embedded into the end of the original change trend sequence in chronological order, and the historical segments in the original sequence corresponding to the current time window are replaced or truncated to form an updated change trend sequence reflecting the latest evolution state. Based on this, referring to the implementation method of the feedback correction mechanism in existing industrial control, the updated change trend sequence is re-segmented and identified, and the continuous segments, reverse segments, and alternating segments are marked respectively. According to the preset sequence reconstruction rules, the various segments are rearranged to generate a correction sequence corresponding to the current actual evolution state. The correction sequence is consistent with the updated change trend sequence in time and is used to replace the original pre-suppression sequence to participate in subsequent control execution, thereby completing the sequence update and control correction process based on real-time feedback.

[0130] Step S4: During the liquid level and temperature regression process, continuously update the coupling deviation value, and adaptively redistribute the water replenishment, drainage and heat exchange intensity based on the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, maintain the circulating water tank in a disturbance collaborative control state within the threshold range.

[0131] In this embodiment, during control execution, the current values ​​of liquid level and temperature parameters are acquired according to a preset sampling period. Combined with the corresponding change direction and deviation attribute indicators, the relative deviation relationship between liquid level and temperature is updated, thus forming a time-progressing sequence of coupling deviation values. Subsequently, the updated coupling deviation values ​​are segmented and categorized according to their deviation type and change trend. Based on the existing multi-variable coordinated adjustment allocation principle in process control, the control commands corresponding to water replenishment, drainage, and heat exchange intensity are divided into several adjustable units, and an allocation relationship corresponding to the coupling deviation values ​​is established between each unit. On this basis, according to the changes in the coupling deviation values ​​in each segment, the adjustable units are proportionally or sequentially adjusted. The sequence switching allows the water replenishment, drainage, and heat exchange intensity to form a coordinated control combination within the same time window. When the liquid level parameter and temperature parameter stably fall within the preset liquid level range and preset temperature range for several consecutive sampling periods, the current control combination is frozen, and only parameters with boundary fluctuations are locally adjusted. This allows the circulating water tank to maintain a coordinated control state with limited disturbances within a predetermined range. The preset liquid level range refers to the upper and lower limits of the liquid level designed for the circulating water tank. For example, if the full scale is 1m, the preset liquid level range is 0.8m-0.95m. The preset temperature range is determined according to the target temperature based on process requirements, such as 50℃, or based on the temperature fluctuation range statistically analyzed from historical normal operation data, such as 49.5℃-50.5℃.

[0132] It should be further noted that all threshold parameters, weighting coefficients, and judgment conditions involved can be preset based on the equipment specifications, process requirements, and historical operating data of the circulating water tank. Specifically, the range of values ​​for each parameter can be determined by statistical analysis of historical operating data or by experimental calibration. At the same time, during actual operation, the above parameters can be adaptively adjusted according to the real-time operating status to improve the adaptability of the control strategy to different operating conditions.

[0133] Example 2

[0134] Please see Figure 4 Another embodiment of the present invention provides a liquid level and temperature linkage control system for the circulating water tank of an underwater pelletizer, comprising: a data acquisition module, a control target generation module, a linkage control module, and an update control module;

[0135] The data acquisition module is used to collect the liquid level parameters and water temperature parameters in the circulating water tank in real time, and to construct the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range.

[0136] The control target generation module is used to classify the circulating water tank into layers based on the current operating state, dividing it into a steady-state coordination zone, a liquid level-dominated intervention zone, a temperature-dominated intervention zone, and a liquid level-temperature conflict zone, and generating corresponding linkage control targets for different zones. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level parameters and water temperature parameters.

[0137] The linkage control module is used to perform linkage control with priority to main variables and compensation for secondary variables when it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, and to perform joint control based on the trend of change for pre-suppression and correction when it is determined to be a liquid level-temperature conflict zone.

[0138] The update control module is used to continuously update the coupling deviation value during the liquid level and temperature regression process, and adaptively redistribute the water replenishment, drainage and heat exchange intensity according to the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, the circulating water tank is kept in a disturbance collaborative control state within the threshold range.

[0139] In addition, the parts of the technical solutions provided in the embodiments of this application that are consistent with the implementation principles of the corresponding technical solutions in the prior art have not been described in detail, so as to avoid excessive elaboration.

[0140] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for linking the liquid level and temperature of the circulating water tank in an underwater pelletizer, characterized in that, include: Real-time acquisition of liquid level and water temperature parameters in the circulating water tank, and construction of the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range. The circulating water tank is divided into four zones based on the current operating status: steady-state coordination zone, liquid level-dominated intervention zone, temperature-dominated intervention zone, and liquid level-temperature conflict zone. Corresponding linkage control targets are generated for each zone. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level and water temperature parameters. When the area is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, a linkage control with priority given to the main variable and compensation for the secondary variable is implemented. When the area is determined to be a liquid level-temperature conflict zone, a joint control based on the trend of change is implemented with pre-suppression and correction. During the liquid level and temperature regression process, the coupling deviation value is continuously updated, and the water replenishment, drainage and heat exchange intensity are adaptively redistributed according to the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, the circulating water tank is kept in a disturbance collaborative control state within the threshold range. The system collects real-time liquid level and temperature parameters within the circulating water tank and constructs the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and the coupling deviation between liquid level and temperature, including: Within a preset acquisition period, the liquid level and water temperature parameters of the circulating water tank at corresponding times are acquired, and liquid level time series and temperature time series are formed in chronological order. The liquid level time series and temperature time series are aligned according to the same decision window, isolated collected values ​​that do not fall into the decision window are removed, and the aligned liquid level time series and temperature time series are paired as homogeneous state segments. Based on each homogeneous state segment, the direction of change, continuity of change, and stage transition position of the liquid level parameter, as well as the direction of change, continuity of change, and stage transition position of the water temperature parameter are identified in sequence, and the liquid level evolution segment corresponding to the liquid level change rate and the temperature evolution segment corresponding to the temperature change rate are determined accordingly. The liquid level evolution segment and the temperature evolution segment are sequentially correlated to generate a liquid level-temperature correlation segment. Based on the order of deviation, duration of deviation and overlap of deviation of each correlation segment relative to the preset cooperative working range, a coupling deviation value between liquid level and temperature is generated. The current operating state of the circulating water tank is calibrated based on the liquid level evolution segment, the temperature evolution segment, and the coupling deviation value. The current operating state includes one of the following: liquid level deviation state, temperature deviation state, liquid level and temperature synchronous deviation state, and liquid level and temperature alternating deviation state.

2. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 1, characterized in that, The liquid level evolution segment and the temperature evolution segment are sequentially correlated to generate a liquid level-temperature correlation segment. Based on the order of deviation, duration of deviation, and overlap of deviations of each correlation segment relative to a preset cooperative working interval, a coupling deviation value between liquid level and temperature is generated, including: The liquid level evolution segment and the temperature evolution segment are rearranged on a unified time axis to form a continuous time slice unit, and a candidate association set is established for the liquid level evolution segment and the temperature evolution segment that fall into the same time slice unit. For each candidate association set, based on the order of triggering of the liquid level evolution segment and the temperature evolution segment, the overlap of the start and end of the change, and the consistency of the change direction, it is determined that it belongs to one of the following: sequential triggering relationship, synchronous change relationship, or reverse restraint relationship, and each candidate association set is assigned a corresponding relationship identifier. Based on the relationship identifier, the continuous time slice units are merged or divided to construct the liquid level-temperature associated segment; Each associated segment is serialized according to its entry order, dwell time, and overlapping range relative to the preset collaborative working area, forming a coupling deviation value that includes deviations from the order, deviations from the duration, and deviations from the overlapping segments.

3. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 2, characterized in that, Based on the liquid level evolution segment, the temperature evolution segment, and the coupling deviation value, the current operating status of the circulating water tank is calibrated, including: The liquid level evolution segment and the temperature evolution segment are cross-arranged according to the corresponding time slice units to form a state sequence unit composed of the interweaving of the liquid level evolution trajectory and the temperature evolution trajectory. The coupling deviation value is introduced into each state sequence unit, and the deviation attribute is assigned to each state sequence unit to form a composite state segment with deviation attribute. The composite state segments are merged according to a preset state mapping rule, and composite state segments with the same combination of deviation attributes and continuous in time are merged into state candidate segments. Priority screening and boundary determination are performed among the candidate state segments, and the candidate state segments that meet the priority determination conditions are determined as the current operating state of the circulating water tank.

4. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 1, characterized in that, Based on the current operating status, the circulating water tank is divided into four zones: a steady-state coordination zone, a level-dominated intervention zone, a temperature-dominated intervention zone, and a level-temperature conflict zone. Corresponding linkage control objectives are generated for each zone, including: The current operating state is deconstructed into liquid level state components and temperature state components, and the corresponding change direction identifier, change stage identifier, and deviation attribute identifier are extracted to form a basic set for state determination. Based on the aforementioned state determination base set, the dominant attributes of liquid level and temperature are determined according to the prior relationship, continuity relationship, and alternation relationship between the liquid level state component and the temperature state component, resulting in one of the following: liquid level dominant identifier, temperature dominant identifier, or no dominant identifier. Based on the dominant attribute identifier and the corresponding deviation attribute identifier, the operating status of the circulating water tank is divided into zones, including the steady-state coordination zone, the dominant intervention zone, and the liquid level and temperature conflict zone. Based on each zone, a linkage control target formed by the combination of liquid level parameters and temperature parameters is constructed, forming a target constraint sequence for the corresponding zone; During the control execution process, the linkage control targets of each partition are dynamically revised according to the target constraint sequence. When the change direction indicator or deviation attribute indicator is switched, the corresponding partition is switched and the linkage control targets are updated synchronously.

5. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 4, characterized in that, Based on the dominant attribute identifier and the corresponding deviation attribute identifier, the operating status of the circulating water tank is divided into zones, including a steady-state coordination zone, a dominant intervention zone, and a liquid level / temperature conflict zone. The dominant attribute identifier and the corresponding deviation attribute identifier are aligned and combined in chronological order to form a judgment sequence consisting of dominant type and deviation type pairing; The determination sequence is segmented, and the pairing set that maintains the same dominant attribute identifier and extends continuously from the deviation attribute identifier is divided into the same determination segment, and the segment boundary is formed at the position where the dominant attribute identifier or the deviation attribute identifier switches. For each judgment segment, it is partitioned and classified according to the combination type of its dominant attribute identifier and deviation attribute identifier to obtain the initial partitioning and classification results; Based on the initial partition classification results, the partition boundaries between adjacent judgment segments are calibrated, and the range of each partition is redefined according to the processed judgment segments to form the partition generation results, including the steady-state coordination zone, the dominant intervention zone, and the liquid level and temperature conflict zone.

6. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 5, characterized in that, Based on each zone, a linkage control target formed by a combination of liquid level parameters and temperature parameters is constructed, forming a target constraint sequence for the corresponding zone, including: Extract the liquid level parameter sequence and temperature parameter sequence corresponding to each partition, and then split the liquid level parameter sequence and temperature parameter sequence synchronously according to the time segment within the partition to form multiple parameter alignment units; The liquid level parameter and temperature parameter in each parameter alignment unit are combined to construct the relationship, and the pairing of change direction, deviation sequence, and deviation superposition are generated respectively. The different types of pairing results are classified and labeled. Based on the classification identifier, the parameter alignment units are sequentially connected to form the target generation sequence by connecting parameter alignment units with continuous pairing relationships within the same partition. Based on the target generation sequence, the change direction pairing, deviation sequence pairing, and deviation superposition pairing are sequentially embedded into the preset constraint expression structure to form the target constraint sequence of the corresponding partition.

7. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 6, characterized in that, During the control execution process, the linkage control targets of each partition are dynamically revised according to the target constraint sequence. When the change direction indicator or deviation attribute indicator changes, the corresponding partition is switched and the linkage control targets are updated synchronously, including: During the control execution process, the target constraint sequence is read cyclically, and the sequence segment corresponding to the current control moment is used as the execution reference unit; Based on the execution reference unit, the change direction identifier and deviation attribute identifier of the liquid level parameter and temperature parameter are continuously acquired and compared with the pairing relationship recorded in the execution reference unit to identify whether there is an identifier switching event. When the identifier switching event is detected, the current partition is re-determined based on the changed direction identifier and deviation attribute identifier after the switch, and the original execution reference unit is terminated at the switching time. At the same time, the sequence start unit corresponding to the new partition is generated. For the sequence starting unit of the new partition, reconstruct the target constraint sequence corresponding to it.

8. The underwater pelletizer circulating water tank level and temperature linkage control method as described in claim 1, characterized in that, When the system is determined to be in a liquid level-dominated intervention zone or a temperature-dominated intervention zone, a linkage control system prioritizing the primary variable and compensating for the secondary variable is implemented. When the system is determined to be in a liquid level-temperature conflict zone, a joint control system based on trend-based pre-suppression and correction is implemented, including: Read the current partition identifier. When it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, determine the corresponding main and secondary variables, and extract the change direction identifier and deviation attribute identifier of the main and secondary variables at the current moment. Within the dominant intervention zone, a main variable control sequence is generated based on the change direction identifier and deviation attribute identifier of the main variable. During the advancement of the main variable control sequence, a compensation sequence corresponding to the secondary variable is generated synchronously, so that the change process of the secondary variable maintains a pairing relationship with the main variable control sequence. When a liquid level and temperature conflict zone is identified, a change trend sequence is constructed based on the change order of liquid level parameters and temperature parameters in historical time slices and the current change direction identifier. A pre-suppression sequence is then generated based on the change trend sequence to intervene in the parameter change path that has entered the zone. After the pre-suppression sequence is executed, the trend sequence is updated by combining the currently acquired liquid level and temperature parameters, and a correction sequence is generated based on the updated trend sequence.

9. A liquid level and temperature linkage control system for an underwater pelletizer circulating water tank, used to implement the liquid level and temperature linkage control method for an underwater pelletizer circulating water tank as described in any one of claims 1-8, characterized in that, include: Data acquisition module, control target generation module, linkage control module, and update control module; The data acquisition module is used to collect the liquid level parameters and water temperature parameters in the circulating water tank in real time, and to construct the current operating status of the circulating water tank based on the liquid level change rate, temperature change rate, and coupling deviation value between liquid level and temperature. The coupling deviation value is used to characterize the degree of deviation between the current liquid level state and the current temperature state relative to the preset cooperative working range. The control target generation module is used to classify the circulating water tank into layers based on the current operating state, dividing it into a steady-state coordination zone, a liquid level-dominated intervention zone, a temperature-dominated intervention zone, and a liquid level-temperature conflict zone, and generating corresponding linkage control targets for different zones. The linkage control targets are dynamically determined based on the combination relationship, change direction, and deviation degree of liquid level parameters and water temperature parameters. The linkage control module is used to perform linkage control with priority to main variables and compensation for secondary variables when it is determined to be a liquid level-dominated intervention zone or a temperature-dominated intervention zone, and to perform joint control based on the trend of change for pre-suppression and correction when it is determined to be a liquid level-temperature conflict zone. The update control module is used to continuously update the coupling deviation value during the liquid level and temperature regression process, and adaptively redistribute the water replenishment, drainage and heat exchange intensity according to the updated coupling deviation value. When the liquid level meets the stable pelletizing requirements and the water temperature meets the allowable fluctuation range of the process, the circulating water tank is kept in a disturbance collaborative control state within the threshold range.

Images