An intelligent industrial steam flow regulation system

Abstract

This invention relates to the field of industrial process automatic control technology, specifically to an intelligent industrial steam flow regulation system. This invention, by performing time-series identification of the continuous changing trend of steam state, can dynamically classify steam state under multi-variable conditions, thereby anticipating steam state switching paths. Based on this, it introduces an information coupling measurement mechanism between enthalpy change and flow rate, and combines the calculation of mutual information between enthalpy and flow rate to accurately identify abnormal sections with regulation deviations in actual operation. Furthermore, it uses the quantification result of the deviation between the target flow rate and the actual output flow rate to determine the possibility of regulation, effectively dividing controllable and uncontrollable sections, and setting a fallback value range limit for uncontrollable sections. Thus, in subsequent target value adjustments, the target settings for the remaining sections can be reset specifically, and a priority order can be constructed based on the regulation sequence and magnitude to achieve dynamic sorting and output of regulation strategies.

Description

Technical Field

[0001] This invention relates to the field of industrial process automatic control technology, and in particular to an intelligent regulation system for industrial steam flow. Background Technology

[0002] The field of industrial process automatic control technology mainly involves the real-time monitoring, adjustment and control of various equipment and process parameters in industrial production processes to achieve automated and efficient management of the production process.

[0003] Among them, the traditional industrial steam flow intelligent regulation system refers to a control system that needs to adjust the flow rate of steam in real time according to process requirements when steam is used as a heat source or power medium in pipelines during industrial processes, thereby solving the problem of steam supply and demand matching and energy utilization efficiency.

[0004] Existing technologies mainly rely on real-time adjustment of steam flow to cope with changes in process requirements. However, their control strategies are mostly based on single-point data feedback or set-value rule adjustments, lacking the ability to systematically judge the trend of steam state changes. Especially when the steam state fluctuates in stages, they cannot identify potential mismatch risks. In addition, such systems usually use the current flow deviation as the sole basis for adjustment, ignoring the energy conversion relationship between steam enthalpy and flow rate. This leads to the system still adjusting according to the original target when coupling deviation occurs, which can easily cause local energy waste or system fluctuations. For example, under high temperature and high pressure, if the trend of state interval switching is not identified, the system will continue to execute the static target flow command, resulting in over-adjustment or under-adjustment, affecting heat load balance and operational safety. Furthermore, there is no closed-loop backoff mechanism for adjustment failure history, and it is impossible to set response restriction strategies for abnormal sections. This results in insufficient adaptability of the control system in dynamic and complex environments and frequent adjustment failures. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and propose an intelligent industrial steam flow regulation system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent industrial steam flow regulation system, the system comprising:

[0007] Steam State Identification Module: Collects time series data on steam temperature, pressure, and density in industrial pipelines over a specified period, determines steam state changes based on the time series, and obtains a state change marker sequence.

[0008] Enthalpy-flow coupling analysis module: Referring to the state change marking sequence, the change in steam enthalpy for each time period is calculated back, the steam output flow rate for the corresponding time period is collected, the mutual information between the change in steam enthalpy and the steam output flow rate is compared, and the coupling abnormal section is marked.

[0009] Adjustment space determination module: compares the deviation between the preset steam target flow rate and the current steam output flow rate, determines the adjustment space of each of the coupling abnormal sections, and sets a forced return threshold for the unadjustable sections to obtain the adjustment lock section and the set of return thresholds;

[0010] Target setting adjustment module: Based on the adjustment locking segment and backoff threshold set, adjust the target setting value of steam output for each time period and generate a set of flow regulation control instructions.

[0011] The present invention is improved in that the state change marking sequence includes a state type mark, the location of the change, and a continuous marking segment; the coupling abnormal segment includes an abnormal segment number, the start and end of the corresponding time period, and the range of steam flow affected; the adjustment locking segment and backoff threshold set includes a locking time period, a set threshold value range, and an execution restriction label; and the flow regulation control instruction set includes the adjusted target flow, the regulation execution identifier, and the time period control instruction.

[0012] The present invention is improved in that the steam state identification module includes:

[0013] Status Acquisition Submodule: Collects steam temperature, pressure and density data in industrial pipelines for a specified period of time, and constructs a basic data sequence of steam status.

[0014] Interval Classification Submodule: Based on the temperature and pressure combination in the steam state basic data sequence, and by referring to the IAPWS steam reference boundary value, determine the current steam state interval, and label the type of each interval with density trend information to generate a state-assigned interval label set;

[0015] Change recognition submodule: Based on the state attribution segment label set, compare the changes in state categories in adjacent time periods, extract the time point and segment number of state switching, record the state path changes according to the time sequence structure, and generate a state change tag sequence.

[0016] The present invention is improved in that the enthalpy-flow coupling analysis module includes:

[0017] Enthalpy calculation submodule: acquires the temperature, pressure and density data corresponding to each time period in the state change marker sequence, calculates the specific enthalpy value of steam in each time period, and the difference in specific enthalpy values ​​within consecutive time periods, and generates the steam enthalpy change.

[0018] Flow extraction submodule: acquires steam flow meter monitoring data within the time period corresponding to the change in steam enthalpy, and generates steam output flow rate;

[0019] Coupled Section Marking Submodule: Calculates the mutual information value between the change in steam enthalpy and the steam output flow rate in the corresponding time period, marks the time periods where there is a deviation from the coupling relationship, and generates coupling abnormal sections.

[0020] The present invention is improved in that the adjustment space determination module includes:

[0021] Target value acquisition submodule: Acquires the preset steam target flow rate for each time period and constructs a preset steam target flow rate set;

[0022] Deviation Judgment Submodule: Compares the preset steam target flow rate set with the steam output flow rate of the coupling abnormal section segment by segment, filters out the segments whose deviation exceeds the allowable threshold for steam regulation, and generates a set of abnormal regulation sections;

[0023] Section marking submodule: Marks the abnormal sections in the set of abnormal adjustment sections as unadjustable, and sets the corresponding flow back threshold according to the section number, generating the adjustment lock section and back threshold set.

[0024] The present invention is improved in that the target setting adjustment module includes:

[0025] Locking response submodule: Based on the locked sections in the adjustment locking segment and backoff threshold set, perform a reset operation on the steam flow target value within the relevant time period, mark the unadjustable state, remove the original target value corresponding to this type of section, and generate the locking segment removal result;

[0026] Threshold comparison submodule: Based on the remaining time period in the locked segment elimination result, extract the corresponding steam output flow rate, compare it with the same segment back-off threshold, reset some target values ​​according to the comparison result, and generate target value adjustment result;

[0027] Set up the summary submodule: unify the adjustment values ​​of each time period in the target value adjustment results with the lock segment status identifier, and generate a flow regulation control instruction set.

[0028] The present invention is improved by further including an execution order optimization module, which sorts the flow regulation control instruction set according to execution priority and generates a flow regulation sequence index.

[0029] The flow regulation sequence index includes a sorting number, regulation timing relationship, and priority execution identifier.

[0030] The present invention is improved in that the execution order optimization module includes:

[0031] Validity filtering submodule: Obtain all adjustment setting instructions in the flow regulation control instruction set, merge valid entries according to time tags, and generate a set of adjustment target times;

[0032] Sorting Generation Submodule: Based on the set of adjustment target times, it sets sorting rules by combining two dimensions: time order and set value size, and outputs a sorting priority time series list;

[0033] Index labeling submodule: For the sorted priority time series list, number the sequential execution labels from front to back, record the time period and adjustment instruction number corresponding to the sorting position, and generate a flow adjustment sequence index.

[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0035] This invention identifies the continuous changing trend of steam state over time, enabling dynamic classification of steam state under multi-variable conditions. This allows for early detection of steam state switching paths. Furthermore, an information coupling measurement mechanism between enthalpy change and flow rate is introduced. By calculating the mutual information between enthalpy and flow rate, abnormal sections with regulation deviations in actual operation are accurately identified. The possibility of regulation is further judged by quantifying the deviation between the target flow rate and the actual output flow rate, effectively dividing controllable and uncontrollable sections. A fallback value limit is set for uncontrollable sections, allowing for targeted resetting of target settings for remaining sections during subsequent target value adjustments. A priority order is constructed based on the regulation sequence and magnitude, achieving dynamic sorting and output of regulation strategies. This process improves the matching accuracy of steam system regulation while ensuring feasibility, avoiding the application of invalid commands to sections without regulation space. This effectively improves the overall system's energy utilization efficiency and regulation response capability under complex operating conditions. In practical applications, through comparison with on-site industrial data and simulation verification, this intelligent regulation system effectively improves the accuracy and dynamic response capability of steam flow regulation. Especially under complex operating conditions such as frequent steam state switching, high temperature and high pressure fluctuations, and variable loads, this system can improve energy utilization efficiency by an average of about 8% to 15% and shorten the adjustment response time by about 25% compared with traditional control strategies. It significantly reduces the risk of waste caused by over- or under-adjustment of energy and improves the overall stability and economy of the system operation. Attached Figure Description

[0036] Figure 1 This is a system module diagram of the present invention;

[0037] Figure 2 This is a system framework diagram of the present invention;

[0038] Figure 3 This is a schematic diagram of the steam state identification module of the present invention;

[0039] Figure 4 This is a schematic diagram of the enthalpy-flow coupling analysis module of the present invention;

[0040] Figure 5This is a schematic diagram of the adjustment space determination module of the present invention;

[0041] Figure 6 A schematic diagram of the adjustment module for setting the objective of this invention;

[0042] Figure 7 This is a schematic diagram of the execution order optimization module of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0045] Please see Figure 1 This invention provides a technical solution: an intelligent industrial steam flow regulation system, the system comprising:

[0046] Steam State Identification Module: Collects time series data on steam temperature, pressure, and density in industrial pipelines over a specified period, determines steam state changes based on the time series, and obtains a state change marker sequence.

[0047] Enthalpy-flow coupling analysis module: Referring to the state change marker sequence, the steam enthalpy change for each time period is calculated back, the steam output flow rate for the corresponding time period is collected, the mutual information between the steam enthalpy change and the steam output flow rate is compared, and the coupling abnormal section is marked.

[0048] Adjustment space determination module: compares the deviation between the preset steam target flow rate and the current steam output flow rate, determines the adjustment space of each coupling abnormal section, and sets a forced return threshold for the unadjustable section to obtain the adjustment lock section and the set of return thresholds;

[0049] Target setting and adjustment module: Based on the adjustment lock segment and backoff threshold set, adjust the target set value of steam output flow rate for each time period and generate a flow rate regulation control instruction set;

[0050] The status change marker sequence includes status type marker, location of change, and continuous marker segment; the coupled abnormal segment includes abnormal segment number, corresponding time period start and end, and range of affected steam flow; the regulation lock segment and backoff threshold set includes lock time period, set threshold value range, and execution restriction label; the flow regulation control instruction set includes adjusted target flow, regulation execution identifier, and time period control instruction.

[0051] Please see Figure 2 and Figure 3 The steam status identification module includes:

[0052] Status Acquisition Submodule: Collects steam temperature, pressure and density data in industrial pipelines for a specified period of time, and constructs a basic data sequence of steam status.

[0053] To collect steam temperature, pressure, and density data for a specified period in an industrial pipeline, temperature, pressure, and density sensors need to be installed at key locations along the pipeline. The acquisition cycle is typically set by the control system, which automatically reads the sensor outputs at fixed time intervals. The data acquisition module obtains the temperature, pressure, and density information for the current time point from each sensor and marks it with a unified timestamp to ensure that all types of data have a time correspondence. During this process, the control system interacts with each sensor through communication protocols, receiving sampled values ​​in real time and caching them in an edge computing device. Subsequently, the cached data is synchronously uploaded to the database. To avoid outliers from corrupting the data sequence, a data preprocessing procedure needs to be set up, including setting and verifying the range of temperature, pressure, and density sampled values. After identifying outliers, they are marked and isolated. To ensure data stability, the sensor devices are baseline calibrated daily using standard media or reference values ​​for comparison. The acquisition system can record continuous data sequences over an entire period according to the configuration strategy, thus forming a basic data sequence of steam state.

[0054] The interval classification submodule: Based on the temperature and pressure combination in the steam state basic data sequence, and by comparing it with the IAPWS steam reference boundary value, it determines the current steam state interval, and combines the density trend information to label the type of each interval, generating a state-assigned interval label set;

[0055] When determining the state based on the temperature and pressure combination in the steam state basic data sequence and comparing it with the IAPWS steam reference boundary values, the temperature and pressure values ​​of each sampling point need to be substituted into the state determination logic for boundary judgment. According to the temperature and pressure critical boundary table provided by IAPWS, the corresponding saturation pressure limit is found at the set temperature. Then, the current pressure value is compared with the boundary value to determine whether it is higher, equal to, or lower than the boundary pressure. This can preliminarily determine whether it is a liquid phase, saturated state, or superheated state. On this basis, the density data sequence is combined to compare the density values ​​in continuous time periods. The density change trend is determined by difference analysis or the method of averaging adjacent multiple points. If the density increases continuously, it can be marked as a density rising zone; if it decreases, it is marked as a decreasing zone; when the trend is stable, it is marked as a stable zone. By combining the temperature and pressure state with the density change trend, labels such as "superheated - density decreasing" and "saturated - density stable" can be generated, and each data point is assigned a corresponding interval label. This operation is performed on each sampling point in the entire data sequence to form a set of state assignment interval labels.

[0056] Change recognition submodule: Based on the state attribution segment label set, compare the changes in state category in adjacent time periods, extract the time point and segment number of state switching, record the state path changes according to the time sequence structure, and generate a state change tag sequence.

[0057] When identifying changes based on the state-attribute segment label set, the entire label sequence is first traversed in chronological order. The state label content of two adjacent time points is compared. If there is an inconsistency in the state label content, it can be determined that the state has changed. The current time point is recorded as the state transition time, and its position number in the overall data sequence is marked. Then, the subsequent label content is traversed to identify all state change nodes. Segments with consecutive label content are grouped and uniformly numbered. The state change nodes are constructed into a state path sequence. The start time, end time, start label, end label, and segment number of each segment in the path are recorded. By comparing the differences in state categories of consecutive segments in the path, it is determined whether there are features such as repetition, alternation, and transition in the state path. During the execution process, a state transition record table should be constructed to assist in the analysis of data temporal characteristics. This table consists of state path number, preceding and following state labels, and transition time point. After processing all time periods, a complete state change label sequence can be obtained.

[0058] Please see Figure 2 and Figure 4 The enthalpy-flow coupling analysis module includes:

[0059] Enthalpy Calculation Submodule: Obtains temperature, pressure and density data corresponding to each time period in the state change marker sequence, calculates the specific enthalpy value of steam in each time period, and the difference in specific enthalpy values ​​within consecutive time periods, and generates the steam enthalpy change.

[0060] After obtaining the temperature, pressure, and density data for each time period in the state change marker sequence, it is necessary to first classify the state and determine which enthalpy calculation path to use. For steam in a superheated state, the specific enthalpy calculation expression in Region 3 of the IAPWS-IF97 industrial steam property formula needs to be used. Region 3 is applicable to high-pressure and high-temperature conditions, and temperature and pressure can be used as input parameters. By referring to the relevant specific enthalpy calculation expression for Region 3 in the IAPWS industrial steam standard table, the partial derivative form based on the state equation is selected. The input temperature is 260 and the pressure is 3. After substituting into the specific enthalpy calculation formula, the specific enthalpy is found to be 2800. If the temperature of the next time period is 255 and the pressure is 3, the specific enthalpy is found to be 2750. The difference between the two is -50, which is recorded as the enthalpy change for that period. For steam in a saturated state, the saturated steam in Table 2 of IAPWS-IF97 needs to be referred to. The specific enthalpy data for steam shows that at a temperature of 250°C, the specific enthalpy of saturated liquid is 1080, and the specific enthalpy of saturated vapor is 2800. By comparing the current density with the theoretical saturated liquid and vapor densities, if the current density is closer to the vapor density, the specific enthalpy of steam is taken as 2800. Then, the density is compared again for the next time period. If the state transitions to saturated liquid, the corresponding specific enthalpy becomes 1080, with a change in specific enthalpy of -1720. If the steam is in the liquid phase, then... The specific enthalpy formula for Region 1 of IAPWS-IF97 is applicable to low-temperature and medium-pressure conditions. The specific enthalpy can be calculated using an empirical polynomial. At a temperature of 230°C and a pressure of 3.5, the specific enthalpy is 1080 when substituted into the empirical expression. The difference between this specific enthalpy and the specific enthalpy of the next time period is used to obtain the change in enthalpy for that period. The above calculation is performed cyclically for all state segments. The specific enthalpy results of each segment are then processed and the difference is calculated to form a sequence of changes in steam enthalpy.

[0061] Flow extraction submodule: Acquires steam flow meter monitoring data within the time period corresponding to the change in steam enthalpy, and generates steam output flow rate;

[0062] When acquiring steam flow meter monitoring data corresponding to the steam enthalpy change within a given time period, a time synchronization mechanism is needed to pair the enthalpy change sequence with the flow data. First, the start and end times of each enthalpy change segment are extracted from the records. For example, if the enthalpy change segment corresponds to 10:00 to 10:00:10, then all valid sampled values ​​within that time period are extracted from the flow meter data. Assuming the flow meter records data once per second, ten data points can be extracted from that time period. These ten data points are summed and averaged to represent the average flow rate for that time period. If the flow rate change is small in some periods, the difference between the start and end times can be directly used as the representative output flow rate. Then, this average flow rate value is paired with the enthalpy change value for that time period to form a unified data structure. If flow data is missing or abruptly changes during the process, data interpolation between preceding and following time periods can be used to fill in the anomalies and ensure data continuity. Subsequently, the enthalpy changes of all time periods are correlated with the steam output flow rate and organized into a sequence structure.

[0063] Coupled Section Marking Submodule: Calculates the mutual information value between the change in steam enthalpy and the steam output flow rate in the corresponding time period, marks the time periods with deviations from the coupling relationship, and generates coupling anomalous sections;

[0064] Calculating the statistical dependence between changes in steam enthalpy and steam output flow rate requires the use of mutual information as a metric. Mutual information reflects how much information one variable contains about the other, and when applied to continuous time-series data, it can reveal the coupling strength between enthalpy and flow rate changes. The core steps of this process include: constructing a joint probability distribution, calculating mutual information values, developing a threshold strategy, and identifying coupling deviations.

[0065] The change in enthalpy and steam output flow A set of observation samples is formed by pairing samples in each corresponding time period, assuming that there exists... For each sample point, normalize the two variables separately (linear transformation to...). (interval), then on and Divided into For example, two identical interval segments. This results in a combination of 5 enthalpy ranges and 5 flow rate ranges. Next, for each group... Data points are statistically analyzed to determine their corresponding enthalpy ranges. and flow range The frequencies of all combinations are counted to form a joint frequency matrix, which is then transformed into a joint probability distribution. .

[0066] The formula for calculating mutual information is:

[0067] ;

[0068] in, Enthalpy change With steam flow The mutual information value between them is compared with the mutual information threshold obtained by subtracting the standard deviation from the mean of all mutual information. If the mutual information value is lower than the mutual information threshold in a certain period, it is marked as an abnormal period with a deviation from the coupling relationship. Enthalpy data falls into the first The interval and traffic data fall into the first... The joint probability of each interval; Enthalpy data falls into the first The marginal probabilities of each interval; Traffic data falls into the first The marginal probabilities of each interval; The total number of intervals into which the variable is discretely divided.

[0069] Mutual information is essentially a measure of the difference between the joint and independent distributions of variables. and Completely independent, then Mutual information is zero if there is no statistical dependency; if there is a statistical dependency, the mutual information is positive, and the higher the value, the stronger the coupling.

[0070] Suppose we are analyzing samples from 10 time periods, and we have already completed normalization and interval mapping. What is the value of the enthalpy change? With steam flow Each region was divided into 5 intervals. After statistical analysis, the following joint probability and marginal probability (in terms of relative frequency) were obtained: The enthalpy is in the third interval, and the flow rate is also in the third interval; The enthalpy is in the fourth interval, and the flow rate is in the fourth interval; The enthalpy is in the 5th interval, the flow rate is in the 5th interval, and the corresponding marginal probabilities are: , ; , ; , .

[0071] Project 1: Calculation Mutual information items:

[0072] ;

[0073] Project Two: Calculation Mutual information items:

[0074] ;

[0075] Project 3: Calculation Mutual information items:

[0076] .

[0077] The sum of the three terms yields the total mutual information value: .

[0078] This value represents the total mutual information contributed by these three typical combinations. If other combinations are zero, this is the total mutual information value. In actual calculations, all non-zero values ​​should be considered. The total mutual information value is obtained by calculating the terms and then summing them. .

[0079] To identify coupling deviation sections, evaluation criteria need to be established based on the mutual information values ​​calculated for all time periods. The first step is to statistically analyze the mutual information values ​​for all corresponding time periods, forming a complete mutual information value sequence. The second step is to calculate the average value of this sequence, serving as the baseline for the overall coupling level between enthalpy change and steam flow rate under the current operating conditions. The third step is to calculate the standard deviation of the mutual information value sequence to reflect the fluctuation of coupling strength. The fourth step is to set a judgment threshold: subtracting a standard deviation from the average value serves as the lower limit for judging coupling anomalies. The fifth step is to iterate through the mutual information values ​​for each time period; if the mutual information value for a certain time period is lower than this threshold, it is determined to be an abnormal section where enthalpy and flow rate deviate from their coupling relationship.

[0080] Suppose a system, after continuous observation over 30 time periods, calculates the mutual information value for each segment, obtaining an average value of 1.8 and a standard deviation of 0.4. The threshold for determining this threshold is [threshold value missing]. If the mutual information of segment 12 is 1.2 and segment 19 is 1.1, both below 1.4, they are marked as coupling anomaly segments. These segments will be further extracted to form a coupling anomaly segment marking table, containing the anomaly time period number, start and end times, corresponding mutual information value, and a combination of enthalpy interval number and flow rate interval number, used for subsequent state analysis and fault backtracking.

[0081] Please see Figure 2 and Figure 5 The adjustment space determination module includes:

[0082] Target value acquisition submodule: Acquires the preset steam target flow rate for each time period and constructs a preset steam target flow rate set;

[0083] To obtain the preset steam target flow rate for each time period, the target values ​​corresponding to each time period must first be extracted from the scheduling plan, process settings, or control parameters. These target values ​​may come from the daily production scheduling plan in the process control system, the prediction results of historical operation models, or manually set operation curves. The system matches the target values ​​according to the timestamp comparison structure. For example, if the target flow rate from 9:00 AM to 9:05 AM is 1,500 cubic meters per hour, then the preset flow rate for all time points within that time period is uniformly set to this value. If there are variable rate requirements for certain time periods, the target values ​​are updated according to the segmented setting rules. In addition, to facilitate comparison with actual measurement data, the target value structure needs to be organized into a time period index and target flow rate key-value pair structure, stored in the system cache, and kept aligned with the time periods output by the coupled analysis module. Organizing all target values ​​in order can form a complete preset steam target flow rate set.

[0084] Deviation Judgment Submodule: Compares the preset steam target flow rate set with the steam output flow rate of the coupled abnormal section segment by segment, filters out the segments whose deviation exceeds the allowable threshold for steam regulation, and generates a set of abnormal regulation sections;

[0085] When comparing the preset steam target flow rate set with the steam output flow rate of the coupled abnormal section segment by segment, the target flow rate and the actual output flow rate are first matched one by one according to the time series. The difference between the target flow rate and the actual value is calculated in each time period. Then, the difference is compared and judged according to the steam regulation allowable threshold set by the system. For example, if the set allowable deviation is 5%, the percentage error of the actual flow rate relative to the target value is calculated for each time period. When the error is greater than 5%, it is judged as a deviation exceeding the limit segment. After comparing all the data segment by segment, the segment numbers of all error exceeding the limit are extracted. These segments are arranged by time and recorded as a unified set of regulation abnormal segments. This set contains the start and end time, target value, actual value and exceedance range of each abnormal segment, which is convenient for subsequent regulation limit labeling.

[0086] Section Marking Submodule: Marks abnormal sections in the set of abnormal adjustment sections as unadjustable, and sets the corresponding flow back-off threshold according to the section number, generating adjustment lock-up sections and back-off threshold sets;

[0087] When marking abnormal segments in the set of abnormal adjustment segments as unadjustable, it is first necessary to identify which segments have lost effective adjustment response capability during operation. This identification is based on the following criteria: the flow deviation within the segment has not been corrected by adjustment commands for multiple consecutive time periods, and the deviation trend continues to expand or remains stable in the high deviation range for a long period. Combined with the system's adjustment response records for similar historical operating conditions, if such segments have experienced multiple adjustment failures, they are determined to be unadjustable. Subsequently, a flow rollback threshold needs to be set for each segment marked as unadjustable. This threshold serves as a subsequent operational limit benchmark for segments already determined to be unadjustable, used to determine whether there is a serious risk of loss of control or the need for load reduction. The flow rollback threshold is not used to determine whether the segment has entered an unadjustable state, but rather, based on the established unadjustable state, it sets the next stage control lower limit benchmark for the segment. The specific calculation method comprehensively considers the target flow of the segment, the maximum flow deviation value that has occurred within the segment, and the residual deviation amount from adjustment failures in the segment or similar historical segments, using the following formula:

[0088] ;

[0089] in, It is the flow rate reduction threshold for this section, used to reduce or limit the steam flow rate; The target steam flow rate set for this section of the dispatch; The maximum actual deviation value observed in this section is taken from all sampling points. The maximum value; This represents the mean residual deviation after historical adjustment failures, indicating the system's persistent error when adjustments failed to take effect under similar historical conditions. , These are the weighting coefficients for the two factors, and their values ​​are determined based on the following: When the deviation in a certain segment is significantly greater than a certain percentage of the target value (e.g., exceeding 10%), it indicates a significant discrepancy between the segment and the system settings, and the weight of the largest deviation term should be increased. Make the backoff threshold more stringent, for example, set it to 0.6; and if the residual after adjustment failure is too large, it indicates that the control system's response capability is insufficient, then increase the threshold. Weights make the system more sensitive to its historical regulation failure characteristics, for example, set to 0.3.

[0090] Assuming the target flow rate for section 21 is 1600, the maximum deviation is 220, and the historical residual is 50, the calculation is as follows: For the non-adjustable section numbered 21, the corresponding flow back-off threshold is 1483. The system writes this value into the "adjustment lock section and back-off threshold set". In subsequent operation, once the flow of this section is lower than 1483, strategies such as load limiting, alarm or control logic switching can be triggered, thereby completing the limit binding and status structure output of each non-adjustable section.

[0091] Please see Figure 2 and Figure 6 The target setting adjustment module includes:

[0092] Locking response submodule: Based on the locking segments in the adjustment locking segment and backoff threshold set, the target value of steam flow in the relevant time period is set to empty, and the unadjustable state is marked. The original target value corresponding to this type of segment is removed, and the locking segment removal result is generated.

[0093] Based on the locked segments in the adjustment lock segment and backoff threshold set, it is necessary to extract all segment numbers and corresponding time period start and end ranges from the list of segments that have been determined to be in an unadjustable state. Then, search for all target flow records involving this time period in the original target flow table, set the target flow field of these records to null, and simultaneously mark this time period as unadjustable. This marking method can use a Boolean flag field recorded in the target flow table structure. At the same time, set the coverage range of the locked segment to prevent it from participating in the target value adjustment calculation in the future. If a segment spans multiple sampling time points during the operation, the nulling operation should be performed point by point to ensure consistency. For example, if the locked segment numbered 18 covers the time period from 10:15:00 to 10:15:30, then the target value field corresponding to all sampling points in the target flow table for this time period is set to null, and the locking flag is added as true. All target flow data items that have completed the nulling process are removed from the original target value set structure, forming the locked segment removal result.

[0094] Threshold comparison submodule: Based on the remaining time period in the locked segment elimination result, extract the corresponding steam output flow rate, compare it with the same segment back-off threshold, reset some target values ​​according to the comparison result, and generate target value adjustment results;

[0095] When extracting the corresponding steam output flow rate for the remaining time periods based on the locked segment elimination results, it is necessary to match the corresponding timestamp data points from the actual flow record table and compare them with the backoff threshold corresponding to that time period. During the comparison process, judgment is made segment by segment. If the current actual flow rate is higher than the backoff threshold set for that segment, it means that the current flow rate has moved away from the unadjustable critical value and has adjustment space. In this case, the target value for that time period can be restored or appropriately increased. If the current actual flow rate is lower than the backoff threshold, the current target value is kept unchanged or the target is further reduced to prevent the regulation system from misjudging and issuing invalid commands. The adjustment range of the target value can be set according to the ratio of the actual flow rate exceeding the threshold. For example, if the actual flow rate of a certain segment is 1520 and the backoff threshold is 1480, the excess ratio is about 2.7%. Then the adjustment target value is set as the lower limit of the actual flow rate value, such as taking 1490 as the correction target. This judgment and adjustment operation is performed on all remaining time periods one by one, and the new target value adjustment results are compiled.

[0096] Set up the summary submodule: unify the adjustment values ​​of each time period in the target value adjustment results with the status identifier of the locked segment, and generate a set of flow regulation control instructions;

[0097] When unifying the adjustment values ​​for each time period in the target value adjustment results with the lock segment status identifier, a complete data structure needs to be established. This structure includes a timestamp field, a lock status identifier field, and a target flow field. First, the target value adjustment results are parsed line by line, and the target value for each time period is merged and mapped with whether it belongs to a lock segment. Then, the target values ​​within the lock segment are uniformly set to invalid values ​​or default flags, and the remaining unlocked segments are updated with new target values ​​based on the adjustment results. The entire unblocking process needs to maintain consistency with the data format of the regulation and control system interface. For example, data alignment and output should be performed uniformly at a rate of one record every five seconds to form a structured instruction set. Each record in the instruction set includes time, target value, adjustment flag, lock status, and other content.

[0098] Please see Figure 2 and Figure 7 It also includes an execution order optimization module, which sorts the flow regulation control instruction set according to execution priority and generates a flow regulation sequence index.

[0099] The flow regulation sequence index includes sort number, regulation timing relationship, and priority execution identifier;

[0100] The execution order optimization module includes:

[0101] Validity Filtering Submodule: Retrieves all adjustment setting instructions in the flow regulation control instruction set, merges valid entries by time tag, and generates a set of adjustment target times;

[0102] When retrieving all adjustment setting instructions from the flow regulation control instruction set, it is necessary to traverse the entire instruction set data structure, extract entries containing specific numerical settings from each record, remove null value identifiers, placeholder fields, or records marked as locked, and retain only instruction entries that are truly meaningful for regulation. Each instruction must contain at least a timestamp field and a target flow setting field. Then, all valid records are merged by time period according to the time label, that is, multiple setting instructions appearing at the same time point are deduplicated or merged. If multiple adjustment instructions appear at a certain time point, the median of the setting value or the latest record is used as the representative value of that time point. For example, if three setting instructions appear at 10:01:00 AM, namely 1520, 1530, and 1510, they can be merged into 1520 as the valid target. All valid adjustment records merged by time constitute a new set of adjustment target times.

[0103] Sorting Generation Submodule: Based on the target time set, it sets sorting rules by combining time order and set value size, and outputs a sorting priority time series list;

[0104] When sorting based on the target time set, sorting rules need to be constructed based on two dimensions. The first dimension is the chronological order, that is, sorting all target time points from earliest to latest according to the timestamp. The second dimension is the order of the size of the target values. In some operation strategies that require load balancing, it may be necessary to prioritize the processing of instructions with larger or smaller target values. Therefore, the system needs to set a combined sorting logic, such as prioritizing sorting by target value in descending order, and then sorting by time in ascending order when the target values ​​are the same. Specifically, the primary key can be the set value size, and the secondary key can be the time tag.

[0105] Index labeling submodule: For the sorted priority time series list, number the sequential execution labels from front to back, record the time period and adjustment instruction number corresponding to the sorting position, and generate a flow adjustment sequence index;

[0106] When establishing sequential execution labels for a priority time series list, each record must be numbered sequentially from the first time point after sorting. This number serves as the priority identifier for the flow regulation system's scheduling execution. Simultaneously, during the number generation process, the original time period timestamp corresponding to that number and the regulation instruction number in the flow regulation control instruction set must also be recorded. For example, if a record's sorting number is 1, the time is 10:06:00, and the instruction number is 35, this information must be written into the flow regulation sequence index structure. The structure fields include: execution sequence number, time period, instruction number, target value, etc. After numbering and binding all sorting results item by item, a complete sequence index table is formed.

[0107] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent industrial steam flow regulation system, characterized in that: The system includes: Steam State Identification Module: Collects time series data on steam temperature, pressure, and density in industrial pipelines over a specified period, determines steam state changes based on the time series, and obtains a state change marker sequence. Enthalpy-flow coupling analysis module: Referring to the state change marking sequence, the change in steam enthalpy for each time period is calculated back, the steam output flow rate for the corresponding time period is collected, the mutual information between the change in steam enthalpy and the steam output flow rate is compared, and the coupling abnormal section is marked. Adjustment space determination module: compares the deviation between the preset steam target flow rate and the current steam output flow rate, determines the adjustment space of each of the coupling abnormal sections, and sets a forced return threshold for the unadjustable sections to obtain the adjustment lock section and the set of return thresholds; Target setting adjustment module: Based on the adjustment locking segment and backoff threshold set, adjust the target setting value of steam output for each time period and generate a set of flow regulation control instructions.

2. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: The state change marker sequence includes a state type marker, the location of the change, and a continuous marker segment. The coupling anomaly segment includes an anomaly segment number, the start and end of the corresponding time period, and the range of steam flow affected. The adjustment lock segment and backoff threshold set includes a lock time period, a set threshold value range, and an execution restriction label. The flow regulation control instruction set includes the adjusted target flow, the regulation execution identifier, and the time period control instruction.

3. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: The steam state identification module includes: Status Acquisition Submodule: Collects steam temperature, pressure and density data in industrial pipelines for a specified period of time, and constructs a basic data sequence of steam status. Interval Classification Submodule: Based on the temperature and pressure combination in the steam state basic data sequence, and by referring to the IAPWS steam reference boundary value, determine the current steam state interval, and label the type of each interval with density trend information to generate a state-assigned interval label set; Change recognition submodule: Based on the state attribution segment label set, compare the changes in state categories in adjacent time periods, extract the time point and segment number of state switching, record the state path changes according to the time sequence structure, and generate a state change tag sequence.

4. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: The enthalpy-flow coupling analysis module includes: Enthalpy calculation submodule: acquires the temperature, pressure and density data corresponding to each time period in the state change marker sequence, calculates the specific enthalpy value of steam in each time period, and the difference in specific enthalpy values ​​within consecutive time periods, and generates the steam enthalpy change. Flow extraction submodule: acquires steam flow meter monitoring data within the time period corresponding to the change in steam enthalpy, and generates steam output flow rate; Coupled Section Marking Submodule: Calculates the mutual information value between the change in steam enthalpy and the steam output flow rate in the corresponding time period, marks the time periods where there is a deviation from the coupling relationship, and generates coupling abnormal sections.

5. The intelligent industrial steam flow regulation system according to claim 4, characterized in that: The formula is used to calculate the mutual information value between the change in steam enthalpy and the steam output flow rate over a corresponding time period: ； in, It is the change in enthalpy. With steam flow The mutual information value between the two values ​​is compared with the mutual information threshold obtained by subtracting the standard deviation from the mean of all mutual information values. If the mutual information value is lower than the mutual information threshold in a certain period, it is marked as an abnormal period with a deviation from the coupling relationship. The enthalpy data falls into the first The interval and traffic data fall into the first... The joint probability of each interval. The enthalpy data falls into the first The marginal probabilities of each interval Is the traffic data falling into the first The marginal probabilities of each interval It represents the total number of intervals into which the variable is discretely divided.

6. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: The adjustment space determination module includes: Target value acquisition submodule: Acquires the preset steam target flow rate for each time period and constructs a preset steam target flow rate set; Deviation Judgment Submodule: Compares the preset steam target flow rate set with the steam output flow rate of the coupling abnormal section segment by segment, filters out the segments whose deviation exceeds the allowable threshold for steam regulation, and generates a set of abnormal regulation sections; Section marking submodule: Marks the abnormal sections in the set of abnormal adjustment sections as unadjustable, and sets the corresponding flow back threshold according to the section number, generating the adjustment lock section and back threshold set.

7. The intelligent industrial steam flow regulation system according to claim 6, characterized in that: To set the corresponding traffic rollback threshold, use the following formula: ； in, It is the flow rate reduction threshold for the target section, used to reduce or limit the steam flow. The target steam flow rate set for the target section scheduling The maximum actual deviation observed in the target section. This represents the mean residual deviation after historical adjustment failures. , They are respectively and The weighting coefficients.

8. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: The target setting adjustment module includes: Locking response submodule: Based on the locked sections in the adjustment locking segment and backoff threshold set, perform a reset operation on the steam flow target value within the relevant time period, mark the unadjustable state, remove the original target value corresponding to this type of section, and generate the locking segment removal result; Threshold comparison submodule: Based on the remaining time period in the locked segment elimination result, extract the corresponding steam output flow rate, compare it with the same segment back-off threshold, reset some target values ​​according to the comparison result, and generate target value adjustment result; Set up the summary submodule: unify the adjustment values ​​of each time period in the target value adjustment results with the lock segment status identifier, and generate a flow regulation control instruction set.

9. The intelligent industrial steam flow regulation system according to claim 1, characterized in that: It also includes an execution order optimization module, which sorts the flow regulation control instruction set according to execution priority and generates a flow regulation sequence index. The flow regulation sequence index includes a sorting number, regulation timing relationship, and priority execution identifier.

10. The intelligent industrial steam flow regulation system according to claim 9, characterized in that: The execution order optimization module includes: Validity filtering submodule: Obtain all adjustment setting instructions in the flow regulation control instruction set, merge valid entries according to time tags, and generate a set of adjustment target times; Sorting Generation Submodule: Based on the set of adjustment target times, it sets sorting rules by combining two dimensions: time order and set value size, and outputs a sorting priority time series list; Index labeling submodule: For the sorted priority time series list, number the sequential execution labels from front to back, record the time period and adjustment instruction number corresponding to the sorting position, and generate a flow adjustment sequence index.

Images