A data-driven based control method for coal chemical shift process

Abstract

The present application relates to coal chemical process control technical field, and disclose a kind of based on data-driven coal chemical transformation process control method.The temperature, pressure sensor is calibrated on line by forming independent acquisition channel, constructs whole field temperature and pressure state vector, establishes partition index and threshold according to data distribution, and real-time depicts spatial difference;Then use sample to generate dimensionless error model, introduce comprehensive load constraint and construct dual lagrangian function, unify safety constraint and optimization target, form temperature and pressure pair scheduling index, obtain executable control increment after conversion, and be allocated to each measuring point according to partition proportion difference, closed-loop operation and dual factor dynamic switching in fixed cycle, realize whole process adaptive, safe and reliable collaborative control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of coal chemical process control, in particular to a coal chemical shift process control method based on data driving. BACKGROUND

[0002] In a coal chemical plant, the shift process or shift process is usually targeted at coal gas, and the source of coal gas includes coke oven gas and coal gasification gas. There are differences between the two in composition, pressure level, dust and water content characteristics, and working condition fluctuation amplitude, etc., but temperature and pressure need to be controlled in the process of separation, purification, heat exchange, shift reaction and the connecting pipe sections before and after the shift reaction to ensure reaction conversion, equipment safety and stable operation. In existing engineering applications, the control mode mainly uses single loop regulation, such as setting temperature control loop and pressure control loop respectively, relying on fixed set value or segmented set value to maintain the working condition within the allowable range; when the load of the device changes or the source of coal gas is switched, manual adjustment of the set value, switching of the valve position or changing of the heat exchange / throttling working condition is often used to maintain stability. This kind of control mode usually takes a small number of key point measurements as the basis, and it is difficult to fully depict the spatial distribution differences along the process section, and it is easy to have local overheating, local overpressure or abnormal pressure drop while the global average value is normal.

[0003] In addition, the existing technology often uses uniform distribution or fixed weight distribution in the distribution of multiple measurement point control quantities, that is, after determining the overall adjustment direction, the same or preset proportion of the adjustment quantity is applied to the execution link corresponding to each measurement point, and there is a lack of mechanism for differentiating the distribution of control quantities based on the deviation degree of the measurement point, the regional characteristics and the real-time operation state. For the cooperative relationship between temperature and pressure, it is often treated as independent control variables in engineering, and there is a lack of paired design and unified scheduling index for temperature regulation and pressure regulation, which leads to the contradiction that the temperature has reached the standard but the pressure is close to the upper limit or the pressure is stable but the temperature field distribution is uneven when the working condition fluctuates. At the same time, the safety load constraint of the device is realized through alarm, interlock or experience limiting in the existing technology, and the safety constraint and process precision target are often separated: when the load approaches the upper limit, the control strategy is easily conservative or relies on manual intervention, and it is difficult to continuously maintain the working condition precision while ensuring that it does not exceed the limit. In the data processing level, the existing method often relies on experience proportion coefficient or fixed normalization for joint evaluation of temperature and pressure variables with different dimensions, lacks a way to automatically form a scale combined with the field data distribution, and is easily affected by abnormal values, sensor drift or sampling noise; at the same time, the measurement point partitioning often relies on process experience or static regional division, and it is difficult to dynamically update with the change of coal gas source and load, leading to mismatch between partitioning and control strategy.

[0004] To this end, the case aims to propose a data-driven coal chemical shift process control method, based on the data-driven method, through online calibration of temperature and pressure sensors, real-time state vector construction, partition index division, dimensionless error modeling and introduction of comprehensive load constraints, a dual Lagrange structure is formed, forming a scheduling and execution system that takes into account process accuracy and load safety. SUMMARY

[0005] The present application provides a data-driven coal chemical shift process control method, which solves the problems mentioned in the background art.

[0006] The present application provides the following technical solutions: a data-driven coal chemical shift process control method, comprising:

[0007] Temperature and pressure sensors are laid out and calibrated using the two-point method to form independent voltage acquisition channels and calibration coefficients;

[0008] The temperature and pressure samples of each measuring point are combined to form a temperature and pressure state vector, and a partition index is established and divided into a main zone and a regulation zone according to the partition threshold;

[0009] Collecting operation samples, obtaining total dimensionless target values and forming a dimensionless error model;

[0010] Introducing comprehensive load constraints, constructing a dimensionless Lagrange function and a dual factor;

[0011] Generating temperature scheduling indicators and pressure scheduling indicators, setting conversion coefficients and obtaining global temperature control increments and global pressure control increments;

[0012] Distributing temperature and pressure control amounts of each measuring point according to the main zone proportion and the regulation zone proportion;

[0013] Performing temperature and pressure regulation and completing state update within a fixed control period;

[0014] Storing operation data and performing dual factor switching update.

[0015] Optionally, the temperature and pressure sensors are laid out and calibrated using the two-point method to form independent voltage acquisition channels and calibration coefficients, specifically comprising:

[0016] Equidistantly laying out a plurality of physical measuring points along the reaction pipeline and sequentially numbering them;

[0017] Installing a temperature sensor and a pressure sensor at each physical measuring point, respectively, and connecting them to independent voltage acquisition channels, the total number of temperature sensors, the total number of pressure sensors and the total number of physical measuring points are consistent;

[0018] Temperature calibration: select two groups of known temperatures, and record the original voltages of each measuring point at the two groups of temperatures;

[0019] If the two voltages are the same, reselect the calibration point or check the connection;

[0020] If not, calculate the linear slope and linear intercept of each measuring point temperature channel by two-point method to establish the linear conversion relationship from voltage to temperature;

[0021] Pressure calibration: select two groups of known pressure, record the original voltage of each measuring point under the two groups of pressure;

[0022] If the two voltages are the same, reselect the calibration point or check the connection;

[0023] If not, calculate the linear slope and linear intercept of each measuring point pressure channel by two-point method to establish the linear conversion relationship from voltage to pressure;

[0024] During sampling operation, convert the temperature channel voltage into actual temperature and the pressure channel voltage into actual pressure according to the respective linear conversion relationship to form the real-time temperature and real-time pressure of each measuring point.

[0025] Optionally, the combination of the temperature and pressure of each measuring point forms a temperature-pressure state vector, establishes a partition index and divides the main area and the adjustment area according to the partition threshold, specifically including:

[0026] At any sampling time, the real-time temperature and real-time pressure of all measuring points are combined into a temperature-pressure state vector according to the measuring point number;

[0027] Calculate the global average temperature and the global average pressure;

[0028] According to the temperature scale and the pressure scale, for each measuring point, calculate the deviation of each measuring point temperature relative to the global average temperature and the deviation of each measuring point pressure relative to the global average pressure, sum up after normalization according to the corresponding scale, and generate a partition index for each measuring point;

[0029] Within a 600-second time window after system startup, collect all measuring point partition indexes and calculate the median as the partition threshold;

[0030] At any time, if the partition index of a measuring point is not greater than the partition threshold, the measuring point is divided into the main area set; if the partition index of a measuring point is greater than the partition threshold, the measuring point is divided into the adjustment area set.

[0031] Optionally, the operation sample is collected, the total dimensionless target value is obtained, and a dimensionless error model is formed, specifically including:

[0032] Within a 600-second time window after system startup, collect the temperature sample set and the pressure sample set;

[0033] respectively acquire 5% quantile and 95% quantile of the temperature sample, 5% quantile and 95% quantile of the pressure sample;

[0034] acquire the minimum resolvable temperature difference of the temperature sensor and the minimum resolvable pressure difference of the pressure sensor;

[0035] construct the temperature scale and the pressure scale: when the corresponding quantile difference is not zero, the quantile difference is adopted; when the corresponding quantile difference is zero, the minimum resolvable temperature difference or the minimum resolvable pressure difference is adopted;

[0036] for each measuring point, firstly normalize around the target temperature and the target pressure set by the process, then square and add the temperature deviation and the pressure deviation respectively to obtain the double-variable dimensionless deviation of each measuring point;

[0037] average the dimensionless deviations of the measuring points in the main zone set to obtain the main zone error; average the dimensionless deviations of the measuring points in the adjustment zone set to obtain the adjustment zone error; add the main zone error and the adjustment zone error to obtain the total dimensionless target value.

[0038] Optionally, the introducing comprehensive load constraint, constructing the dimensionless Lagrange function and the dual factor, specifically includes:

[0039] after the system is started, collect the temperature sample set and the pressure sample set within a 600-second time window;

[0040] respectively acquire 5% quantile and 95% quantile of the temperature sample, 5% quantile and 95% quantile of the pressure sample;

[0041] acquire the minimum resolvable temperature difference of the temperature sensor and the minimum resolvable pressure difference of the pressure sensor;

[0042] construct the temperature scale and the pressure scale: when the corresponding quantile difference is not zero, the quantile difference is adopted; when the corresponding quantile difference is zero, the minimum resolvable temperature difference or the minimum resolvable pressure difference is adopted;

[0043] for each measuring point, firstly normalize around the target temperature and the target pressure set by the process, then square and add the temperature deviation and the pressure deviation respectively to obtain the double-variable dimensionless deviation of each measuring point;

[0044] average the dimensionless deviations of the measuring points in the main zone set to obtain the main zone error; average the dimensionless deviations of the measuring points in the adjustment zone set to obtain the adjustment zone error; add the main zone error and the adjustment zone error to obtain the total dimensionless target value.

[0045] Optionally, the generating temperature scheduling index and pressure scheduling index, setting conversion coefficient and obtaining global temperature control increment and global pressure control increment, specifically includes:

[0046] A temperature scheduling index is constructed: the difference between the global average temperature and the target temperature is converted by a temperature scale, and an item obtained by weighting the standardized load deviation obtained by standardizing the comprehensive load by a temperature load coefficient is superimposed, to obtain the temperature scheduling index;

[0047] A pressure scheduling index is constructed: the difference between the global average pressure and the target pressure is converted by a pressure scale, and an item obtained by weighting the standardized load deviation obtained by standardizing the comprehensive load by a pressure load coefficient is superimposed, to obtain the pressure scheduling index;

[0048] A conversion coefficient corresponding to the temperature scale and the pressure scale is set, and the conversion coefficient is equal to the inverse of the corresponding scale;

[0049] According to the scheduling index and the conversion coefficient, the global temperature control increment and the global pressure control increment are generated in the opposite direction of the scheduling index, and the corresponding analytical relationship is given.

[0050] Optionally, the temperature and pressure control amounts of each measuring point are distributed according to the main area proportion and the adjustment area proportion, and specifically include:

[0051] The main area proportion and the adjustment area proportion are calculated based on the number of measuring points in the main area set and the adjustment area set, respectively;

[0052] For each measuring point in the main area, the global temperature control increment is distributed to the temperature control amount of each measuring point according to the main area proportion, and the global pressure control increment is distributed to the pressure control amount of each measuring point according to the main area proportion;

[0053] For each measuring point in the adjustment area, the global temperature control increment is distributed to the temperature control amount of each measuring point according to the adjustment area proportion, and the global pressure control increment is distributed to the pressure control amount of each measuring point according to the adjustment area proportion.

[0054] Optionally, the temperature and pressure adjustment and state update are performed in a fixed control period, and specifically include:

[0055] The control period length is set to a preset number of seconds;

[0056] At the end of each control period, the current temperature and the current pressure of each measuring point are updated according to the corresponding temperature control amount and the pressure control amount, to obtain the temperature and pressure at the next time;

[0057] Each temperature actuator and each pressure actuator act according to the temperature control amount of the corresponding measuring point and the pressure control amount of the corresponding measuring point, respectively.

[0058] Optionally, the running data is stored and the dual factor switching update is performed, and specifically includes:

[0059] At the end of each control cycle, the current cycle operation data is stored, including the temperature and pressure of each measuring point, the main zone set and the adjustment zone set, the global temperature control increment and the global pressure control increment, the temperature control amount of each measuring point and the pressure control amount of each measuring point, the comprehensive load value and the standardized load deviation, and the dual factor value;

[0060] According to whether the current cycle comprehensive load exceeds the upper limit of the comprehensive load safety, the dual factor in the next cycle is switched and updated between zero and one.

[0061] The present application has the following beneficial effects:

[0062] 1. Based on the two-point method, the output voltages of the temperature and pressure sensors under two known working conditions are collected, and the real physical quantities are converted through a linear relationship; at the same time, a voltage collection channel and a corresponding calibration coefficient are independently allocated to each sensor channel. The traditional calibration process is embedded in the online sampling process, and the calibration parameters can be dynamically updated without additional downtime or manual intervention, thereby improving the calibration efficiency and data reliability. This method beneficially solves the problems of sensor drift and long-term running error accumulation in the coal chemical field, ensuring that all subsequent decision-making steps are based on reliable data.

[0063] 2. The real-time temperature and pressure information of all measuring points are concatenated as a state vector according to the measuring point number, and on this basis, a partition index function is introduced to normalize and weight the sum of the temperature and pressure deviations of each measuring point, thereby quantifying the regional differences of each measuring point relative to the global average level. The median threshold is used to divide the main zone and the adjustment zone, making the partitioning discrimination robust and resistant to anomalies. It can dynamically identify hot spots or bottleneck areas in the process, helping the control system to allocate resources or adjustment weights to the areas that need adjustment most. Compared with the traditional method of dividing based on global average or empirical threshold, the data-driven partitioning strategy of this scheme is more suitable for nonlinear coupling and multivariate coal chemical application scenarios, ensuring the real-time and reliability of the partitioning discrimination.

[0064] 3. Based on the temperature and pressure sample data in the initialization stage, the quantile and minimum resolution difference are calculated to construct the temperature scale and pressure scale; then the deviation of each measuring point is normalized and squared to form a two-variable dimensionless deviation, and the average deviation of the main zone and the adjustment zone is calculated, and the total target value is obtained by adding them. The second-order judgment strategy of quantile and minimum resolution difference makes the scale construction reflect the data distribution characteristics and cope with extreme deviation values; at the same time, it takes into account the dual needs of the main zone and the adjustment zone in error aggregation. It realizes the smooth mapping from multivariate deviation to a single total target, providing a unified quantitative standard for subsequent optimization; compared with the traditional error model which relies only on a single variable or empirical weight, this scheme is more stable and robust in multivariate coupled systems, effectively avoiding error amplification or conflict.

[0065] 4、From the equipment nameplate to obtain temperature and pressure load coefficient, the global average temperature and pressure are linearly combined according to the coefficient to obtain the comprehensive load, and after further standardization, the dual factor is introduced to form the dimensionless Lagrange function. The process safety load constraint and the target control quantity are fused through the dual Lagrange structure to realize the dynamic adjustment of the dual factor influence under the premise of not exceeding the limit. Both the strict compliance with the safety constraint and the real-time relaxation or tightening mechanism provided by the dual factor for the system ensure that the control system can intelligently switch in the multi-objective trade-off. Compared with the existing method of hard limiting or soft constraint at the end of control, the safety constraint and the target optimization are coupled in the mathematical structure in the present scheme, which not only enhances the safety of the system, but also improves the control performance.

[0066] 5、Under the Lagrange framework, the temperature and pressure scheduling indicators are calculated respectively, and then the dimensionless scheduling indicators are converted into corresponding temperature and pressure control increments by setting conversion coefficients (i.e. scale inverses). An explicit analytical mapping relationship is established between the dimensionless indicators and the physical control quantities, so that the decision output has executability and physical meaning. The conversion steps from the decision layer to the execution layer are simplified, and complex nonlinear mapping or artificial experience adjustment is avoided; at the same time, the order of magnitude of the control increment is ensured to be consistent with the process demand through the accurate conversion coefficient. Compared with the existing method which relies on empirical formula or offline simulation mapping, the present scheme realizes truly data-driven, online-calibratable scheduling indicator conversion, improving the control precision and adaptability.

[0067] 6、According to the number of measuring points in the main zone and the adjustment zone, the partition ratio is calculated, and the global temperature and pressure control increments are distributed to each measuring point according to the ratio. The dynamic partition set size is used as the weight, so that each measuring point control strategy has the adaptive group response attribute. The distribution ratio can be adjusted in real time according to the number of measuring points in the partition, avoiding the waste of control resources or response delay caused by uniform distribution; at the same time, the partition set is automatically reduced in the case of measuring point failure or abnormality, ensuring the robustness of the distribution mechanism. Compared with the traditional static or experience ratio distribution, the present scheme realizes adaptive resource scheduling based on real-time online partition, improving the control efficiency and system fault tolerance.

[0068] 7、Set a fixed control period, update the state of each measuring point according to the allocated temperature and pressure control quantity at the end of each period, and realize physical adjustment through the actuator. The data-driven control decision is embedded into the bounded periodic execution framework, realizing the organic integration of decision and execution. While ensuring real-time, the system has predictability and verifiability, which is helpful for timely evaluation and adjustment of the execution result; at the same time, the introduction of fixed period avoids the issuance of too frequent or too sparse control commands. Compared with the existing event-triggered or completely time-triggered method, the present scheme enhances the stability and maintainability of the system on the basis of ensuring the response speed.

[0069] 8. At the end of each control cycle, the data including the measurement point state, the partition set, the control variable, the comprehensive load and the dual factor are stored comprehensively, and the dual factor value is switched dynamically according to whether the comprehensive load is out of limit. The data storage is coupled with the safety control adaptive mechanism to realize complete record of the process running history and intelligent switching of the dual factor state. High-quality data basis is provided for subsequent fault diagnosis, performance analysis and model updating, and meanwhile, the dual factor can intervene in the control strategy in real time when outside the safety boundary. Compared with the traditional method of only storing key variables or not recording full data on line, the panoramic data and dual factor linkage updating method of the present scheme provide valuable historical support and safety guarantee for process optimization and improvement. BRIEF DESCRIPTION OF DRAWINGS

[0070] Figure 1 The flowchart of the present application. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0072] Embodiment, refer to Figure 1 A data-driven coal chemical transformation process control method, comprising:

[0073] Temperature and pressure sensors are laid out and a two-point method is used to complete calibration, forming an independent voltage acquisition channel and a calibration coefficient;

[0074] The temperature and pressure samples of each measurement point are combined to form a temperature-pressure state vector, a partition index is established, and a main zone and a regulation zone are divided according to a partition threshold;

[0075] Operation samples are collected, a total dimensionless target value is obtained, and a dimensionless error model is formed;

[0076] A comprehensive load constraint is introduced, and a dimensionless Lagrange function and a dual factor are constructed;

[0077] Temperature scheduling indicators and pressure scheduling indicators are generated, conversion coefficients are set, and global temperature control increments and global pressure control increments are obtained;

[0078] The temperature and pressure control variables of each measurement point are distributed according to the main zone proportion and the regulation zone proportion;

[0079] Temperature and pressure regulation is performed and state updating is completed within a fixed control cycle;

[0080] Storing operation data and performing dual factor switching update.

[0081] Firstly, by calibrating the temperature and pressure sensors online and forming independent voltage acquisition channels and calibration coefficients, the problem of signal error accumulation caused by precision drift of field sensors and long-period operation is solved, ensuring the reliability of measurement data from the source; Subsequently, the temperature and pressure data of all measurement points are combined into a unified state vector, and partition indicators and partition thresholds are constructed according to data distribution, realizing real-time dynamic description of process space differences, and avoiding excessive regulation caused by traditional control relying on average value or empirical threshold; Then, by collecting samples to generate a dimensionless error model, multivariate deviations are aggregated into a structured target value, providing a unified quantitative standard for subsequent scheduling decisions; The comprehensive load constraint is introduced and the dual Lagrange function is constructed, which integrates safety load constraint and optimization target into one, so that process precision and safety are considered under the premise of not exceeding the limit; Finally, based on the Lagrange structure, a scheduling index is generated, and through conversion coefficient, an executable temperature and pressure control increment is obtained, and then the control amount is distributed to each measurement point according to the partition ratio, forming an implementable and adaptive control output; The whole process is closed-loop executed in a fixed period, and the dual factor is dynamically switched to ensure that the system automatically converges to the optimal solution under different load conditions. A whole-process closed-loop system from data acquisition to control execution is constructed, which not only improves the measurement and control accuracy, but also ensures the real-time balance of intelligent decision and safety constraint. Compared with existing segmented optimization or static scheduling methods, the system's adaptability and reliability are enhanced.

[0082] The temperature and pressure sensors are laid out and calibrated by two-point method to form independent voltage acquisition channels and calibration coefficients, specifically including:

[0083] Multiple physical measurement points are laid out equidistantly along the reaction pipeline and sequentially numbered;

[0084] A temperature sensor and a pressure sensor are respectively installed at each physical measurement point and connected to independent voltage acquisition channels, and the total number of temperature sensors, pressure sensors and physical measurement points is consistent;

[0085] Temperature calibration: Select two groups of known temperatures and record the original voltages of each measurement point at the two groups of temperatures;

[0086] If the two voltages are the same, reselect calibration points or check the connection;

[0087] If not, calculate the linear slope and linear intercept of each measurement point temperature channel according to the two-point method, and establish the linear conversion relationship from voltage to temperature;

[0088] Pressure calibration: Select two groups of known pressures and record the original voltages of each measurement point at the two groups of pressures;

[0089] If the two voltages are the same, reselect the calibration point or check the connection.

[0090] If they are different, calculate the linear slope and linear intercept of the pressure channel at each measuring point using the two-point method to establish a linear conversion relationship from voltage to pressure.

[0091] During sampling, the voltage of the temperature channel is converted into the actual temperature and the voltage of the pressure channel is converted into the actual pressure according to their respective linear conversion relationships, thus forming the real-time temperature and real-time pressure of each measuring point.

[0092] Further specific implementation steps include:

[0093] In the controlled gas process section of the coal chemical conversion process, the gas source is coke oven gas or coal gasification gas; the controlled gas process section is a gas separation-conversion process section, including but not limited to separation, purification, heat exchange, conversion reaction and its upstream and downstream connecting pipe sections; reaction pipelines are equidistantly arranged along the reaction pipelines of the process section. There are 1 physical measurement point, numbered as follows: ;in, This represents the total number of physical measuring points laid out along the reaction pipeline; Number the physical measurement points;

[0094] At each physical measurement point, one temperature sensor and one pressure sensor are installed, each connected to an independent voltage acquisition channel; the total number of temperature sensors is denoted as . The total number of pressure sensors is recorded as ,satisfy ;

[0095] Temperature sensor calibration: Select two sets of known temperatures Record physical measurement points Voltage output of the temperature sensor at two sets of temperatures ;in, , These are the known calibration temperatures for the first and second groups of temperature channels, respectively. , Physical measuring points Temperature sensor at calibration temperature , The original voltage reading below;

[0096] like If so, reselect the calibration point or check the sensor connection;

[0097] If the two voltages are different, the linear calibration coefficient is calculated as follows: , ;in, For physical measurement points The voltage-temperature linear slope coefficient of the temperature sensor; physical measurement point linear calibration intercept of the temperature sensor at the physical measurement point;

[0098] Pressure sensor calibration: Select two sets of known pressures Record the voltage output of the pressure sensor at the physical measurement point at the two sets of pressures ; wherein, P1, P2 are the first and second sets of known calibration pressures for the pressure channel, respectively; V1, V2 are the raw voltage readings of the pressure sensor at the calibration pressures , at the physical measurement point, respectively;

[0099] If , then reselect calibration points or check sensor connections;

[0100] If the two voltages are different, then calculate the linear calibration coefficient as: , ; wherein, m is the linear slope coefficient of the voltage of the pressure sensor at the physical measurement point ; b is the linear calibration intercept of the pressure sensor at the physical measurement point ;

[0101] At any sampling time , the true temperature and pressure at the physical measurement point are calculated as: , ; wherein, t is a continuous time variable; Vt is the voltage output of the temperature sensor at the physical measurement point at time t; Pt is the voltage output of the pressure sensor at the physical measurement point at time t; Tt is the actual temperature at the physical measurement point at time t; Pt is the actual pressure at the physical measurement point at time t.

[0102] ​​​​​The one-to-one acquisition structure is formed by arranging the measuring points equidistantly along the key reaction pipeline and independently accessing the temperature and pressure sensor channels at each measuring point, which effectively avoids the mutual interference or signal crosstalk of the sensor channels; in the calibration link, two groups of known calibration points are used for real-time calibration, if the two voltage readings are consistent, the re-calibration or connection check step is triggered, so as to solve the blind spot that the dead zone and saturated output cannot be identified in the calibration process; when the voltage response is normal, the linear slope and intercept are calculated to map the voltage signal into physical quantity, the whole process does not need to be offline or use a special calibration table, and the dynamic calibration in the online continuous running state is realized; especially at the sampling time, the original voltage is converted into actual temperature and pressure in real time, so as to ensure that the data relied on by all subsequent logic and algorithm modules are reliable real-time values.

[0103] The temperature and pressure sampling of each measuring point of the combination forms a temperature-pressure state vector, and a partition index is established and the main area and the adjustment area are divided according to the partition threshold, specifically including:

[0104] At any sampling time, the real-time temperature and real-time pressure of all measuring points are combined into a temperature-pressure state vector according to the measuring point number;

[0105] The global average temperature and the global average pressure are calculated;

[0106] According to the temperature scale and the pressure scale, for each measuring point, the deviation of the temperature of each measuring point relative to the global average temperature and the deviation of the pressure of each measuring point relative to the global average pressure are calculated, and the sum is calculated after normalization according to the corresponding scale, so as to generate a partition index for each measuring point;

[0107] Within 600 seconds after the system starts, collect all measuring point partition indexes, and calculate the median as the partition threshold;

[0108] At any time, if the partition index of a measuring point is not greater than the partition threshold, the measuring point is divided into the main area set; if the partition index of a measuring point is greater than the partition threshold, the measuring point is divided into the adjustment area set.

[0109] Further specific implementation operations include:

[0110] At the sampling time , the temperature and pressure of all physical measuring points are combined in order according to the measuring point number to construct a temperature-pressure state vector: ; wherein, is the state vector composed of the temperature and pressure of all physical measuring points at time ;

[0111] The global average temperature and the global average pressure are calculated, specifically:

[0112] , .

[0113] wherein, is the arithmetic mean of all physical measurement point temperatures at time is the arithmetic mean of all physical measurement point pressures at time

[0114] The partition index function at the physical measurement point is constructed, specifically as: wherein, is the partition index value at time at the physical measurement point is the temperature scale; is the pressure scale;

[0115] Within a time interval of 600 seconds after system startup, a partition index set of all measurement points within the time period is constructed: and the median is taken as the partition threshold: wherein, is the continuous time at system startup; is the partition threshold; is the median of a finite sample set;

[0116] If there exists a time such that , the physical measurement point is classified into the main zone physical measurement point set ; otherwise, it is classified into the adjustment zone physical measurement point set .

[0117] First, the real-time temperature and pressure of all measurement points are concatenated in order of numbering to form a unified state vector, providing a basis for multi-dimensional data fusion; then the global average is calculated, and a partition index function is further constructed to normalize and weight-sum the deviation of each measurement point, converting it into a single quantitative index; the median of all indices within a fixed time window after system startup is used as the partition threshold, which cleverly utilizes the anti-exceptional characteristics of the median to solve the interference of extreme deviation or noise on threshold calculation; further, at any time, the measurement points are dynamically divided into the main zone (small deviation) and the adjustment zone (large deviation) according to the size relationship between the partition index and the threshold, realizing the real-time depiction of process spatial heterogeneity.

[0118] The running samples are collected, the total dimensionless target value is obtained, and a dimensionless error model is formed, specifically including:

[0119] Within a 600-second time window after system startup, the temperature sample set and the pressure sample set are collected;

[0120] ​​​​respectively, and the 5th and 95th percentiles of the pressure samples;

[0121] obtaining the minimum resolvable temperature difference of the temperature sensor and the minimum resolvable pressure difference of the pressure sensor;

[0122] constructing the temperature scale and the pressure scale: using the corresponding percentile difference when the corresponding percentile difference is not zero, and using the minimum resolvable temperature difference or the minimum resolvable pressure difference when the corresponding percentile difference is zero;

[0123] for each measurement point, first normalizing around the target temperature and the target pressure set by the process, then squaring and adding the temperature deviation and the pressure deviation respectively to obtain the two-variable dimensionless deviation of each measurement point;

[0124] averaging the dimensionless deviations of the measurement points in the main zone set to obtain the main zone error, averaging the dimensionless deviations of the measurement points in the adjustment zone set to obtain the adjustment zone error, and adding the main zone error and the adjustment zone error to obtain the total dimensionless target value.

[0125] Further specific implementation operations include:

[0126] in the time interval of the initial running phase seconds, respectively, the temperature sample set and the pressure sample set are counted.

[0127] the 5th percentile of the temperature samples is denoted as , , the 95th percentile of the temperature samples is denoted as .

[0128] the 5th percentile of the pressure samples is denoted as , , the 95th percentile of the pressure samples is denoted as .

[0129] obtaining the minimum resolvable temperature difference of the temperature sensor, denoted as .

[0130] obtaining the minimum resolvable pressure difference of the pressure sensor, denoted as .

[0131] respectively, the temperature scale and the pressure scale are constructed as:

[0132] , .

[0133] constructing the two-variable deviation function at the physical measurement point , specifically: ; wherein target temperature set for the process; target pressure set for the process; dimensionless comprehensive deviation value of the physical measuring point dimensionless comprehensive deviation value of the physical measuring point dimensionless comprehensive deviation value of the physical measuring point

[0134] The error functions of the main zone and the adjustment zone are respectively constructed, and specifically are as follows:

[0135]

[0136] wherein, dimensionless comprehensive deviation value of the physical measuring point average value of dimensionless deviations of all measuring points in the main zone at time average value of dimensionless deviations of all measuring points in the adjustment zone at time average value of dimensionless deviations of all measuring points in the main zone at time average value of dimensionless deviations of all measuring points in the adjustment zone at time number of elements in the set number of elements in the set number of elements in the set number of elements in the set

[0137] The total target function of the system is set as the sum of the error functions of the main zone and the adjustment zone, and specifically is as follows: wherein, total dimensionless target value at time total dimensionless target value at time

[0138] By collecting samples in the initial running stage, the quantile deviation and the minimum distinguishable difference of temperature and pressure are counted, the scale parameter is constructed, so that the normalization process not only reflects the actual distribution characteristics of the data, but also avoids the zero deviation leading to the inability to divide; the deviation square accumulation method is used to form the dimensionless deviation of two variables, and the overall deviation of each measuring point relative to the process target is quantified; then the average deviation of the main zone and the adjustment zone is calculated respectively and summed, to obtain the total dimensionless target value, which provides a unified optimization index for subsequent scheduling.

[0139] The comprehensive load constraint is introduced, and the dimensionless Lagrange function and the dual factor are constructed, and specifically include:

[0140] Within a 600-second time window after the system starts, the temperature sample set and the pressure sample set are collected;

[0141] The 5% quantile and the 95% quantile of the temperature sample and the 5% quantile and the 95% quantile of the pressure sample are respectively obtained;

[0142] The minimum distinguishable temperature difference of the temperature sensor and the minimum distinguishable pressure difference of the pressure sensor are obtained;

[0143] ​​Constructing temperature scale and pressure scale: using quantile difference when corresponding quantile difference is not zero, using minimum resolvable temperature difference or minimum resolvable pressure difference when corresponding quantile difference is zero;

[0144] For each measuring point, first normalize around the target temperature and target pressure set by the process, then square and add the temperature deviation and pressure deviation respectively to obtain the double-variable dimensionless deviation of each measuring point;

[0145] Average the dimensionless deviations of the measuring points in the main zone set to obtain the main zone error; average the dimensionless deviations of the measuring points in the adjustment zone set to obtain the adjustment zone error; add the main zone error and the adjustment zone error to obtain the total dimensionless target value.

[0146] Further specific implementation operations include:

[0147] Obtain the temperature load coefficient from the equipment nameplate, denoted as ; obtain the pressure load coefficient, denoted as ;

[0148] Construct a thermal pressure comprehensive load function, specifically: ; wherein, is the thermal pressure comprehensive load value at time ;

[0149] Obtain the maximum comprehensive load safety upper limit allowed by the equipment, denoted as ;

[0150] Standardize the load function to dimensionless form to construct a standardized load function, specifically: ; wherein, is the standardized load deviation at time ;

[0151] Introduce a dual factor to construct a dimensionless Lagrange function, specifically: ; wherein, is the dual factor at time ; is the Lagrange function at time ;

[0152] At the system startup time , the initial value of the dual factor is determined according to the relationship between the comprehensive load and the safety upper limit, specifically: ; wherein, is the initial value of the dual factor at the startup time .

[0153] First, the temperature and pressure load coefficients are obtained according to the equipment nameplate, the global average temperature and pressure are linearly combined to obtain the comprehensive load value; by comparison with the equipment rated upper limit and standardization, the safety load is introduced into the target framework; then the safety constraint is added to the dimensionless target value by using the dual factor coefficient to form the Lagrange function; at the system startup, the dual factor value is initialized according to whether the load is over limit, and the safety relaxation or tightening basis is provided for the subsequent algorithm.

[0154] The generation temperature scheduling index and pressure scheduling index are set, and the conversion coefficient is obtained to obtain the global temperature control increment and the global pressure control increment, specifically including:

[0155] The temperature scheduling index is constructed: the difference between the global average temperature and the target temperature is converted according to the temperature scale, and the standardized load deviation obtained by standardizing the comprehensive load is weighted according to the temperature load coefficient to obtain the item, and the temperature scheduling index is obtained;

[0156] The pressure scheduling index is constructed: the difference between the global average pressure and the target pressure is converted according to the pressure scale, and the standardized load deviation obtained by standardizing the comprehensive load is weighted according to the pressure load coefficient to obtain the item, and the pressure scheduling index is obtained;

[0157] The conversion coefficient corresponding to the temperature scale and the pressure scale is set, and the conversion coefficient is equal to the inverse of the corresponding scale;

[0158] According to the scheduling index and the conversion coefficient, the global temperature control increment and the global pressure control increment are generated in the opposite direction of the scheduling index, and the corresponding analytical relationship is given.

[0159] Further specific implementation operations include:

[0160] The temperature regulation index function and the pressure regulation index function are constructed respectively, specifically: , ; wherein, is the temperature scheduling index at time ; is the pressure scheduling index at time ;

[0161] The temperature scale and the pressure scale are used to set , ; wherein, represents the inverse of the temperature scale, which is used to convert the dimensionless temperature scheduling index into the temperature control variable; is the inverse of the pressure scale, which is used to convert the dimensionless pressure scheduling index into the pressure control variable;

[0162] The temperature control variable is denoted as The pressure control variable is denoted as Analytical expressions for the temperature and pressure control quantities are constructed respectively, as follows: , ;in, For at any time Temperature control increments calculated from scheduling indicators; For at any time Pressure control increment.

[0163] By calculating the scheduling indices for temperature and pressure separately within the Lagrange framework, and converting the dimensionless indices into actual control increments based on pre-set conversion coefficients (i.e., scale reciprocals), a one-to-one mapping between decision quantities and physical quantities is ensured. At the same time, the standardized deviation of the comprehensive load is weighted and superimposed into the scheduling indices according to the load coefficient, so that safety constraints directly affect the magnitude of the control quantity, thus solving the problem that it is difficult to balance safety and accuracy in control increments under multi-objective environments.

[0164] The allocation of temperature and pressure control values ​​at each measuring point according to the ratio of the main zone to the adjustment zone specifically includes:

[0165] Based on the number of measuring points in the main area set and the adjustment area set, the proportion of the main area and the proportion of the adjustment area are calculated respectively.

[0166] For each measuring point in the main area, the global temperature control increment is allocated to the temperature control quantity of each measuring point according to the proportion of the main area, and the global pressure control increment is allocated to the pressure control quantity of each measuring point according to the proportion of the main area.

[0167] For each measuring point within the regulation zone, the global temperature control increment is allocated to the temperature control quantity of each measuring point according to the regulation zone ratio, and the global pressure control increment is allocated to the pressure control quantity of each measuring point according to the regulation zone ratio.

[0168] Further specific implementation steps include:

[0169] Based on the set of physical measurement points in the main area With the set of physical measurement points in the regulation zone The allocation ratio factors for the main area and the adjustment area are constructed separately, as follows: , ;in, For a moment The proportion of the number of measuring points in the main area to the total number of measuring points; For a moment The proportion of measuring points in the adjustment zone to the total number of measuring points;

[0170] For physical measurement points in the main area The allocated temperature control and pressure control values ​​are as follows: , ;in, , respectively are the temperature control increment and the pressure control increment of the physical measuring point in the main area at time ;

[0171] For the physical measuring point in the adjustment area, the allocated temperature control increment and the pressure control increment are respectively: , ; wherein, , respectively are the temperature control increment and the pressure control increment of the physical measuring point in the adjustment area at time .

[0172] First, the partition ratio is calculated according to the number of measuring points in the main area and the adjustment area; then the global temperature and pressure control increments are allocated to each measuring point according to the corresponding ratio, realizing the adaptive response of the group; when some measuring points are abnormal or out of line, the size of the partition set automatically changes, and the allocation ratio adjusts accordingly, so as to ensure that the fault point will not affect the overall allocation.

[0173] The temperature and pressure regulation and state update in the fixed control period are specifically:

[0174] The control period length is set to be a preset number of seconds;

[0175] At the end of each control period, for each measuring point, the current temperature and the current pressure are updated according to the corresponding temperature control increment and the pressure control increment, to obtain the temperature and pressure at the next time;

[0176] Each temperature actuator and each pressure actuator act according to the temperature control increment of the corresponding measuring point and the pressure control increment of the corresponding measuring point.

[0177] Further specific implementation operations include:

[0178] The control period length is set to be seconds;

[0179] The temperature and pressure of the physical measuring point at time are updated to: , ; wherein, , respectively are the temperature and pressure of the physical measuring point at the next time ;

[0180] Each temperature actuator and pressure actuator act according to the and of the corresponding physical measuring point The signal acts to realize the coordinated scheduling of the temperature field and pressure field of the gas (coke oven gas or coal gasification gas) in the coal gas separation-shift process section.

[0181] By setting a uniform cycle length, the real-time nature of the control instructions is ensured, and the system has a predictable execution rhythm; at the end of each cycle, the temperature and pressure of each measuring point are updated by the control increment, and the physical adjustment is implemented through the actuator, realizing the closed-loop feedback of data-driven decision-making.

[0182] The storage of operation data and the implementation of dual factor switching update specifically include:

[0183] At the end of each control cycle, the operation data of the current cycle is stored, including the temperature and pressure of each measuring point, the main zone set and the adjustment zone set, the global temperature control increment and the global pressure control increment, the temperature control amount of each measuring point and the pressure control amount of each measuring point, the comprehensive load value and the standardized load deviation, and the dual factor value.

[0184] According to whether the comprehensive load of the current cycle exceeds the upper limit of the comprehensive load safety, the dual factor of the next cycle is switched and updated between zero and one.

[0185] Further specific implementation operations include:

[0186] At the end of each control cycle, the operation data of the current cycle is stored, including but not limited to: the temperature and pressure of all physical measuring points, the calculated main zone physical measuring point and adjustment zone physical measuring point set , the temperature control amount and pressure control amount , and the measuring point control amount , the current load function value and the standardized load function value , and the current dual factor .

[0187] The dual factor is updated, specifically: ; wherein, is the dual factor value at the next moment , which is switched only according to whether the load at the current moment is over-limit.

[0188] At the end of each cycle, the key data including the state of the measuring point, the partition result, the control increment, the comprehensive load deviation and the dual factor are stored, which provides a complete data base for the subsequent diagnosis, optimization and model retraining. Meanwhile, only according to whether the current load is over limit, the dual factor is switched, and the dynamic intervention of the security constraint is realized.

[0189] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0190] The above only describes the preferred embodiments of the present application. It should be noted that for those skilled in the art, without departing from the technical principles of the present application, a number of improvements and refinements can be made, which should also be considered as the protection scope of the present application.

Claims

1. A data-driven control method for coal chemical conversion processes, characterized in that, include: Temperature and pressure sensors were deployed and calibrated using a two-point method to form an independent voltage acquisition channel and calibration coefficient. The temperature and pressure samples from each measuring point are combined to form a temperature and pressure state vector, and zoning indicators are established and the main zone and regulation zone are divided according to the zoning threshold. Collect running samples, obtain the total dimensionless target value, and form a dimensionless error model; By introducing comprehensive load constraints, a dimensionless Lagrangian function and dual factor are constructed. Generate temperature and pressure scheduling indicators, set conversion factors, and obtain global temperature control increment and global pressure control increment. The temperature and pressure control values ​​for each measuring point are allocated according to the ratio of the main zone to the regulation zone. Perform temperature and pressure regulation and complete status updates within a fixed control cycle; Store runtime data and perform dual factor switching updates.

2. The data-driven coal chemical conversion process control method according to claim 1, characterized in that, The temperature and pressure sensors are deployed and calibrated using a two-point method, forming an independent voltage acquisition channel and calibration coefficients, specifically including: Multiple physical measurement points are arranged at equal intervals along the reaction pipeline and numbered sequentially. At each physical measurement point, a temperature sensor and a pressure sensor are installed and connected to an independent voltage acquisition channel. The total number of temperature sensors and the total number of pressure sensors are the same as the total number of physical measurement points. Temperature calibration: Select two sets of known temperatures and record the original voltage at each measuring point under the two sets of temperatures; If the two voltages are the same, reselect the calibration point or check the connection. If they are different, calculate the linear slope and linear intercept of the temperature channel at each measuring point using the two-point method, and establish a linear conversion relationship from voltage to temperature. Pressure calibration: Select two sets of known pressures and record the original voltage at each measuring point under the two sets of pressures; If the two voltages are the same, reselect the calibration point or check the connection. If they are different, calculate the linear slope and linear intercept of the pressure channel at each measuring point using the two-point method to establish a linear conversion relationship from voltage to pressure. During sampling, the voltage of the temperature channel is converted into the actual temperature and the voltage of the pressure channel is converted into the actual pressure according to their respective linear conversion relationships, thus forming the real-time temperature and real-time pressure of each measuring point.

3. The data-driven coal chemical conversion process control method according to claim 2, characterized in that, The combined temperature and pressure sampling at each measuring point forms a temperature and pressure state vector, establishes zoning indicators, and divides the main zone and regulation zone according to zoning thresholds. Specifically, this includes: At any sampling moment, the real-time temperature and real-time pressure of all measuring points are combined into a temperature and pressure state vector according to the measuring point number; Calculate the global average temperature and global average pressure; Based on the temperature and pressure scales, for each measuring point, the deviation of the temperature at each measuring point relative to the global average temperature and the deviation of the pressure at each measuring point relative to the global average pressure are calculated. After normalization according to the corresponding scale, the results are summed to generate a zoning index for each measuring point. Within a 600-second time window after system startup, all measurement point zoning indicators are collected, and the median is calculated as the zoning threshold. At any given time, if the zoning index of a certain measuring point is not greater than the zoning threshold, the measuring point is assigned to the main zone set; if the zoning index of a certain measuring point is greater than the zoning threshold, the measuring point is assigned to the adjustment zone set.

4. The data-driven coal chemical conversion process control method according to claim 3, characterized in that, The process of collecting and running samples, obtaining the total dimensionless target value, and forming a dimensionless error model specifically includes: Within a 600-second time window after system startup, temperature and pressure sample sets are collected. The 5th and 95th percentiles of the temperature sample and the 5th and 95th percentiles of the pressure sample were obtained respectively. Obtain the minimum resolvable temperature difference of the temperature sensor and the minimum resolvable pressure difference of the pressure sensor; Construct temperature and pressure scales: use the quantile difference when the corresponding quantile difference is not zero, and use the minimum resolvable temperature difference or the minimum resolvable pressure difference when the corresponding quantile difference is zero. For each measuring point, based on the target temperature and target pressure set by the process, the data is first normalized according to the corresponding scale, and then the temperature deviation and pressure deviation are squared and added together to obtain the bivariate dimensionless deviation of each measuring point. The dimensionless deviation of each measuring point in the main region set is averaged to obtain the main region error; the dimensionless deviation of each measuring point in the adjustment region set is averaged to obtain the adjustment region error; the main region error and the adjustment region error are added together to obtain the total dimensionless target value.

5. The data-driven coal chemical conversion process control method according to claim 4, characterized in that, The introduction of comprehensive load constraints and the construction of a dimensionless Lagrangian function and dual factor specifically include: Within a 600-second time window after system startup, temperature and pressure sample sets are collected. The 5th and 95th percentiles of the temperature sample and the 5th and 95th percentiles of the pressure sample were obtained respectively. Obtain the minimum resolvable temperature difference of the temperature sensor and the minimum resolvable pressure difference of the pressure sensor; Construct temperature and pressure scales: use the quantile difference when the corresponding quantile difference is not zero, and use the minimum resolvable temperature difference or the minimum resolvable pressure difference when the corresponding quantile difference is zero. For each measuring point, based on the target temperature and target pressure set by the process, the data is first normalized according to the corresponding scale, and then the temperature deviation and pressure deviation are squared and added together to obtain the bivariate dimensionless deviation of each measuring point. The dimensionless deviation of each measuring point in the main region set is averaged to obtain the main region error; the dimensionless deviation of each measuring point in the adjustment region set is averaged to obtain the adjustment region error; the main region error and the adjustment region error are added together to obtain the total dimensionless target value.

6. The data-driven coal chemical conversion process control method according to claim 5, characterized in that, The generation of temperature and pressure scheduling indices, setting conversion factors, and obtaining global temperature control increments and global pressure control increments specifically includes: Constructing temperature scheduling indicators: The temperature scheduling indicators are obtained by converting the difference between the global average temperature and the target temperature using a temperature scale, and by superimposing the standardized load deviation obtained by standardizing the comprehensive load according to the temperature load coefficient. The pressure scheduling index is constructed by converting the difference between the global average pressure and the target pressure using a pressure scale, and then adding the standardized load deviation obtained by standardizing the comprehensive load and weighting it by the pressure load coefficient. Set conversion factors corresponding to temperature and pressure scales. The conversion factor is equal to the reciprocal of the corresponding scale. Based on the scheduling indicators and conversion coefficients, global temperature control increments and global pressure control increments are generated in the opposite direction to the scheduling indicators, and the corresponding analytical relationships are given.

7. The data-driven coal chemical conversion process control method according to claim 6, characterized in that, The allocation of temperature and pressure control values ​​at each measuring point according to the ratio of the main zone to the adjustment zone specifically includes: Based on the number of measuring points in the main area set and the adjustment area set, the proportion of the main area and the proportion of the adjustment area are calculated respectively. For each measuring point in the main area, the global temperature control increment is allocated to the temperature control quantity of each measuring point according to the proportion of the main area, and the global pressure control increment is allocated to the pressure control quantity of each measuring point according to the proportion of the main area. For each measuring point within the regulation zone, the global temperature control increment is allocated to the temperature control quantity of each measuring point according to the regulation zone ratio, and the global pressure control increment is allocated to the pressure control quantity of each measuring point according to the regulation zone ratio.

8. The data-driven coal chemical conversion process control method according to claim 7, characterized in that, The process of performing temperature and pressure regulation and completing state updates within a fixed control cycle specifically includes: Set the control cycle length to the preset number of seconds; At the end of each control cycle, the current temperature and pressure at each measuring point are updated according to the corresponding temperature control and pressure control values ​​to obtain the temperature and pressure at the next moment. Each temperature actuator and each pressure actuator operates according to the temperature control value and pressure control value of the corresponding measuring point, respectively.

9. A data-driven coal chemical conversion process control method according to claim 8, characterized in that, The storage of runtime data and the execution of dual factor switching updates specifically include: At the end of each control cycle, store the operating data for that cycle, including the temperature and pressure of each measuring point, the main zone set and the regulation zone set, the global temperature control increment and the global pressure control increment, the temperature control quantity and the pressure control quantity of each measuring point, the comprehensive load value and the standardized load deviation, and the value of the dual factor. Based on whether the current period's comprehensive load exceeds the comprehensive load safety limit, the dual factor for the next period is switched and updated between zero and one.

Images