A method and system for regulating an electrically operated regulating valve in a steam line

Abstract

The present application belongs to the technical field of steam pipeline regulation, and discloses a kind of regulating method and regulating system of electric regulating valve in steam pipeline, through the multi-sensor redundancy design of acquisition module and fusion noise reduction algorithm, accurately capture key parameters such as steam dryness, enthalpy, etc., adjustment module constructs the five-parameter collaborative dynamic model of temperature, pressure, dryness, enthalpy and phase change rate, with three-level adaptive regulation logic of benchmark regulation, working condition mutation correction and phase change compensation, combined with closed-loop feedback mechanism dynamic correction parameter, can quickly respond to working condition mutation scene, effectively suppresses overshoot phenomenon, solves the regulation lag problem under the scene of unbalanced phase change, meets the requirement of high-precision production process on steam parameter;Operation and maintenance early warning module integrates a variety of monitoring units, captures various fault characteristic signals, constructs steam special fault characteristic database, and can accurately locate fault type and position by extracting fault characteristics.

Description

Technical Field

[0001] This invention belongs to the field of steam pipeline regulation technology, specifically a method and system for regulating an electric regulating valve in a steam pipeline. Background Technology

[0002] Steam is a crucial heat energy carrier in industrial production, widely used in chemical, power, textile, and food processing industries. The stability of its pipeline operation directly affects production efficiency, product quality, and energy utilization efficiency. As the core actuator for flow and pressure regulation in steam pipelines, the electric regulating valve plays a key role in optimizing the overall operating conditions of the steam system due to its regulation accuracy, response speed, and operational stability. Currently, electric regulating systems for steam pipelines still face the following technical challenges: During the transmission of steam through pipelines, temperature, pressure and phase change processes interact. Existing systems mostly use single-parameter linear regulation logic, which is difficult to match the synergistic effects of the above-mentioned multiple parameters. Under variable load and variable operating conditions, it is easy to encounter situations such as regulation lag, overshoot and excessive steady-state error, which cannot meet the strict requirements of high-precision production processes for steam parameters. The existing system can only passively report valve failures and cannot monitor the gradual failure process such as valve core wear and seal aging in real time. It also cannot predict failure trends by combining historical operating data, which leads to delayed operation and maintenance decisions and can easily cause safety hazards such as sudden changes in pipeline pressure and steam leakage, while increasing unplanned downtime losses. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for regulating an electric regulating valve in a steam pipeline, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides a regulating system for an electric regulating valve in a steam pipeline, comprising a data acquisition module, a regulating module, an operation and maintenance early warning module, a drive module, a control module, a fault-tolerant control module, and a collaborative optimization module; Preferably, the acquisition module integrates a pressure sensor, temperature sensor, phase change state sensor, and pipeline stress sensor with dual-redundancy design, as well as a steam dryness detection unit based on a throttling and infrared composite method; wherein, each redundant sensor of the pressure sensor, temperature sensor, phase change state sensor, and pipeline stress sensor synchronously acquires the same parameter to achieve data consistency verification and provide a basis for fault prediction for the fault-tolerant control module; The steam dryness detection unit releases steam to a superheated state through a throttling orifice, and simultaneously collects temperature and pressure data after depressurization to calculate enthalpy. Combined with infrared thermometry, it corrects the detection error in the wet steam region, and realizes detection within the preset dryness range. The phase change capture adopts a three-pronged technology of microwave resonance, infrared thermometry and dryness detection. Based on the dual resistance model, it identifies the gas-liquid double film resistance distribution at the steam condensation phase change interface, and captures the phase change critical point, phase change rate and non-equilibrium interface temperature jump characteristics. Data processing employs a fusion noise reduction algorithm based on Kalman filtering and wavelet thresholding to preprocess data collected synchronously from multiple sensors, removing interference signals from pipeline vibration and heat dissipation, and outputting standardized datasets of temperature, pressure, phase change state, pipeline stress, steam dryness, and enthalpy. These datasets are then pushed to the regulation module and the operation and maintenance early warning module, while historical data is stored.

[0005] Preferably, the adjustment module receives standardized data output by the acquisition module, establishes a five-parameter collaborative dynamic model of temperature, pressure, dryness, enthalpy and phase change rate through deep learning training, introduces a dual-resistance model to quantify the thermal resistance correlation of the gas-liquid film at the phase change interface, and characterizes the correspondence between comprehensive parameters and valve adjustment amount under different working conditions. The deep learning model adopts a fully connected neural network architecture, with the specific layer settings as follows: The input layer has a dimension of 5, corresponding to 5 parameters: temperature, pressure, dryness, enthalpy, and phase transition rate. There are 3 hidden layers: the first layer has 64 neurons, the second layer has 32 neurons, and the third layer has 16 neurons. The output layer has a dimension of 1, corresponding to the valve adjustment amount. The activation function of each layer is the ReLU function, and the output layer uses the Sigmoid function to normalize the adjustment amount to the opening range of [0, 100%].

[0006] The model training steps are as follows: 1) Data set partitioning: The historically collected standardized dataset is divided into training set, validation set and test set in a ratio of 7:2:1, and the data covers the full operating condition range of steam dryness 0.9~1.0, temperature 100~400℃ and pressure 0.1~10MPa; 2) Training parameter settings: The optimizer is Adam, the initial learning rate is 0.001, the exponential decay strategy is adopted, the decay rate is 0.95, the decay period is 50 rounds, the batch size is set to 32, the number of iterations is 200 rounds, and the loss function is mean squared error. 3) Training termination conditions: Training stops when the loss function value on the validation set does not decrease for 10 consecutive rounds or the number of iterations reaches 200 rounds, and the final model's fitting accuracy on the test set is ≥98.5%.

[0007] The logic for combining the model with the steam pipeline scenario is as follows: the input parameters are all standardized core steam parameters output by the acquisition module, and the output adjustment amount is directly mapped to the valve opening percentage. By introducing the gas-liquid film thermal resistance calculated by the dual-resistance model as an implicit constraint condition of the model, it is ensured that the output adjustment amount is adapted to the non-equilibrium phase change characteristics of steam.

[0008] The design incorporates a three-level adaptive regulation algorithm: baseline regulation, operating condition abrupt change correction, and phase change compensation. The operating condition abrupt change identification unit detects scenarios including sudden load changes and phase change bursts by measuring dryness and pressure change rate, prioritizing the activation of a rapid response strategy. Under non-abrupt operating conditions, baseline regulation is executed, while the phase change compensation submodule corrects the valve regulation rate in real time based on the gas-liquid film resistance data output by the collaborative dynamic model. A closed-loop feedback mechanism is established to receive valve opening and operating current data from the drive module in real time, and dynamically corrects model parameters by combining the dryness and enthalpy values ​​from the acquisition module, ensuring regulation accuracy.

[0009] Preferably, the operation and maintenance early warning module receives steam condition data from the acquisition module and operating status data from the drive module, and establishes an integrated intelligent operation and maintenance system that combines steam condition, fault characteristics, life prediction and operation and maintenance decision-making throughout the entire life cycle; it integrates a valve core vibration sensor, a sealing cavity pressure sensor, an actuator current sensor, a high-frequency acoustic valve core erosion monitoring unit and a sealing cavity temperature field distribution acquisition unit to capture valve core erosion characteristic signals, high-temperature aging data of seals and actuator operating status parameters; A feature database of steam faults, including valve core erosion, high-temperature aging of seals, and valve seat deformation caused by steam scouring, is constructed. A convolutional neural network is used to extract fault features from multi-dimensional monitoring signals, and rule-based reasoning is combined to locate the fault type and location. The specific structure of the convolutional neural network is as follows: The input layer is a time-series data matrix of multi-dimensional monitoring signals, with dimensions of N×M, where N is the number of sampling points and M is the number of monitoring parameters. M=5, corresponding to valve core vibration, sealing cavity pressure, actuator current, valve core erosion acoustic signal, and sealing cavity temperature field distribution. The convolutional layer consists of two layers. The first layer has 32 convolutional kernels with a kernel size of 3×1, and the second layer has 64 convolutional kernels with a kernel size of 2×1. Both layers use the SamePadding pattern. The pooling layer uses max pooling with a pooling kernel size of 2×1 and a stride of 2. There is one fully connected layer with 128 neurons. The output layer is a fault feature vector with a dimension of 1×K, where K=3, corresponding to the feature weights of three core faults: valve core erosion, seal aging, and valve seat deformation.

[0010] The logic of combining the model with the steam fault diagnosis scenario: The input data is five types of monitoring signals collected in real time by the operation and maintenance early warning module. After standardization preprocessing (normalized to the [-1, 1] interval), the data is input into the CNN. The time-frequency joint features of the signal are extracted through the convolutional layer, the pooling layer compresses redundant information, and the fully connected layer maps the signal into a fault feature vector. The output feature vector is matched with the standard feature template in the fault feature database using cosine similarity. The similarity threshold is set to 0.85. Combined with rule-based reasoning, for example, if the valve core vibration feature weight is greater than 0.6 and the acoustic signal feature weight is greater than 0.5, it is determined to be a valve core erosion fault, thus realizing the fault type and location.

[0011] A model is established based on an improved LSTM neural network to correlate operating parameters with failure rates. The model is input with historical operating data and real-time failure signals from the acquisition module to predict the remaining service life of key components, including valve cores and seals. Differentiated operation and maintenance plans are generated for different steam dryness ranges, and operation and maintenance priorities and operation guidelines are output.

[0012] The specific improvements and structure of the improved LSTM neural network are as follows: 1) Improvement: An attention mechanism layer is added between the forget gate and the input gate of the traditional LSTM to enhance the influence of key operating parameters on lifetime prediction through weight allocation; 2) Network structure: The input layer has a dimension of 5, corresponding to pressure, temperature, dryness, runtime, and real-time failure signals. The attention mechanism layer adopts the Bahdanau attention mechanism. There are 2 LSTM hidden layers with 64 hidden units in each layer and a dropout rate of 0.2. There is 1 fully connected layer. The output layer has a dimension of 1, corresponding to the remaining service life of key components in hours.

[0013] Model training parameters and steps: 1) Training dataset: Contains 5,000 sets of historical operating data, covering a 3-year operating cycle, including component life data under different failure modes, and divided into training set and test set in an 8:2 ratio; 2) Training parameters: The optimizer used is RMSProp, the learning rate is 0.0008, the batch size is set to 64, the number of iterations is 150, and the loss function is the mean absolute error. 3) Training and validation: A validation set test is conducted every 10 rounds. Training is stopped when MAE ≤ 50 hours. The final model prediction error is ≤ ±3%.

[0014] The logic for combining the model with the steam pipeline operation and maintenance scenario is as follows: the input data is directly related to the core operating parameters of the steam pipeline and the operating status of the components. The attention mechanism highlights the accelerating effect of severe operating conditions such as dryness <0.95 on component failure. The output remaining life data is mapped to different operation and maintenance strategies according to the dryness range (≥0.95, 0.9~0.95, <0.9), ensuring that the prediction results are directly related to the differentiated operation and maintenance solutions.

[0015] Preferably, the drive module receives the opening adjustment command from the adjustment module and drives the valve to complete the action; the drive architecture adopts an integrated design of permanent magnet synchronous motor and precision ball screw, combined with vector control technology of model predictive control, to achieve millisecond-level response of valve opening; the actuator housing adopts a double-layer heat insulation structure, the sealing components are made of high-temperature resistant fluororubber material, and an integrated sealing cavity dehumidification unit is used.

[0016] An integrated regenerative energy recovery unit feeds back the regenerative energy from the motor to the system power supply during valve deceleration and stable operation. Combined with the steam dryness adaptation adjustment strategy output by the collaborative optimization module, it reduces energy consumption caused by ineffective regulation. An actuator winding temperature monitoring unit is configured to collect real-time thermal operation data of the drive mechanism and feed it back to the regulation module, operation and maintenance early warning module, and collaborative optimization module.

[0017] Preferably, the edge computing unit of the control module deploys a collaborative adjustment model and a fault early warning algorithm to process the sensing data from the acquisition module in real time, receive the adjustment parameters from the adjustment module, and assist in generating emergency adjustment commands. The cloud platform realizes centralized monitoring of the collaborative operating conditions of the multi-pipeline system, remote optimization of adjustment parameters, review and distribution of operation and maintenance plans pushed by the operation and maintenance early warning module, and historical data tracing. At the same time, it establishes a production load prediction model based on the historical operating condition data of the acquisition module, outputs short-term production load prediction results to the collaborative optimization module, and completes global control decisions.

[0018] It integrates industrial Ethernet and 5G dual-mode communication interfaces, supports multi-protocol adaptation, and develops a dedicated protocol for data interaction in steam systems. It adopts end-to-cloud encrypted transmission and data desensitization technology, and designs a monitoring interface that adapts to multiple terminals, supporting real-time display and remote adjustment of steam-specific parameters.

[0019] Preferably, the fault-tolerant control module is equipped with a primary and backup dual-core control unit, which synchronizes the parameters of the collaborative adjustment model of the adjustment module, the operation and maintenance monitoring data of the operation and maintenance early warning module, and the adjustment instructions generated by the adjustment module in real time. It adopts a fault prediction trigger switching mechanism based on actuator current fluctuation and consistency verification of redundant sensor data of the acquisition module, and completes hot switching in advance when abnormal signs appear in the primary control unit. At the same time, the control authority is synchronized to the cloud backup unit of the management and control module.

[0020] It is equipped with a dual power supply system consisting of a main power supply and a backup lithium battery, featuring voltage monitoring and automatic switching functions, and an integrated voltage fluctuation compensation unit to adapt to grid fluctuations caused by the start-up and shutdown of the steam system. It also features a mechanical emergency adjustment mechanism adapted to high-pressure steam, employing a worm gear drive design. In the event of a complete failure of the electronic system, the valve opening can be manually adjusted and locked, while simultaneously triggering a cloud-based safety warning from the control module. The cloud platform provides real-time backup of critical system operating data and adjustment strategies.

[0021] Preferably, the collaborative optimization module integrates the steam temperature, pressure, dryness and enthalpy parameters collected by the acquisition module, the energy consumption data of the driving module and the component loss data of the operation and maintenance early warning module to establish a database linking energy consumption and operation and maintenance costs. A multi-objective genetic algorithm is used to calculate the energy consumption and maintenance cost balance point under different production loads and dryness ranges based on the energy consumption data of the driving module and the component loss data of the operation and maintenance early warning module. Combined with the production load prediction, the collaborative adjustment parameters of the adjustment module are optimized in advance to achieve the matching of steam supply and production demand and reduce ineffective energy consumption. A linkage report of energy consumption, operation and maintenance cost and dryness is generated, and optimization suggestions that take into account energy saving and low-cost operation and maintenance are output and pushed to the control module.

[0022] The present invention also provides a method for regulating an electrically operated regulating valve in a steam pipeline, based on the above-described system, comprising the following steps: Data Acquisition Phase: By integrating pressure, temperature, phase change state, pipeline stress sensors, and a steam dryness detection unit based on a throttling and infrared composite method, the temperature, pressure, phase change state, pipeline stress, and dryness parameters in the steam pipeline are acquired simultaneously. The enthalpy value is calculated through the dryness detection unit. The acquired and calculated data are preprocessed using a Kalman filter and wavelet threshold fusion noise reduction algorithm to output a standardized dataset. Adjustment phase: Based on the dual-resistance model, a five-parameter collaborative dynamic model is constructed. Standardized data from the acquisition phase is received. A three-level adaptive adjustment logic of benchmark adjustment, working condition change correction and phase change compensation is used to generate adjustment commands to drive the electric actuator to complete the valve opening adjustment. At the same time, the opening and current data fed back by the actuator are received, and the model parameters are dynamically corrected to form a closed-loop adjustment. Operation and maintenance early warning stage: Integrate steam condition data from the acquisition stage with actuator operation data from the regulation stage, extract fault features through convolutional neural networks, diagnose steam faults including valve core erosion and seal aging, predict the remaining life of key components based on the improved LSTM model, and generate differentiated operation and maintenance solutions. During the control phase: the edge computing unit processes the sensing data in real time and assists in generating emergency adjustment instructions; the cloud platform realizes centralized monitoring of multi-pipeline operating conditions, remote optimization of adjustment parameters, and review and issuance of operation and maintenance plans; and achieves end-to-cloud data interaction through encrypted communication protocols. Redundancy and fault tolerance protection phase: Real-time monitoring of the operating status of core modules; when abnormal signs are detected, millisecond-level hot switching of the main and backup control units is triggered and permissions are synchronized to the cloud backup; when the main power supply is abnormal, it automatically switches to the backup lithium battery; when the electronic system fails, the mechanical emergency mechanism is activated for manual adjustment and safety warnings are triggered. Collaborative optimization phase: The energy consumption and operation and maintenance cost balance point is calculated through a multi-objective genetic algorithm, the adjustment parameters are optimized and pushed to the adjustment module and the cloud, so as to achieve the matching of steam supply and production demand.

[0023] The beneficial effects of this invention are as follows: 1. This invention accurately captures key parameters such as steam dryness and enthalpy by using a multi-sensor redundancy design and fusion noise reduction algorithm in the acquisition module. The adjustment module constructs a five-parameter collaborative dynamic model of temperature, pressure, dryness, enthalpy and phase change rate. Combined with a three-level adaptive adjustment logic of benchmark adjustment, operating condition change correction and phase change compensation, and a closed-loop feedback mechanism to dynamically correct parameters, it can quickly respond to operating condition change scenarios, effectively suppress overshoot, solve the adjustment lag problem in non-equilibrium phase change scenarios, and meet the requirements of high-precision production processes for steam parameters.

[0024] 2. The operation and maintenance early warning module of this invention integrates multiple monitoring units, captures various fault characteristic signals, and constructs a steam-specific fault characteristic database. By extracting fault characteristics, it can accurately locate the fault type and location. Based on the correlation model between operating parameters and failure rate, it can predict the remaining life of key components and generate differentiated operation and maintenance plans for different operating conditions, transforming passive operation and maintenance into proactive prediction, effectively avoiding safety risks such as sudden changes in pipeline pressure and steam leakage, reducing unplanned downtime losses, and lowering overall operation and maintenance costs.

[0025] 3. The fault-tolerant control module of this invention constructs a multi-dimensional redundancy protection system through a dual-core control unit, a dual power supply system, and a mechanical emergency adjustment mechanism. It can quickly switch when the core module is abnormal or the power supply fluctuates, ensuring that the adjustment process is uninterrupted. The mechanical emergency adjustment can ensure the basic operation of the system. At the same time, the drive module integrates a regenerative energy recovery unit, and the collaborative optimization module integrates steam parameters, energy consumption data, and component loss data. Through a multi-objective genetic algorithm, it finds the balance point between energy consumption and operation and maintenance costs, and optimizes the adjustment parameters in combination with production load prediction, reducing ineffective energy consumption and resource waste. This improves the operational reliability of the system under high temperature, high pressure, and complex operating conditions, and achieves the dual goals of energy saving and low-cost operation and maintenance. Attached Figure Description

[0026] Figure 1 This is a flowchart of the adjustment system of the present invention;

[0027] Figure 2 This is a flowchart illustrating the core regulating process of the electric regulating valve for steam pipelines in this invention.

[0028] Figure 3 This is a flowchart illustrating the entire lifecycle fault early warning and maintenance process of this invention.

[0029] Figure 4 This is a flowchart illustrating the redundancy and fault tolerance guarantee process of the system in this invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] like Figures 1 to 4 As shown in the figure, this embodiment of the invention provides a regulating system for an electric regulating valve in a steam pipeline, including a data acquisition module, a regulating module, an operation and maintenance early warning module, a drive module, a control module, a fault-tolerant control module, and a collaborative optimization module. The specific implementation of each module is as follows: The acquisition module integrates a pressure sensor, temperature sensor, phase change state sensor, pipeline stress sensor, and steam dryness detection unit based on a throttling and infrared composite method, all with dual-redundant design. Each redundant sensor synchronously acquires the same parameter to achieve data consistency verification, providing a basis for fault prediction for the fault-tolerant control module. The steam dryness detection unit releases steam to a superheated state through a throttling orifice, and simultaneously collects temperature and pressure data after depressurization to calculate enthalpy. Combined with infrared thermometry, it corrects the detection error in the wet steam region, achieving high-precision detection within the dryness range of 0.9~1.0 with an error ≤±0.5%, avoiding the detection blind spot of the single throttling method in low-pressure, high-dryness scenarios. The phase change capture adopts a triple technology of microwave resonance, infrared thermometry, and dryness detection. Based on the dual-resistance model, it identifies the gas-liquid double-film resistance distribution at the steam condensation phase change interface, accurately capturing the phase change critical point, phase change rate, and non-equilibrium interface temperature jump characteristics. The dual-resistance model is a superposition model of gas-liquid two-phase heat transfer resistance, where the total resistance equals the gas film resistance plus the liquid film resistance. The gas film resistance is calculated using the gas film thickness detected by microwave resonance and combined with the real-time steam thermal conductivity; the liquid film resistance is calculated using the liquid film thickness derived from infrared thermometry and combined with the fixed thermal conductivity of condensate. The final output total resistance data is used for subsequent adjustment and compensation.

[0032] Data processing employs a fusion noise reduction algorithm based on Kalman filtering and wavelet thresholding to preprocess data collected synchronously from multiple sensors, removing interference signals from pipeline vibration and heat dissipation, and outputting standardized datasets of temperature, pressure, phase change state, pipeline stress, steam dryness, and enthalpy. These datasets are then pushed to the regulation module and the operation and maintenance early warning module, while historical data is stored.

[0033] The specific implementation logic of the fusion noise reduction algorithm is as follows: 1) Kalman filter preprocessing: The state equation is set as follows The observation equation is set as ;

[0034] in Let k be the system state variable at time k, corresponding to the filtered value of the collected steam parameter, with a standardized range of [0, 1]. , is the state transition matrix of the constant velocity model; Let k-1 be the system state variables; =0, indicating the control input matrix; =0, which represents the control input at time k; The noise at time k follows a Gaussian distribution with a variance of 0.01. The observation value at time k is the raw data collected by the sensor. =1, which is the observation matrix, directly mapping state variables to observation values; The noise observed at time k follows a Gaussian distribution, and the variance is obtained by statistical analysis of historical sensor data: 0.005 for pressure sensor, 0.003 for temperature sensor, and 0.002 for dryness detection unit. The output is a filtered data sequence.

[0035] 2) Wavelet thresholding denoising: The Kalman filter output data is decomposed into 3 levels using a db4 wavelet basis. Low-frequency coefficients are directly retained, while high-frequency coefficients are decomposed using an improved soft thresholding function. ;

[0036] These are the high-frequency coefficients after wavelet decomposition; These are the high-frequency coefficients after thresholding. Wavelet threshold; The noise standard deviation is calculated statistically from the raw sensor data. The sampling length for a single set of data; the threshold calculation formula is: ;

[0037] 3) Fusion Weight Allocation: The Kalman filter result and the wavelet denoising result are fused according to dynamic weights, with weight coefficients... SNR / 100; where The weighting coefficients for the Kalman filter result range from [0, 1]; SNR is the signal-to-noise ratio of the current data, in dB, calculated by the ratio of signal power to noise power. The final output is standardized data to ensure the signal-to-noise ratio of the denoised data. .

[0038] The adjustment module receives standardized data output by the acquisition module, establishes a five-parameter collaborative dynamic model of temperature, pressure, dryness, enthalpy and phase change rate through deep learning training, introduces a dual-resistance model to quantify the thermal resistance correlation of gas-liquid film at the phase change interface, and characterizes the correspondence between comprehensive parameters and valve adjustment amount under different working conditions.

[0039] The design incorporates a three-level adaptive adjustment algorithm: baseline adjustment, operating condition abrupt change correction, and phase change compensation. The operating condition abrupt change identification unit detects scenarios such as sudden load changes and phase change bursts by measuring dryness and pressure change rate (thresholds can be dynamically configured), prioritizing the activation of a rapid response strategy and dynamically compressing the adjustment step size to 0.1%. Under non-abrupt operating conditions, baseline adjustment is executed, while the phase change compensation submodule corrects the valve adjustment rate in real time based on the gas-liquid film resistance data output by the collaborative dynamic model to suppress overshoot. A closed-loop feedback mechanism is established to receive valve opening and operating current data from the drive module in real time, and dynamically corrects model parameters by combining the dryness and enthalpy values ​​from the acquisition module to ensure adjustment accuracy, with steady-state errors ≤ ±0.1℃ and ±0.01MPa.

[0040] Phase change compensation rule: The greater the total resistance, the slower the valve adjustment rate. When the dryness is ≥0.97, adjust according to the reference rate (0.5% / ms); when the dryness is 0.9~0.97, the rate is reduced to 50% of the reference value; when the dryness is <0.9, the rate is reduced to 20% of the reference value to avoid overshoot.

[0041] The specific implementation of the benchmark adjustment is based on the comprehensive adjustment coefficient output by the five-parameter collaborative dynamic model. ,in This is the comprehensive adjustment coefficient, with a value range of [0, 1]. The temperature is standardized to map the actual temperature to the interval [0, 1]. To standardize the pressure value, map the actual pressure to... interval; The value is the steam dryness fraction, ranging from [0.9, 1.0]. This is a standardized value for enthalpy, mapping the actual enthalpy value to the interval [0, 1]. To standardize the phase transition rate, map the phase transition rate to the interval [0, 1].

[0042] Initial adjustment amount ,in This represents the single valve adjustment amount (in %), corresponding to the percentage change in valve opening. =5% is the maximum single adjustment step size;

[0043] Dynamic configuration rules for operating condition change thresholds: Dryness change rate threshold s (default value), pressure change rate threshold MPa / s; where This represents the change in dryness. The change over time is expressed in seconds (s). This represents the pressure change, expressed in MPa.

[0044] This threshold can be adjusted according to the production scenario via the cloud platform of the control module; the threshold is increased by 20% for chemical scenarios and decreased by 15% for food processing scenarios. When a detection is made... or If the threshold is exceeded and continues for two sampling cycles, the operating condition change correction strategy is triggered.

[0045] The operation and maintenance early warning module receives steam condition data from the acquisition module and operating status data from the drive module, and establishes an integrated intelligent operation and maintenance system that combines steam condition, fault characteristics, life prediction and operation and maintenance decision-making throughout the entire life cycle; it integrates a valve core vibration sensor, a sealing cavity pressure sensor, an actuator current sensor, a high-frequency acoustic valve core erosion monitoring unit and a sealing cavity temperature field distribution acquisition unit to capture valve core erosion characteristic signals, high-temperature aging data of seals and actuator operating status parameters; A feature database containing steam faults such as valve core erosion, high-temperature aging of seals, and valve seat deformation caused by steam scouring is constructed. A convolutional neural network is used to extract fault features from multi-dimensional monitoring signals, and rule-based reasoning is combined to achieve accurate location of fault type and location.

[0046] The structure of the steam-specific fault feature database is as follows: 1) Feature vector dimension: 12-dimensional, including valve core vibration (amplitude, frequency), sealing cavity pressure (deviation, fluctuation), actuator current (fluctuation, slope), high-frequency acoustic signal (peak value, energy), sealing cavity temperature (gradient, temperature difference), pipeline stress change rate, and actuator vibration acceleration. Each feature can be input into the model after standardization. 2) Standard fault characteristic template: valve core erosion (vibration amplitude > 30Hz and acoustic peak value > 8kHz), seal aging (sealing cavity pressure deviation > 0.08MPa and temperature gradient > 3℃ / cm), valve seat deformation (actuator current fluctuation > 4% and vibration amplitude > 25Hz);

[0047] The rule-based reasoning logic adopts a feature weight voting mechanism: when the number of features that meet the corresponding fault template in the 12-dimensional features is ≥8, it is determined to be this type of fault, and the positioning accuracy is ≤5cm; the fault confidence C is calculated as C = number of matching features / 12 × 100%; the number of matching features is the number of feature dimensions in the current monitoring signal that meet the fault template. 80% directly trigger maintenance alerts, 60% 80% were marked as suspected faults and are continuously monitored.

[0048] Based on an improved LSTM neural network, a correlation model between operating parameters and failure rate is established. The historical operating data (pressure, temperature, dryness, and running time) and real-time failure signals of the acquisition module are input to accurately predict the remaining service life of key components such as valve cores and seals. Differentiated operation and maintenance plans are generated for different steam dryness ranges. For example, when the dryness is <0.95, the replacement cycle of seals is shortened by 30%, and operation and maintenance priorities and operation guidelines are output.

[0049] The drive module receives the opening adjustment command from the adjustment module and drives the valve to complete the action. The drive architecture adopts an integrated design of permanent magnet synchronous motor and precision ball screw, combined with vector control technology of model predictive control, to achieve millisecond-level (response time ≤5ms) precise response of valve opening. The actuator housing adopts a double-layer heat insulation structure, the sealing components are made of high-temperature resistant fluororubber material, and an integrated sealing cavity dehumidification unit is used to effectively resist the high temperature and high pressure environment of steam and prevent component corrosion caused by steam leakage.

[0050] Vector control decomposes the motor current into excitation and torque components through coordinate transformation. The control torque component tracks the command, while the excitation component is set to 0 to achieve a fast response. The prediction step size is 1ms to ensure that the valve response time is ≤5ms.

[0051] An integrated regenerative energy recovery unit feeds back the regenerative energy from the motor to the system power supply during valve deceleration and stable operation. Combined with the steam dryness adaptation adjustment strategy output by the collaborative optimization module, it reduces energy consumption caused by ineffective regulation. An actuator winding temperature monitoring unit is configured to collect real-time thermal operation data of the drive mechanism and feed it back to the regulation module, operation and maintenance early warning module, and collaborative optimization module.

[0052] The edge computing unit of the control module deploys a collaborative adjustment model and a fault early warning algorithm. It is responsible for processing the sensing data of the acquisition module in real time, receiving the adjustment parameters of the adjustment module, and assisting in generating emergency adjustment commands. The response delay is ≤10ms, ensuring the real-time performance of local adjustment.

[0053] The cloud platform enables centralized monitoring of the collaborative operating conditions of multi-pipeline systems, remote optimization of adjustment parameters, review and distribution of operation and maintenance plans pushed by the operation and maintenance early warning module, and historical data traceability. Simultaneously, it establishes a production load prediction model based on historical operating data from the acquisition module, outputting short-term production load prediction results to the collaborative optimization module to complete global control decisions. It integrates industrial Ethernet and 5G dual-mode communication interfaces, supports adaptation to multiple protocols such as Modbus and OPCUA, and develops a dedicated data interaction protocol for the steam system to improve the stability of data transmission with various modules under high temperature and strong electromagnetic interference environments. Employing end-to-end encrypted transmission and data anonymization technology, it designs a monitoring interface compatible with multiple terminals such as computers, mobile phones, and tablets, supporting real-time display and remote adjustment of steam-specific parameters such as dryness and enthalpy.

[0054] 1) Encrypted transmission: Core data is encrypted using the AES-128 algorithm, and the key is exchanged using the RSA algorithm to prevent data tampering; 2) Data anonymization: Sensitive parameters such as current and fault location are only displayed in range or area numbers on non-control terminals, without showing specific values.

[0055] The production load prediction model uses a gradient boosting regression tree (GBRT) architecture: 1) Model structure: Number of decision trees = 100, tree depth = 5, learning rate = 0.05, minimum number of sample splits = 20; 2) Input parameters: historical steam temperature, pressure, dryness, enthalpy, production shift (0-1 code, 0 for day shift, 1 for night shift), and seasonal coefficient (spring 0.9, summer 1.1, autumn 1.0, winter 1.2) of the acquisition module. 3) Output parameters: Production load forecast for the next 24 hours (Q, in t / h, corresponding to steam demand); 4) Training method: The historical operating data of one year is used as the training set. The training set and the validation set are divided in a 7:3 ratio. The loss function is the mean squared error (MSE). The training is terminated when the MSE of the validation set is ≤0.02 or the number of iterations reaches 100. The relative error of the prediction results is ≤±5%, which provides a basis for load prediction for the collaborative optimization module.

[0056] The fault-tolerant control module is equipped with a primary and backup dual-core control unit. It synchronizes the parameters of the collaborative adjustment model of the adjustment module, the operation and maintenance monitoring data and adjustment instructions of the operation and maintenance early warning module in real time. It adopts a fault prediction trigger switching mechanism based on actuator current fluctuation and consistency verification of redundant sensor data of the acquisition module. When the primary control unit shows abnormal signs, it completes the hot switch in advance and synchronizes the control authority to the cloud backup unit of the management and control module to ensure that the adjustment process is not interrupted.

[0057] The specific criteria for fault prediction are as follows: 1) Redundant sensor data consistency verification: Deviation of the same parameter collected by dual redundant sensors ;in This represents the absolute deviation of the data acquired by the dual-channel sensors. The unit is consistent with the corresponding parameter, such as MPa for pressure deviation and °C for temperature deviation. This is the value collected by the first sensor, and the unit is consistent with the corresponding parameter. This is the value collected by the second sensor; the unit is consistent with the corresponding parameter. when Furthermore, if the sensor malfunctions for three consecutive sampling periods, it is determined to be faulty; among which... The threshold for sensor data consistency verification is set according to parameter type: pressure sensor MPa, temperature sensor °C and dryness detection unit =0.01; 2) Actuator current fluctuation determination: Actuator rated current When the real-time current satisfy 0.2 If the condition persists for 50ms, it is determined that the executor is malfunctioning; 3) Triggering logic: When any of the above-mentioned anomaly detection conditions are met, the fault-tolerant control module initiates a hot switchover of the primary and backup control units. The switching time is... 20ms, to ensure that the adjustment command is not interrupted.

[0058] It is equipped with a dual power supply system consisting of a main power supply and a backup lithium battery, featuring voltage monitoring and automatic switching functions, and an integrated voltage fluctuation compensation unit to adapt to grid fluctuations caused by the start-up and shutdown of the steam system. It also features a mechanical emergency adjustment mechanism adapted to high-pressure steam, employing a worm gear drive design. In the event of a complete failure of the electronic system, the valve opening can be manually adjusted and locked, simultaneously triggering a cloud-based safety warning from the control module. The cloud platform provides real-time backup of critical system operating data and adjustment strategies, supporting rapid synchronization of parameters with each core module after fault recovery.

[0059] The collaborative optimization module integrates comprehensive parameters such as steam temperature, pressure, dryness and enthalpy collected by the acquisition module, drives the energy consumption data of the module and the component loss data of the operation and maintenance early warning module, and establishes a database linking energy consumption and operation and maintenance costs.

[0060] The core of the database stores three types of data: 1) Basic steam parameters, including temperature, pressure, dryness fraction, and enthalpy; 2) Drive energy consumption data, including power, runtime, and recovered energy; 3) Operation and maintenance cost data, including component name, replacement unit price, and downtime. These three types of data are linked by timestamp and support algorithm query and retrieval.

[0061] A multi-objective genetic algorithm is used to calculate the balance point of energy consumption and operation and maintenance costs under different production loads and dryness ranges based on the energy consumption data of the drive module and the component loss data of the operation and maintenance early warning module. Combined with the production load prediction, the coordinated adjustment parameters of the adjustment module are optimized in advance to achieve the matching of steam supply and production demand and reduce ineffective energy consumption. For example, when the dryness is >0.97, the valve opening is optimized to reduce the waste of superheated steam.

[0062] The specific configuration of the multi-objective genetic algorithm is as follows: 1) Objective function: ① Minimize energy consumption in Total energy consumption of the drive module, in kWh; The real-time power of the drive module at time i is expressed in kW. Let be the runtime at time i, in hours (h). This represents the number of time segments within a statistical period. ② Minimize operation and maintenance costs ;in Total maintenance cost, in yuan; The unit price for replacing a single critical component is expressed in yuan per component. This refers to the number of times components were replaced within a given time period. The downtime loss is expressed as yuan / hour. Total downtime during the statistical period, in hours; 2) Constraints: Steam dryness ∈[0.9, 1.0], Steam dryness; temperature , The actual temperature of the steam is expressed in °C; the pressure... , The actual steam pressure is expressed in MPa; valve opening degree. , Valve opening percentage, in % 3) Algorithm parameters: Population size is set to 50, number of iterations is set to 100, crossover probability = 0.7, mutation probability = 0.05. A non-dominated sorting genetic algorithm is used to find the Pareto optimal solution. The solution closest to the ideal solution is selected as the optimal adjustment parameter. The ideal solution is... =0、 =0, meaning theoretically there is no energy consumption and no maintenance cost;

[0063] The mean squared error involved is uniformly labeled as MSE, and it is calculated by dividing the sum of the squares of the differences between the predicted and actual values ​​by the amount of data.

[0064] Generate linked reports on energy consumption, operation and maintenance costs, and dryness, and output optimization suggestions that take into account both energy saving and low-cost operation and maintenance, such as adjusting steam pressure to increase dryness and reducing the aging rate of seals, and push them to the management and control module.

[0065] This invention also provides a method for regulating an electrically operated regulating valve in a steam pipeline, based on the above-described system, comprising the following steps: Data Acquisition Phase: By integrating pressure, temperature, phase change state, pipeline stress sensors, and a steam dryness detection unit based on a throttling and infrared composite method, the temperature, pressure, phase change state, pipeline stress, and dryness parameters in the steam pipeline are acquired simultaneously. The enthalpy value is calculated through the dryness detection unit. The acquired and calculated data are preprocessed using a Kalman filter and wavelet threshold fusion noise reduction algorithm to output a standardized dataset. Adjustment phase: Based on the dual-resistance model, a five-parameter collaborative dynamic model is constructed. Standardized data from the acquisition phase is received. A three-level adaptive adjustment logic of benchmark adjustment, working condition change correction and phase change compensation is used to generate adjustment commands to drive the electric actuator to complete the valve opening adjustment. At the same time, the opening and current data fed back by the actuator are received, and the model parameters are dynamically corrected to form a closed-loop adjustment. Operation and maintenance early warning stage: Integrate steam condition data from the acquisition stage with actuator operation data from the regulation stage, extract fault features through convolutional neural networks, diagnose steam faults such as valve core erosion and seal aging, predict the remaining life of key components based on the improved LSTM model, and generate differentiated operation and maintenance solutions. During the control phase: the edge computing unit processes the sensing data in real time and assists in generating emergency adjustment instructions; the cloud platform realizes centralized monitoring of multi-pipeline operating conditions, remote optimization of adjustment parameters, and review and issuance of operation and maintenance plans; and achieves end-to-cloud data interaction through encrypted communication protocols. Redundancy and fault tolerance protection phase: Real-time monitoring of the operating status of each core module (acquisition module, adjustment module, operation and maintenance early warning module, drive module, control module, and collaborative optimization module). When abnormal signs are detected, the main and backup control units are triggered to perform millisecond-level hot switching and synchronize permissions to the cloud backup. When the main power supply is abnormal, it automatically switches to the backup lithium battery. When the electronic system fails, the mechanical emergency mechanism is activated for manual adjustment and safety warnings are triggered. Collaborative optimization phase: Integrate steam temperature, pressure, dryness and enthalpy parameters collected by the acquisition module, energy consumption data of the drive module and component loss data of the operation and maintenance early warning module, calculate the balance point between energy consumption and operation and maintenance costs through a multi-objective genetic algorithm, optimize the adjustment parameters and push them to the adjustment module and the cloud, so as to achieve precise matching between steam supply and production demand.

[0066] During the data acquisition phase, the dryness detection accuracy is ≤ ±0.5%, and the data sampling frequency is 100Hz; during the adjustment phase, the steady-state adjustment error is ≤ ±0.1℃ and ±0.01MPa, and the adjustment step size is compressed to 0.1% when the operating conditions change abruptly.

[0067] During the redundancy and fault tolerance protection phase, the mechanical emergency adjustment accuracy is ≤±1%, and the system operation data and adjustment strategy are backed up in real time in the cloud. After the fault is recovered, the parameters can be quickly synchronized.

[0068] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0069] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A regulating system for an electrically operated regulating valve in a steam pipeline, characterized in that, It includes a data acquisition module, an adjustment module, an operation and maintenance early warning module, a driver module, a management and control module, a fault-tolerant control module, and a collaborative optimization module; Acquisition module: Integrates a data processing mechanism that combines multiple types of sensors and multiple algorithms to capture the non-equilibrium process of steam phase change and key parameters, completes data noise reduction, and outputs standardized data; Regulation module: Based on the standardized data from the acquisition module, a steam-specific adaptive regulation model is established. Based on multi-parameter correlation analysis, a three-level regulation logic is adopted to realize regulation under non-equilibrium phase change scenarios, and regulation commands are generated and sent to the drive module. Operation and maintenance early warning module: Establish a full life cycle operation and maintenance system, integrate steam operating condition data and actuator operating status data, realize fault diagnosis and component life prediction, and generate differentiated operation and maintenance solutions; Drive module: Receives adjustment commands to drive the valve to complete the action, and synchronously feeds back operating status data and energy consumption data; Management and control module: Establish an end-to-cloud collaborative management and control architecture, integrate the operation data of each module to achieve local real-time control and cloud-based global optimization, and support visual monitoring and remote operation; Fault-tolerant control module: Establish a multi-dimensional redundancy protection system, monitor the status of core modules in real time, and set up a main and backup power switching mechanism, an emergency adjustment mechanism, and a real-time backup mechanism; Collaborative optimization module: Establishes a dual-objective optimization model, analyzes the balance point between energy consumption and operation and maintenance costs, and generates optimization suggestions to guide the adjustment module in reverse. The acquisition module integrates a dual-redundant design of pressure sensor, temperature sensor, phase change state sensor, pipeline stress sensor, and steam dryness detection unit based on a throttling and infrared composite method; each redundant sensor synchronously acquires the same parameter to achieve data consistency verification, providing a basis for fault prediction for the fault-tolerant control module; The steam dryness detection unit depressurizes the steam to a superheated state through a throttling orifice, calculates the enthalpy value by combining the temperature and pressure data after depressurization, corrects the detection error in the wet steam region, and realizes dryness detection within a preset range; the phase change capture adopts a three-pronged technology of microwave resonance, infrared thermometry and dryness detection to identify the gas-liquid double film resistance distribution at the steam condensation phase change interface, and captures the phase change critical point, phase change rate and non-equilibrium interface temperature characteristics. Data processing employs a fusion noise reduction algorithm to preprocess data collected synchronously from multiple sensors in a time sequence, remove interference signals, output a standardized dataset, and push it to the adjustment module and the operation and maintenance early warning module, while storing historical data. The adjustment module receives standardized data output by the acquisition module, establishes a five-parameter collaborative dynamic model of temperature, pressure, dryness, enthalpy and phase change rate, quantifies the thermal resistance relationship of gas-liquid film at the phase change interface based on the dual resistance model, and characterizes the correspondence between comprehensive parameters and valve adjustment amount under different operating conditions. A three-level adaptive adjustment algorithm is adopted, consisting of benchmark adjustment, operating condition change correction, and phase change compensation. The operating condition change identification unit detects special scenarios by measuring dryness and pressure change rate and prioritizes the activation of a fast response strategy. Under non-change conditions, benchmark adjustment is performed, while the phase change compensation submodule corrects the valve adjustment rate in real time based on the gas-liquid film resistance data output by the collaborative dynamic model. A closed-loop feedback mechanism is established, which combines the valve opening and operating current data fed back by the drive module with the dryness and enthalpy values ​​of the acquisition module to dynamically correct the model parameters and ensure adjustment accuracy. The operation and maintenance early warning module receives steam condition data from the acquisition module and operating status data from the drive module to establish a full life cycle intelligent operation and maintenance system; it integrates a valve core vibration sensor, a sealing cavity pressure sensor, an actuator current sensor, a high-frequency acoustic valve core erosion monitoring unit, and a sealing cavity temperature field distribution acquisition unit to capture valve core erosion characteristic signals, high-temperature aging data of seals, and actuator operating status parameters. A feature database for steam-specific faults is constructed. Fault characteristics are extracted from multi-dimensional monitoring signals, and fault type and location are determined by rule-based reasoning. Establish a correlation model between operating parameters and failure rate, input historical operating data and real-time failure signals to predict the remaining service life of key components, generate differentiated operation and maintenance plans, and output operation and maintenance priorities and operation guidelines.

2. The regulating system for the electric regulating valve in the steam pipeline according to claim 1, characterized in that, The drive module receives the opening adjustment command from the adjustment module and adopts an integrated drive architecture combined with control technology to achieve rapid valve opening response; the actuator housing adopts a heat-insulating and sealed design and integrates a dehumidification unit. The integrated regenerative energy recovery unit reduces energy consumption caused by ineffective regulation; the actuator winding temperature monitoring unit is configured to collect real-time thermal operation data of the drive mechanism and feed it back to the regulation module, operation and maintenance early warning module and collaborative optimization module.

3. The regulating system for the electric regulating valve in the steam pipeline according to claim 1, characterized in that, The edge computing unit of the control module deploys a collaborative adjustment model and a fault early warning algorithm to process the sensing data from the acquisition module in real time, receive the adjustment parameters from the adjustment module, and assist in generating emergency adjustment commands. The cloud platform enables centralized monitoring of the collaborative operating conditions of multi-pipeline systems, remote optimization of adjustment parameters, review and distribution of operation and maintenance plans, and historical data traceability. At the same time, it establishes a production load prediction model and outputs the prediction results to the collaborative optimization module. It integrates multi-mode communication interfaces to support multi-protocol adaptation, adopts end-to-cloud encrypted transmission and data security technology, and designs a monitoring interface that adapts to multiple terminals, supporting real-time display and remote operation of steam-specific parameters.

4. The regulating system for the electric regulating valve in the steam pipeline according to claim 1, characterized in that, The fault-tolerant control module is equipped with a primary and backup dual core control unit, which synchronizes core data and adjustment commands in real time. It realizes hot switching of primary and backup based on the fault prediction triggering switching mechanism, and at the same time synchronizes the control authority to the management and control module. It is equipped with a dual power supply system consisting of a main power supply and a backup lithium battery, as well as a voltage fluctuation compensation unit to adapt to power grid fluctuations; it also features an added mechanical emergency adjustment mechanism that allows manual adjustment and locking of the valve opening in the event of a complete electronic system failure, while providing cloud-based safety warnings. The cloud platform provides real-time backup of critical system operating data and adjustment strategies.

5. The regulating system for the electric regulating valve in the steam pipeline according to claim 1, characterized in that, The collaborative optimization module integrates the steam parameters collected by the acquisition module, the energy consumption data of the drive module, and the component loss data of the operation and maintenance early warning module to establish a database linking energy consumption and operation and maintenance costs. A multi-objective genetic algorithm is used to calculate the balance point between energy consumption and operation and maintenance costs under different operating conditions. Combined with production load forecasting, the coordinated adjustment parameters are optimized to achieve the matching of steam supply and production demand. A linked report of energy consumption, operation and maintenance costs and dryness is generated, and optimization suggestions are output and pushed to the control module.

6. A method for regulating an electrically operated regulating valve in a steam pipeline, based on the system described in any one of claims 1-5, characterized in that, Includes the following steps: Data Acquisition Phase: By integrating multiple types of sensors and a steam dryness detection unit, key parameters of the steam pipeline are collected synchronously. After calculating the enthalpy value, a fusion noise reduction algorithm is used for preprocessing to output a standardized dataset. Regulation phase: Based on the dual-resistance model, a five-parameter collaborative dynamic model is constructed. Standardized data is received, and a three-level adaptive regulation logic is used to generate regulation commands to drive valve opening adjustment. The model parameters are dynamically corrected in combination with feedback data to form a closed-loop regulation. Operation and maintenance early warning stage: Integrate steam condition data and actuator operation data, extract fault characteristics to diagnose steam faults, predict the remaining life of key components based on intelligent models, and generate differentiated operation and maintenance solutions; During the control phase: the edge computing unit processes the sensing data in real time and assists in generating emergency adjustment instructions, while the cloud platform realizes centralized monitoring, parameter optimization, and operation and maintenance plan review and distribution, and achieves end-to-cloud data interaction through encrypted communication; Redundancy and fault tolerance protection phase: Real-time monitoring of the operating status of core modules; triggering the switching of main and backup control units and synchronizing permissions to cloud backup when an anomaly occurs; automatically switching to backup power when the main power supply is abnormal; activating mechanical emergency mechanisms and triggering safety warnings when electronic system fails. Collaborative optimization phase: The energy consumption and operation and maintenance cost balance point is calculated through a multi-objective genetic algorithm, the adjustment parameters are optimized and pushed to the adjustment module and the cloud, so as to achieve the matching of steam supply and production demand.

Images