A back-pressure power generation control system and power generation device for utilizing waste heat and pressure from steam

Abstract

This invention relates to the field of steam waste heat and pressure recovery technology, and discloses a back-pressure power generation control system and power generation device for utilizing steam waste heat and pressure. The system includes: an inlet regulating valve to control the inlet steam flow; a turbine generator to convert steam into electrical energy and cooled, depressurized process steam; a back-end demand monitoring module to collect real-time steam demand; a control system that constructs a digital twin model through feature extraction and fusion to achieve process steam condition prediction and system optimization control, improving the operating efficiency and stability of the steam waste heat and pressure power generation system, reducing process steam waste; and a decision support module based on knowledge graph technology to provide intelligent decision support for operators. Through the intelligent collaborative work of multiple modules, this invention achieves efficient recovery of steam waste heat and pressure and intelligent, safe operation of the power generation system, resulting in significant economic and social benefits.

Description

Technical Field

[0001] This invention relates to the field of waste heat and pressure recovery technology, and in particular to a back pressure power generation control system and power generation device for utilizing steam waste heat and pressure. Background Technology

[0002] In existing steam waste heat and pressure power generation systems, digital twin applications are mostly limited to static geometric display or single thermodynamic simulation. They cannot simultaneously reflect the unsteady flow of steam in the pipeline network and the expansion-compression coupling process inside the turbine. Traditional models often separate thermodynamic calculations from fluid dynamic calculations and lack a real-time mutual feedback mechanism. As a result, when back pressure demand changes abruptly, the model prediction error is rapidly amplified, making it difficult to support subsequent closed-loop control.

[0003] Existing control methods for waste heat and pressure power generation systems generally employ single-machine PID or simple rule tables, generating valve opening commands based solely on instantaneous measurements. These methods lack global operating condition optimization and online compensation mechanisms. When the system operating point deviates from the design operating condition, parameters must be manually adjusted. This approach cannot cope with the dual time-varying characteristics of steam "load-heat source," resulting in high steam venting rates and low power generation efficiency.

[0004] In existing steam waste heat and pressure power generation systems, turbine-side sensors output high-dimensional data such as pressure and temperature at a frequency of seconds, while downstream steam demand is often reported at the minute level or even in an event-triggered manner. The sampling period, dimensions, and noise levels of the two are vastly different. Existing technologies mostly use fixed-weighted averages or simple sliding filters, which neither solves the problem of heterogeneous timing alignment nor performs online estimation of noise covariance during load abrupt changes, resulting in lagging and unreliable system state estimation.

[0005] In existing steam waste heat and pressure power generation systems, operators have long relied on paper-based SOPs or personal experience for troubleshooting and adjusting operating conditions. They lack tools to retrieve and reason about multi-source information such as DCS real-time data, CMMS historical work orders, and expert experience. As a result, abnormal operating conditions are slow to be identified, maintenance strategies are lagging behind, and the availability and safety of the system are difficult to improve. Summary of the Invention

[0006] This invention provides a back-pressure power generation control system and power generation device for utilizing waste heat and pressure from steam, in order to solve existing technical problems.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: This invention provides a back-pressure power generation control system for utilizing waste heat and pressure from steam, comprising: Inlet regulating valve: Used to control the flow rate of waste heat steam from the waste heat source; Turbine generator: Waste heat steam is fed into a turbine to generate steam and electricity for process use, and the turbine power generation process is monitored in real time to obtain monitoring data of the turbine power generation module; Backend demand monitoring module: Used to monitor the demand of backend steam-consuming equipment in real time and obtain backend demand data; Feature extraction module: preprocesses and extracts features from monitoring data and backend requirement data to obtain system operation features; Data modeling module: Builds a digital twin model based on the system's operational characteristics; Operating condition prediction module: Predicts the operating conditions of the system based on the digital twin model and generates control vectors based on the prediction results; The operation control module generates control commands based on the control vector, thereby controlling the system's operation. Decision Support Module: Used to provide intelligent decision suggestions to operators through knowledge graph technology.

[0008] Furthermore, the inlet regulating valve is used to connect the waste heat source to the turbine generator through a pipeline, and the system controls the flow rate of waste heat steam entering the turbine generator by adjusting the opening of the inlet regulating valve.

[0009] Furthermore, the turbine generator includes a turbine unit, a power generation unit, a bypass valve unit, and a monitoring unit; The turbine unit is used to convert the thermal energy of pretreated steam into mechanical energy through a turbine, output process steam and introduce the process steam into the process header, and the process header is used to transmit the process steam to the downstream steam-using equipment. The power generation unit is used to connect the output shaft of the turbine to the input shaft of the generator through a coupling, convert the mechanical energy of the turbine into electrical energy, and transmit the electrical energy to the downstream electrical equipment. The bypass valve unit is used to immediately open the bypass valve when the turbine trips due to a fault or in an emergency shutdown. The steam inside the turbine is de-cooled and de-pressurized by the desuperheater and pressure reducer and then introduced into the process header. The monitoring unit is used to monitor the inlet steam flow of the turbine in real time through a flow sensor, the inlet steam pressure of the turbine and the steam pressure of the process header in real time through a pressure sensor, the inlet steam temperature of the turbine in real time through a temperature sensor, and the power generation of the generator in real time through an electrical parameter sensor. The monitoring unit collects the above monitoring data through a PLC controller and transmits the monitoring data to the feature extraction module.

[0010] Furthermore, the backend demand monitoring module includes a demand definition unit, a data acquisition unit, and a data transmission unit; The demand definition unit is used to divide the demand for steam from the back-end equipment to be sampled into multiple demand parameters, and define corresponding acquisition rules according to the demand parameters. The acquisition rules include the acquisition equipment corresponding to the demand parameters and its deployment location, signal interface and acquisition cycle. The data acquisition unit is used to collect the demand of back-end steam-consuming equipment through the acquisition device according to the acquisition rules, and obtain back-end demand data. The data transmission unit is used to integrate the backend demand data through the edge gateway and transmit the backend demand data to the feature extraction module via the MQTT protocol.

[0011] Furthermore, the feature extraction module includes a data preprocessing unit, a monitoring feature extraction unit, a demand feature extraction unit, and a data fusion unit; The data preprocessing unit is used to clean, format and normalize the monitoring data and the backend demand data respectively to obtain preprocessed monitoring data and preprocessed backend demand data. The monitoring feature extraction unit is used to perform data dimensionality reduction and principal component extraction on the preprocessed monitoring data using the principal component analysis (PCA) algorithm to obtain the turbine generator operating characteristics. The demand feature extraction unit is used to extract features from the preprocessed backend demand data using the TCN-feature extraction method to obtain backend steam load features; The data fusion unit is used to synchronize and fuse the monitoring features and backend demand features in time using an adaptive Kalman fusion method to obtain system operation features.

[0012] Furthermore, the data modeling module includes a model building unit, a model optimization unit, and a model visualization unit; The model building unit is used to build a geometric 3D model of the system in the MATLAB modeling environment based on the system structure, and to build a thermodynamic-fluid dynamic coupling model of the system based on the system operating characteristics. The model optimization unit is used to fuse the geometric 3D model and the thermodynamic-fluid dynamic coupling model of the system using the multiphysics coupling tool in the MATLAB environment to obtain the digital twin model of the system, and to verify and optimize the digital twin model of the system by the back-substitution-residual analysis method. The model visualization unit is used to build a visualization interface through the App Designer in the MATLAB environment to visualize and display the digital twin model and its operation.

[0013] Furthermore, the operating condition prediction module includes a model loading unit, a prediction execution unit, and a control requirement parsing unit; The model loading unit is used to obtain a digital twin model from the data modeling module, set the model parameters of the digital twin model, and generate a digital twin model instance for working condition simulation. The model parameters include the initial conditions and boundary conditions of the digital twin model, the time step and total duration of the dynamic simulation. The prediction execution unit is used to perform dynamic simulation of the digital twin model instance according to the model parameters, and predict the operating status of the system through the PIML fusion algorithm to obtain prediction information. The prediction information includes the system's performance, load and fault risk under the current and future operating conditions. The performance includes power generation efficiency and heat recovery efficiency. The load includes operating load, maximum load and load change rate. The fault risk includes fault type, fault probability, fault time and fault impact range. The control requirement analysis unit is used to further optimize the dynamic simulation results using a genetic algorithm to obtain the optimal operating condition trajectory, and then convert the optimal operating condition trajectory into a control vector using a rolling time-domain linearization method. The control vector includes the controlled object, control parameters, and control signal strength.

[0014] Furthermore, the operation control module includes an adaptive optimization unit, an instruction decomposition unit, and an instruction execution unit; The adaptive optimization unit is used to perform online compensation of the control vector based on the operating characteristics using an adaptive optimization algorithm; The instruction decomposition unit is used to map the control vector into system control instructions through a static mapping table. The control instructions include inlet regulating valve control instructions and bypass valve control instructions. The instruction execution unit is used to transmit the control instructions via an industrial fieldbus, thereby controlling the operation of the system.

[0015] Furthermore, the decision support module includes a knowledge extraction unit, a graph construction unit, and a decision generation unit; The knowledge extraction unit is used to obtain multi-source heterogeneous information in the field of waste heat power generation from DCS real-time database, CMMS maintenance work orders, standard operating procedures (SOPs) and expert interview records, and generate structured knowledge triples through entity relation extraction method. The graph construction unit is used to construct a knowledge graph in the field of waste heat power generation based on the structured knowledge triples and the Neo4j graph database. The decision generation unit is used to generate operational decision suggestions based on system operating characteristics and knowledge graph, and push the decision suggestions to operators through a web-based visual dashboard. The decision suggestions include equipment maintenance suggestions and fault prevention suggestions.

[0016] This invention provides a back-pressure power generation device for utilizing waste heat and pressure from steam, comprising: Inlet regulating valve: Used to regulate the flow rate of waste heat steam from the heat source; Turbine: Used to turbine waste heat steam, converting the thermal energy of the waste heat steam into mechanical energy, and then transmitting the turbined process steam to downstream steam-using equipment through the process header. Generator: Connected to the turbine via a coupling, it is used to convert the mechanical energy of the turbine into electrical energy and transmit it to downstream electrical equipment; Back-end steam supply equipment: used to receive and use process steam output from the turbine; Back-end electrical equipment: used to receive and use the electrical energy output by the generator; Backend demand monitoring module: used to monitor the steam demand of backend steam-consuming equipment in real time and obtain backend demand data; Intelligent monitoring system: used to monitor the operation of turbines and generators to obtain monitoring data, receive backend demand data from the backend demand monitoring module, preprocess and extract features from the monitoring data and backend demand data to obtain system operation characteristics, perform digital twin modeling and operating condition prediction of the system based on the system operation characteristics, and perform operation control of the system based on the results of the operating condition prediction.

[0017] The beneficial effects of the technical solution provided by this invention include at least the following: This invention constructs a thermodynamic-hydrodynamic bidirectional coupled digital twin model of a steam waste heat and pressure power generation system in the MATLAB environment, and corrects the model parameters online through back-substitution-residual analysis, effectively reducing the error in the model's predicted power generation when the heat source pressure or downstream demand fluctuates significantly, laying a reliable foundation for subsequent precise control.

[0018] This invention designs a three-level progressive optimization control architecture for a steam waste heat and pressure power generation system. The first level uses a genetic algorithm to globally search for the optimal operating trajectory in a digital twin. The second level uses rolling time-domain linearization to convert the trajectory into a valve opening control vector in real time. The third level uses an adaptive optimization unit to perform millisecond-level compensation on the vector based on the latest measurement values, effectively improving the system's power generation efficiency and reducing the steam venting rate.

[0019] This invention introduces TCN-feature extraction to capture long-sequence periodic patterns of back-end steam demand in a waste heat and pressure power generation system. Then, an adaptive Kalman fusion algorithm is used to synchronize the monitoring data and the back-end steam demand data in time and estimate the noise covariance online. This effectively reduces the system state estimation delay when the back-end load suddenly increases, and significantly improves the real-time performance and robustness of the system's automatic control.

[0020] This invention constructs a knowledge graph in the field of waste heat power generation by extracting entity relationships, providing operators with comprehensive and accurate decision-making suggestions, reducing false alarm rates, improving the scientific nature and timeliness of decision-making, and thus enhancing system safety, reliability and operation and maintenance efficiency. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a system flowchart provided in an embodiment of the present invention; Figure 2 This is a structural diagram of the device provided in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0024] This embodiment provides a back-pressure power generation control system and power generation device for utilizing waste heat and pressure from steam. Please refer to... Figure 1 This is a system flowchart provided in an embodiment of the present invention.

[0025] A back-pressure power generation control system for utilizing waste steam heat and pressure in this embodiment includes: 1. Inlet regulating valve: Used to control the flow rate of waste heat steam from the waste heat source; The inlet regulating valve is used to connect the waste heat source to the turbine generator through a pipeline. The system controls the flow rate of waste heat steam entering the turbine generator by adjusting the opening of the inlet regulating valve.

[0026] It should be noted that the opening of the regulating valve is usually driven by an actuator (such as an electric actuator). Opening the regulating valve more will increase the steam intake and power generation of the turbine generator, while closing the regulating valve will reduce the steam intake and power generation of the turbine generator. In conjunction with subsequent monitoring data and control commands, the back pressure of the turbine generator is kept stable, ensuring that "steam production = steam consumption", and minimizing the waste caused by steam venting or pressure reducing valve release.

[0027] II. Turbine Generator: Waste heat steam is fed into the turbine to generate steam and electricity for process use. The turbine power generation process is monitored in real time to obtain monitoring data of the turbine power generation module. The turbine generator includes a turbine unit, a power generation unit, a bypass valve unit, and a monitoring unit; The turbine unit is used to convert the thermal energy of pretreated steam into mechanical energy through a turbine, output process steam and introduce the process steam into the process header, and the process header is used to transmit the process steam to the downstream steam-using equipment. The power generation unit is used to connect the output shaft of the turbine to the input shaft of the generator through a coupling, convert the mechanical energy of the turbine into electrical energy, and transmit the electrical energy to the downstream electrical equipment. The bypass valve unit is used to immediately open the bypass valve when the turbine trips due to a fault or in an emergency shutdown. The steam inside the turbine is de-cooled and de-pressurized by the desuperheater and pressure reducer and then introduced into the process header. The monitoring unit is used to monitor the inlet steam flow of the turbine in real time through a flow sensor, the inlet steam pressure of the turbine and the steam pressure of the process header in real time through a pressure sensor, the inlet steam temperature of the turbine in real time through a temperature sensor, and the power generation of the generator in real time through an electrical parameter sensor. The monitoring unit collects the above monitoring data through a PLC controller and transmits the monitoring data to the feature extraction module.

[0028] It should be noted that a flow sensor is a sensor that converts flow signals into electrical signals to measure the flow rate of fluids. Common flow sensors include turbine flow meters, electromagnetic flow meters, and ultrasonic flow meters. Taking electromagnetic flow meters as an example, they are based on Faraday's law of electromagnetic induction. When a conductive liquid flows in a magnetic field, an induced electromotive force is generated in the conductor. The flow rate can be obtained by measuring the induced electromotive force.

[0029] A pressure sensor is a sensor that converts pressure signals into electrical signals, used to measure the pressure of gases, liquids, or solids. Common types of pressure sensors include strain gauge type, piezoresistive type, and capacitive type. Taking the piezoresistive type as an example, it is based on the piezoresistive effect of semiconductor materials. When pressure is applied to the sensor, it causes a change in the resistance of the semiconductor material. The pressure value can be obtained by measuring the change in resistance.

[0030] A temperature sensor is a sensor that converts temperature signals into electrical signals to measure the temperature of an object. Common types of temperature sensors include thermocouples, resistance temperature detectors (RTDs), and thermistors. Taking the RTD as an example, it is based on the characteristic that the resistance of a metallic conductor changes with temperature. When the temperature changes, the resistance changes accordingly, and the temperature value can be obtained by measuring the resistance value.

[0031] A humidity sensor is a sensor that converts humidity signals into electrical signals to measure the humidity of air or other gases. Common types of humidity sensors include capacitive, resistive, and ceramic sensors. Taking capacitive sensors as an example, they are based on the characteristic that capacitance changes with humidity. When humidity changes, the capacitance changes accordingly, and the humidity value can be obtained by measuring the capacitance value.

[0032] An electrical parameter sensor is a sensor that converts electrical parameter signals into electrical signals, used to measure electrical parameters such as voltage, current, and power. Common electrical parameter sensors include voltage transformers, current transformers, and power sensors. Taking a current transformer as an example, it is based on the principle of electromagnetic induction, converting primary current into secondary current to measure current.

[0033] III. Backend Demand Monitoring Module: Used to monitor the demand of backend steam-consuming equipment in real time and obtain backend demand data; The backend demand monitoring module includes a demand definition unit, a data acquisition unit, and a data transmission unit; The demand definition unit is used to divide the demand for steam from the back-end equipment to be sampled into multiple demand parameters, and define corresponding acquisition rules according to the demand parameters. The acquisition rules include the acquisition equipment corresponding to the demand parameters and its deployment location, signal interface and acquisition cycle. The data acquisition unit is used to collect the demand of back-end steam-consuming equipment through the acquisition device according to the acquisition rules, and obtain back-end demand data. The data transmission unit is used to integrate the backend demand data through the edge gateway and transmit the backend demand data to the feature extraction module via the MQTT protocol.

[0034] It should be noted that the following is an example of a data collection rule defined by the requirement definition unit:

[0035] MQTT (Message Queuing Telemetry Transport) is a standard protocol for lightweight, low-bandwidth communication between edge computing and IoT devices. It is suitable for real-time data transmission in industrial internet scenarios. MQTT supports edge nodes aggregating data and pushing it to the cloud or central system (such as a feature extraction module). In this patent, the system needs to monitor backend requirements in real time, and MQTT's publish / subscribe model can efficiently handle such requirements.

[0036] IV. Feature Extraction Module: Preprocesses and extracts features from monitoring data and backend demand data to obtain system operation features; The feature extraction module includes a data preprocessing unit, a monitoring feature extraction unit, a demand feature extraction unit, and a data fusion unit. The data preprocessing unit is used to clean, format and normalize the monitoring data and the backend demand data respectively to obtain preprocessed monitoring data and preprocessed backend demand data. The monitoring feature extraction unit is used to perform data dimensionality reduction and principal component extraction on the preprocessed monitoring data using the principal component analysis (PCA) algorithm to obtain the turbine generator operating characteristics. The demand feature extraction unit is used to extract features from the preprocessed backend demand data using the TCN-feature extraction method to obtain backend steam load features; The data fusion unit is used to synchronize and fuse the monitoring features and backend demand features in time using an adaptive Kalman fusion method to obtain system operation features.

[0037] It should be noted that data cleaning refers to processing the collected raw data to ensure its quality and usability. In steam waste heat and pressure power generation systems, the data collected by sensors may be affected by noise interference, equipment failure, or human factors. Data cleaning is necessary to remove erroneous, duplicate, missing, or inconsistent data, thereby improving the accuracy and reliability of the data.

[0038] Data formatting refers to converting collected data into a unified format. Since data collected by different types of sensors may have different formats (timestamp format, numerical precision, and data type), data formatting can unify these data into a standard format, which facilitates data storage, processing, and analysis.

[0039] Data normalization refers to scaling data proportionally to make it fall within a small range (usually [0,1]). In a waste heat and pressure power generation system, data collected by different sensors may have different dimensions and numerical ranges. Data normalization is needed to eliminate these differences, make data with different characteristics comparable, and further improve the convergence speed and performance of subsequent algorithms.

[0040] Principal Component Analysis (PCA) is a statistical method that transforms original data into a new coordinate system through linear transformation, maximizing the variance of the data along the coordinate axes in this new system, thereby achieving data dimensionality reduction and feature extraction. In steam waste heat and pressure power generation systems, the collected operational data may contain a large number of correlated features, requiring PCA algorithms for data dimensionality reduction to reduce data redundancy and extract the most important features (i.e., principal components), thus improving the efficiency and performance of subsequent processing.

[0041] Temporal Convolutional Network-based Feature Extraction (TCN) outperforms traditional RNN / LSTM methods in handling long sequences and capturing multi-scale temporal dependencies, and has high parallelism, especially meeting the real-time requirements of industrial applications. TCN can automatically extract local-global temporal features without the need for manual construction of statistics, meeting the requirements of "end-to-end" feature extraction. Its functionality has been extensively validated in scenarios such as energy load forecasting and industrial demand-side management.

[0042] The adaptive Kalman fusion method naturally supports time synchronization, noise suppression, and weight adaptation. It is suitable for handling heterogeneous sampling frequencies and dimensions of turbine generator operating characteristics (from PCA) and back-end demand characteristics (from TCN). Its adaptive form (AKF) can estimate the covariance of process noise and observation noise online, ensuring that the fusion result remains robust when operating conditions change abruptly. This method has mature implementations in industry (such as MATLAB Adaptive Kalman Filter Toolbox and C++ real-time library) and can be directly implemented in PLCs or edge gateways.

[0043] V. Data Modeling Module: Establish a digital twin model based on the system's operational characteristics; The data modeling module includes a model building unit, a model optimization unit, and a model visualization unit; The model building unit is used to build a geometric 3D model of the system in the MATLAB modeling environment based on the system structure, and to build a thermodynamic-fluid dynamic coupling model of the system based on the system's operating characteristics. The model optimization unit is used to fuse the geometric 3D model and the thermodynamic-fluid dynamic coupling model of the system using the multiphysics coupling tool in the MATLAB environment to obtain the digital twin model of the system, and to verify and optimize the digital twin model of the system by back substitution-residual analysis. The model visualization unit is used to build a visualization interface through the App Designer in the MATLAB environment to visualize and display the digital twin model and its operation.

[0044] It should be noted that MATLAB is a high-performance language tool for numerical computation, visualization, and programming, widely used in engineering, science, and mathematics. It supports system modeling, simulation, and control design, and provides a variety of modeling tools (such as Simulink, System Identification Toolbox, etc.) for building complex dynamic system models.

[0045] Geometric 3D models are typically used to represent the geometry, appearance, and behavior of objects. In the digital twin model of a waste heat and pressure power generation system, a 3D visualization model can intuitively show the structure and layout of objects such as turbines and generators.

[0046] The thermodynamic model is a mathematical model built on thermodynamic principles. In this patent, the thermodynamic model is used to describe the energy conversion process in a turbine system.

[0047] The fluid dynamics model is a mathematical model built on the principles of fluid dynamics. In this patent, the fluid dynamics model is used to describe the flow and expansion process of steam in a turbine system.

[0048] Back-substitution-residual analysis is a model validation method mainly used to evaluate the accuracy and reliability of a model. It compares the model's prediction results with actual observation data, calculates the residuals (i.e., the difference between the predicted and actual values), and verifies the model's rationality based on the distribution and magnitude of the residuals.

[0049] App Designer is an interactive application design environment provided by MATLAB for quickly creating graphical user interfaces (GUIs) and applications. It supports drag-and-drop design, allowing users to build feature-rich visual interfaces with simple operations.

[0050] VI. Operating Condition Prediction Module: Predicts the operating conditions of the system based on the digital twin model and generates control vectors based on the prediction results; The operating condition prediction module includes a model loading unit, a prediction execution unit, and a control requirement analysis unit; The model loading unit is used to obtain the digital twin model from the data modeling module, set the model parameters of the digital twin model, and generate a digital twin model instance for working condition simulation. The model parameters include the initial conditions and boundary conditions of the digital twin model, the time step and total duration of the dynamic simulation. The prediction execution unit is used to perform dynamic simulation of the digital twin model instance based on the model parameters, and predict the operating status of the system through the PIML fusion algorithm to obtain prediction information. The prediction information includes the performance, load and failure risk of the back pressure power generation control system for steam waste heat and waste pressure utilization under current and future operating conditions. The performance includes power generation, power generation efficiency and heat recovery efficiency. The load includes operating load, maximum load and load change rate. The failure risk includes failure type, failure probability, failure time and failure impact range. The control demand analysis unit is used to further optimize the dynamic simulation results based on the dynamic simulation results using a genetic algorithm to obtain the optimal operating trajectory, and then transform the optimal operating trajectory into a control vector using a rolling time-domain linearization method. The control vector includes the controlled object, the target operating parameters, and the control signal strength.

[0051] It's important to note that PIML (Physics-Informed Machine Learning) is a method that combines physical laws with machine learning techniques. Its core lies in integrating known physical laws and constraints into machine learning algorithms, enabling the model to learn from data while adhering to physical principles. PIML reduces reliance on large-scale datasets and is particularly suitable for handling complex systems with multiple scales and physical phenomena, such as those in fluid dynamics and thermodynamics. This approach not only improves the model's prediction accuracy but also enhances its interpretability, making it better applicable to complex physical systems.

[0052] Genetic algorithms are a search and optimization technique based on the principles of natural selection and genetics. They simulate the process of biological evolution and gradually improve the quality of candidate solutions through operations such as selection, crossover (recombination), and mutation, ultimately finding the optimal solution or a near-optimal solution. In this patent, genetic algorithms are used to optimize dynamic simulation results (which contain multiple feasible working condition trajectories) to find the optimal working condition trajectory.

[0053] Rolling time-domain linearization is a control method for handling nonlinear systems. It simplifies the control problem by linearizing the nonlinear system near the current operating point at each time step, making it particularly suitable for dynamic systems with complex nonlinear behavior. In the runtime control module, rolling time-domain linearization is used to transform the optimal operating trajectory into a control vector, that is, to convert the optimal operating trajectory into intermediate variables that can be used to execute control commands.

[0054] VII. Operation Control Module: Generates control commands based on the control vector, thereby controlling the operation of the system; The operation control module includes an adaptive optimization unit, an instruction decomposition unit, and an instruction execution unit; The adaptive optimization unit is used to perform online compensation of the control vector based on the operating characteristics using an adaptive optimization algorithm; The instruction decomposition unit is used to map the control vector into system control instructions through a static mapping table. The control instructions include inlet regulating valve control instructions and bypass valve control instructions. The instruction execution unit is used to transmit the control instructions via an industrial fieldbus, thereby controlling the operation of the system.

[0055] It should be noted that adaptive optimization algorithms are algorithms that can dynamically adjust optimization strategies based on the system's operating state. In the operation control module, they are used to perform online compensation of the control vector, dynamically adjusting the control vector through real-time monitoring of system operating characteristics to ensure that the system can adapt to various changes and maintain its optimal operating state during actual operation.

[0056] A static mapping table is a tool that maps high-level control instructions to specific device control instructions. It is usually a predefined table used to convert abstract control vectors into specific control instructions.

[0057] 8. Decision Support Module: Used to provide intelligent decision suggestions to operators through knowledge graph technology; The decision support module includes a knowledge extraction unit, a graph construction unit, and a decision generation unit; The knowledge extraction unit is used to obtain multi-source heterogeneous information in the field of waste heat power generation from DCS real-time database, CMMS maintenance work orders, standard operating procedures (SOPs), and expert interview records, and to generate structured knowledge triples through entity relation extraction methods. The graph construction unit is used to construct a knowledge graph in the field of waste heat power generation based on structured knowledge triples through the Neo4j graph database; The decision generation unit generates operational decision suggestions based on system operating characteristics and knowledge graphs, and pushes these suggestions to operators through a web-based visual dashboard. The decision suggestions include equipment maintenance suggestions and fault prevention suggestions.

[0058] It should be noted that the DCS (Distributed Control System) real-time database is a database system used in industrial automation control systems to store and manage real-time data. CMMS (Computerized Maintenance Management System) work orders are documents that record equipment maintenance and repair activities; SOP (Standard Operating Procedure) is a set of operational steps and requirements established by an enterprise or organization to ensure the standardization and normalization of work processes; Expert interview transcripts are documents that record the experience and insights of experts in the field through interviews.

[0059] Entity relation extraction is a natural language processing technique used to identify entities and the relationships between them from text. Its working principle is as follows: Entity recognition: Identifying entities with specific meanings from text using Named Entity Recognition (NER) technology; Relationship identification: Identifying relationships between entities by analyzing their syntactic structure and semantic information; Knowledge representation: Represent the identified entities and relations as structured knowledge triples (such as "entity 1-relation-entity 2").

[0060] Neo4j is a high-performance graph database specifically designed for storing and querying graph-structured data. In the decision support module, Neo4j constructs and stores a knowledge graph in the field of waste heat power generation through structured knowledge triples, providing support for the decision generation unit.

[0061] Web-based visualization dashboards are data visualization tools implemented using web technologies. They can present complex data to users in the form of intuitive charts, dashboards, and other formats, helping users quickly understand and analyze the data.

[0062] This embodiment provides a back-pressure power generation device that utilizes waste heat and pressure from steam. Please refer to [link / reference]. Figure 2 This is a structural diagram of the device provided in an embodiment of the present invention.

[0063] A back-pressure power generation device utilizing waste steam heat and pressure in this embodiment includes: Inlet regulating valve: Used to regulate the flow rate of waste heat steam from the heat source; Turbine: Used to turbine waste heat steam, converting the thermal energy of the waste heat steam into mechanical energy, and then transmitting the turbined process steam to downstream steam-using equipment through the process header. Generator: Connected to the turbine via a coupling, it is used to convert the mechanical energy of the turbine into electrical energy and transmit it to downstream electrical equipment; Back-end steam supply equipment: used to receive and use process steam output from the turbine; Back-end electrical equipment: used to receive and use the electrical energy output by the generator; Backend demand monitoring module: used to monitor the steam demand of backend steam-consuming equipment in real time and obtain backend demand data; Intelligent monitoring system: used to monitor the operation of turbines and generators to obtain monitoring data, receive backend demand data from the backend demand monitoring module, preprocess and extract features from the monitoring data and backend demand data to obtain system operation characteristics, perform digital twin modeling and operating condition prediction of the system based on the system operation characteristics, and perform operation control of the system based on the results of the operating condition prediction.

[0064] Furthermore, it should be noted that the present invention can be provided as a method, apparatus, or computer program product. Therefore, embodiments of the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0065] Embodiments of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0066] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing terminal equipment to cause a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0067] It should also 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. The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0068] Finally, it should be noted that the above description represents a preferred embodiment of the present invention. It should be pointed out that although preferred embodiments have been described, those skilled in the art, once they understand the basic inventive concept of the present invention, can make various improvements and modifications without departing from the principles described herein. These improvements and modifications should also be considered within the scope of protection of the present invention. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

Claims

1. A back-pressure power generation control system for utilizing waste heat and pressure from steam, characterized in that, include: Inlet regulating valve: Used to control the flow rate of waste heat steam from the waste heat source; Turbine generator: Waste heat steam is fed into a turbine to generate steam and electricity for process use, and the turbine power generation process is monitored in real time to obtain monitoring data of the turbine power generation module; Backend demand monitoring module: Used to monitor the demand of backend steam-consuming equipment in real time and obtain backend demand data; Feature extraction module: preprocesses and extracts features from monitoring data and backend requirement data to obtain system operation features; Data modeling module: Builds a digital twin model based on the system's operational characteristics; Operating condition prediction module: Predicts the operating conditions of the system based on the digital twin model and generates control vectors based on the prediction results; The operation control module generates control commands based on the control vector, thereby controlling the system's operation. Decision Support Module: Used to provide intelligent decision-making suggestions to operators through knowledge graph technology; The turbine generator includes a turbine unit, a power generation unit, a bypass valve unit, and a monitoring unit; The turbine unit is used to convert the thermal energy of pretreated steam into mechanical energy through a turbine, output process steam and introduce the process steam into the process header, and the process header is used to transmit the process steam to the downstream steam-using equipment. The power generation unit is used to connect the output shaft of the turbine to the input shaft of the generator through a coupling, convert the mechanical energy of the turbine into electrical energy, and transmit the electrical energy to the downstream electrical equipment. The bypass valve unit is used to immediately open the bypass valve when the turbine trips due to a fault or in an emergency shutdown. The steam inside the turbine is de-cooled and de-pressurized by the desuperheater and pressure reducer and then introduced into the process header. The monitoring unit is used to monitor the inlet steam flow of the turbine in real time through a flow sensor, the inlet steam pressure of the turbine and the steam pressure of the process header in real time through a pressure sensor, the inlet steam temperature of the turbine in real time through a temperature sensor, and the power generation of the generator in real time through an electrical parameter sensor. The monitoring unit collects the above monitoring data through a PLC controller and transmits the monitoring data to the feature extraction module. The backend demand monitoring module includes a demand definition unit, a data acquisition unit, and a data transmission unit; The demand definition unit is used to divide the demand for steam from the back-end equipment to be sampled into multiple demand parameters, and define corresponding acquisition rules according to the demand parameters. The acquisition rules include the acquisition equipment corresponding to the demand parameters and its deployment location, signal interface and acquisition cycle. The data acquisition unit is used to collect the demand of back-end steam-consuming equipment through the acquisition device according to the acquisition rules, and obtain back-end demand data. The data transmission unit is used to integrate the backend demand data through the edge gateway and transmit the backend demand data to the feature extraction module via the MQTT protocol; The feature extraction module includes a data preprocessing unit, a monitoring feature extraction unit, a demand feature extraction unit, and a data fusion unit. The data preprocessing unit is used to clean, format and normalize the monitoring data and the backend demand data respectively to obtain preprocessed monitoring data and preprocessed backend demand data. The monitoring feature extraction unit is used to perform data dimensionality reduction and principal component extraction on the preprocessed monitoring data using the principal component analysis (PCA) algorithm to obtain the turbine generator operating characteristics. The demand feature extraction unit is used to extract features from the preprocessed backend demand data using the TCN-feature extraction method to obtain backend steam load features; The data fusion unit is used to synchronize and fuse the monitoring features and backend demand features in time using an adaptive Kalman fusion method to obtain system operation features. The data modeling module includes a model building unit, a model optimization unit, and a model visualization unit; The model building unit is used to build a geometric 3D model of the system in the MATLAB modeling environment based on the system structure, and to build a thermodynamic-fluid dynamic coupling model of the system based on the system operating characteristics. The model optimization unit is used to fuse the geometric 3D model and the thermodynamic-fluid dynamic coupling model of the system using the multiphysics coupling tool in the MATLAB environment to obtain the digital twin model of the system, and to verify and optimize the digital twin model of the system by the back-substitution-residual analysis method. The model visualization unit is used to build a visualization interface through the App Designer in the MATLAB environment to visualize and display the digital twin model and its operation. The operating condition prediction module includes a model loading unit, a prediction execution unit, and a control requirement analysis unit. The model loading unit is used to obtain a digital twin model from the data modeling module, set the model parameters of the digital twin model, and generate a digital twin model instance for working condition simulation. The model parameters include the initial conditions and boundary conditions of the digital twin model, the time step and total duration of the dynamic simulation. The prediction execution unit is used to perform dynamic simulation of the digital twin model instance according to the model parameters, and predict the operating status of the system through the PIML fusion algorithm to obtain prediction information. The prediction information includes the system's performance, load and fault risk under the current and future operating conditions. The performance includes power generation efficiency and heat recovery efficiency. The load includes operating load, maximum load and load change rate. The fault risk includes fault type, fault probability, fault time and fault impact range. The control requirement analysis unit is used to further optimize the dynamic simulation results using a genetic algorithm to obtain the optimal operating condition trajectory, and then convert the optimal operating condition trajectory into a control vector using a rolling time-domain linearization method. The control vector includes the controlled object, control parameters, and control signal strength.

2. The back-pressure power generation control system for utilizing waste steam heat and pressure according to claim 1, characterized in that: The inlet regulating valve is used to connect the waste heat source to the turbine generator through a pipeline. The system controls the flow rate of waste heat steam entering the turbine generator by adjusting the opening of the inlet regulating valve.

3. The back-pressure power generation control system for utilizing waste steam heat and pressure according to claim 1, characterized in that: The operation control module includes an adaptive optimization unit, an instruction decomposition unit, and an instruction execution unit; The adaptive optimization unit is used to perform online compensation of the control vector based on the operating characteristics using an adaptive optimization algorithm; The instruction decomposition unit is used to map the control vector into system control instructions through a static mapping table. The control instructions include inlet regulating valve control instructions and bypass valve control instructions. The instruction execution unit is used to transmit the control instructions via an industrial fieldbus, thereby controlling the operation of the system.

4. The back-pressure power generation control system for utilizing waste steam heat and pressure according to claim 1, characterized in that: The decision support module includes a knowledge extraction unit, a graph construction unit, and a decision generation unit. The knowledge extraction unit is used to obtain multi-source heterogeneous information in the field of waste heat power generation from DCS real-time database, CMMS maintenance work orders, standard operating procedures (SOPs) and expert interview records, and generate structured knowledge triples through entity relation extraction method. The graph construction unit is used to construct a knowledge graph in the field of waste heat power generation based on the structured knowledge triples and the Neo4j graph database. The decision generation unit is used to generate operational decision suggestions based on system operating characteristics and knowledge graph, and push the decision suggestions to operators through a web-based visual dashboard. The decision suggestions include equipment maintenance suggestions and fault prevention suggestions.

5. A back-pressure power generation device utilizing waste steam heat and pressure, used in the back-pressure power generation control system for utilizing waste steam heat and pressure as described in claim 1, characterized in that, include: Inlet regulating valve: Used to regulate the flow rate of waste heat steam from the heat source; Turbine: Used to turbine waste heat steam, converting the thermal energy of the waste heat steam into mechanical energy, and then transmitting the turbined process steam to downstream steam-using equipment through the process header. Generator: Connected to the turbine via a coupling, it is used to convert the mechanical energy of the turbine into electrical energy and transmit it to downstream electrical equipment; Back-end steam supply equipment: used to receive and use process steam output from the turbine; Back-end electrical equipment: used to receive and use the electrical energy output by the generator; Backend demand monitoring module: used to monitor the steam demand of backend steam-consuming equipment in real time and obtain backend demand data; Intelligent monitoring system: used to monitor the operation of turbines and generators to obtain monitoring data, receive backend demand data from the backend demand monitoring module, preprocess and extract features from the monitoring data and backend demand data to obtain system operation characteristics, perform digital twin modeling and operating condition prediction of the system based on the system operation characteristics, and perform operation control of the system based on the results of the operating condition prediction.

Images