Intelligent heat-flow coupled variable pressure circulating water and full life cycle scale prevention method and system

Abstract

This invention relates to the fields of non-ferrous metallurgical engineering, industrial circulating water treatment, and industrial artificial intelligence control technology. It discloses an intelligent heat-fluid coupled variable pressure circulating water system and a full-lifecycle scale prevention method and system. The method includes: the system powering on to generate a real-time state vector that has undergone quality verification; the real-time state vector being pushed to an edge computing unit or PLC coprocessor, outputting predicted values ​​of key thermal and water quality quantities within a future prediction window; obtaining risk prediction results and anomaly diagnosis conclusions; inputting the risk prediction results and anomaly diagnosis conclusions to an integrated setter, which is continuously updated in each strategy cycle based on constraint priority and safety coverage logic; and the updated sewage discharge setpoints and chemical dosing setpoints being sent to the heat-fluid coupled execution layer and the water quality dynamic balance execution layer. This invention effectively improves the operating efficiency and safety of the circulating water system, reduces maintenance costs, and extends the service life of the equipment.

Description

Technical Field

[0001] This invention relates to the fields of non-ferrous metal metallurgical engineering, industrial circulating water treatment and industrial artificial intelligence control technology, and in particular to an intelligent heat-fluid coupled variable pressure circulating water system and a scale prevention method and system throughout its entire life cycle. Background Technology

[0002] In the pyrometallurgical process of heavy non-ferrous metals, the cooling system of the smelting furnace, such as top-blown and side-blown furnaces, is the lifeline for ensuring the safe operation of the furnace. The cooling system typically consists of a cooling water jacket or copper water jacket, circulating pipelines, pump stations, cooling towers, and water treatment facilities. With the development of smelting equipment towards larger scale and higher strength, the heat load inside the furnace is extremely high and fluctuates drastically. Existing circulating water systems mainly face two core technical bottlenecks: 1) Mismatch between cooling supply and heat load demand, resulting in high energy consumption and low safety margin. Traditional circulating water systems mostly adopt constant pressure water supply or a simple constant flow mode; however, the heat load in the smelting process is dynamic, affected by the amount of feed, reaction intensity, and slag layer thickness. Overcooling problem: During low heat load periods, such as when holding the furnace for material preparation, a constant large flow water supply leads to wasted power consumption at the pump station and easily causes the surface temperature of the water jacket to become too low, triggering dew point corrosion of the acidic SO3 gas in the flue gas. Risk of insufficient cooling: During periods of high heat load, such as strong oxidizing spraying, insufficient water supply pressure can easily lead to localized nucleation boiling, film boiling, or even vaporization on the inner wall of the water jacket, resulting in deteriorated heat transfer and potentially causing the water jacket to burn through. Lack of pressure-temperature linkage: Focusing only on the outlet water temperature while ignoring the effect of water supply pressure on boiling point makes it impossible to effectively suppress the formation of vaporization nuclei under high-temperature conditions. 2) Scaling and corrosion in the circulating water system lead to heat exchange failure. Scaling, mainly composed of calcium carbonate, calcium sulfate, and suspended solids, is the main reason for reduced heat exchange efficiency and shortened water jacket life. Delayed water quality control: Existing technologies mostly rely on manual periodic sampling to test hardness, alkalinity, and chloride ions, and sewage discharge and chemical dosing are based on experience; this is highly delayed and cannot respond in real time to the drastic fluctuations in COC caused by evaporation and concentration in the cooling tower. Fluid dynamics design defects: Some pipeline layouts have velocity dead zones or vortex zones, leading to suspended solids settling; large pipe wall roughness makes it easy for crystal nuclei to attach and grow. Chemical Dependence and Environmental Burden: Over-reliance on chemical scale inhibitors and bactericides not only increases operating costs but also makes wastewater treatment more difficult, which is inconsistent with the development trend of green metallurgy. Furthermore, modern industrial sites involve large amounts of data, strong variable coupling, and frequent operating condition changes: relying solely on fixed thresholds or traditional PID control is insufficient to achieve optimal control across operating conditions and early risk identification. Therefore, it is urgent to introduce artificial intelligence technology to achieve heat load and water quality trend prediction, anomaly detection and early warning, adaptive multi-objective optimization control, digital twins, and predictive maintenance, forming a safe, efficient, and green intelligent circulating water system throughout its entire lifecycle. Summary of the Invention

[0003] The main objective of this invention is to provide an intelligent thermal-fluid coupled variable pressure circulating water system and a full life-cycle scale prevention method and system to solve the problems in the existing technology where modern industrial sites have large amounts of data, strong variable coupling, and frequent operating condition switching: relying solely on fixed thresholds or traditional PID controllers makes it difficult to achieve optimal control across operating conditions and early risk identification.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A smart heat-fluid coupled variable pressure circulating water system and a full-lifecycle scale prevention method, the smart heat-fluid coupled variable pressure circulating water system and the full-lifecycle scale prevention method comprising: After the system is powered on, the PLC / DCS loads the basic parameters, boundary conditions and engineering data, and collects thermal, water quality and operation and maintenance data in real time to form a real-time status vector that has been verified by quality. The real-time state vector is pushed to the edge computing unit or PLC coprocessor, and rolling prediction is performed based on historical window data and exogenous disturbances. The predicted values ​​of key thermal and water quality quantities within the future prediction window are output. At the same time, rapid risk identification based on physical margin and anomaly diagnosis based on the consistency of prediction-measured residuals are performed in parallel to obtain risk prediction results and anomaly diagnosis conclusions. Risk prediction results and anomaly diagnosis conclusions are input into the integrated setter, which is rolled up in each strategy cycle according to constraint priority and safety coverage logic: determine the zonal flow setpoint, main pipe pressure setpoint and target concentration factor to take the maximum feasible value that meets the water quality boundary conditions, and calculate the sewage discharge setpoint and chemical dosing setpoint. The updated sewage discharge and chemical dosing settings are sent to the heat-fluid coupling execution layer and the water quality dynamic balance execution layer.

[0005] As a further improvement of the present invention, the process of forming a real-time state vector that has undergone quality verification includes the following steps: The system collects thermal hydraulic, water quality and actuator status data via industrial Ethernet or fieldbus, including zone temperature, pressure, flow rate, wall temperature, main pipe pressure, total flow rate, pump and valve frequency or opening feedback, as well as conductivity, pH, turbidity, sewage flow rate and makeup water flow rate. After being timestamped, the data is formed into a multi-source data stream with a unified time scale. Perform range verification, rate of change verification, and consistency verification on the data stream with unified time scale; use moving average or exponential filtering to process short-term noise; and adopt a strategy of preserving the last valid value and conservative margin for missing or unavailable data to obtain a valid dataset that has undergone quality verification and fault tolerance processing. Intermediate derived quantities are calculated online based on effective datasets, including zone temperature rise, zone heat load, water-side wall temperature and saturated vapor pressure. The physical property parameters are obtained by online table lookup or function calculation, and the structural parameters are traced from drawings or material manuals. Based on intermediate derived quantities and calibrable parameters, the final derived quantities are further calculated, including anti-vaporization margin, concentration factor, TDS value and scaling tendency index LSI, forming a real-time state vector for prediction, diagnosis and setpoint calculation.

[0006] As a further improvement of the present invention, the process of outputting the predicted values ​​of key thermal and water quality quantities within the future prediction window includes the following steps: The system collects thermal hydraulic, water quality and actuator status data through industrial Ethernet or fieldbus. After timestamp alignment and data quality verification, it constructs a historical window and status vector containing zone inlet and outlet temperatures, wall temperatures, pressures, flow rates, main pipe pressures, total flow rates, conductivity, pH, turbidity, sewage flow rates, makeup water flow rates and optional furnace-side operation quantities. The historical window and state vector are input into the gray box prediction module of the mechanism model and data-driven residual correction. The thermal and water quality baseline prediction values ​​are obtained by the mechanism calculation of heat load autoregression and water quality conservation. Then, the unmodeled disturbances are corrected by data-driven residual compensation. The key prediction quantities of the zone heat load, wall temperature, anti-vaporization pressure requirement, circulating water conductivity, concentration factor and scaling tendency index in the future prediction window are output. The uncertainty of key forecast quantities is characterized to obtain the standard deviation or quantile interval of the forecast error. Combined with the configurable safety factor, a conservative forecast result is formed and passed to the setpoint rolling update interface as a feedforward correction quantity to generate the setpoints for main pipe pressure, zone flow, target concentration factor, sewage flow and chemical dosage with safety margin. The predictive model deployed on the edge computing unit or PLC coprocessor is put into online operation, and a sliding window is used to perform real-time statistics on the mean and variance of the prediction residuals. When the drift index exceeds the set threshold, the system automatically switches to the pure mechanism conservative prediction mode and triggers the model retraining prompt.

[0007] As a further improvement of the present invention, the process for obtaining risk prediction results and anomaly diagnosis conclusions including types such as vaporization margin, deposition risk, and equipment degradation includes the following steps: The online collected thermal hydraulic and water quality data and rolling prediction results are input into the physical margin calculation module to calculate the anti-vaporization margin, anti-deposition margin and minimum value within the prediction window of each zone in real time. After two-level threshold judgment, the module outputs an instant risk warning signal, including warning, alarm and strong cooling trigger command. The measured and predicted values ​​are used to construct a residual vector. After standardization, the Mahalanobis distance is calculated based on the covariance matrix of historical normal data. The result of the system anomaly detection is obtained by comparing it with a preset threshold. Then, the source of the anomaly is located based on the contribution of the residual components, such as temperature, pressure, flow rate or water quality. The anomaly type and severity are output. Characteristic residuals are constructed for key rotating equipment and pipeline components, including pump head residuals and efficiency consistency indicators, valve command-feedback-flow consistency residuals, and equivalent roughness based on pressure drop inversion. After trend analysis or change point detection, equipment deterioration trend warnings and maintenance suggestions are output. The physical margin warning, residual diagnosis and location results, and equipment degradation trend information are summarized into the anomaly comprehensive assessment unit. After multi-source information fusion and hierarchical logic judgment, the system anomaly level, partition location and suggested actions are output, and the control layer is linked to execute safety coverage, force pressure increase and flow increase, switch to standby equipment, trigger backwashing and upload maintenance work orders.

[0008] As a further improvement of the present invention, the process of sending the updated sewage discharge setpoint and chemical dosing setpoint to the heat-fluid coupling execution layer and the water quality dynamic balance execution layer includes the following steps: The online collected inlet and outlet water temperatures, pressures, and flow rates of each zone are input into the heat load identification module. The instantaneous heat load of each zone is obtained through energy conservation calculation. The heat load of each zone is combined with the preset target temperature rise and minimum anti-deposition flow velocity. After constraint take-maximum operation, the target flow rate set value of each zone is generated. The zone heat load and the measured temperature of the wall temperature thermocouple are input into the heat conduction inversion module. The water-side wall temperature is calculated by combining the known burial depth and thermal conductivity. The water-side wall temperature is input into the saturated vapor pressure calculation module. The saturated pressure is obtained by the standard physical property equation and superimposed with the safety margin to form the minimum target pressure of the zone. The minimum target pressure of the zone is superimposed with the pressure drop and elevation difference along the pipeline network. After the maximum value is calculated for the most unfavorable zone, the target pressure set value of the main pipeline is generated. The target pressure of the main pipeline, the target flow rate of each zone, the performance curve of the pump set, and the resistance characteristics of the pipeline network are input into the optimization solver. With the goal of minimizing the total power of the pump station, under the conditions of meeting the upper limit of wall temperature, temperature difference constraint, anti-vaporization constraint, and minimum flow velocity constraint, the target speed of the parallel pump set and the target opening of the regulating valve of each zone are obtained through nonlinear programming, forming the pump-valve coordinated setpoint. The target speed of the pump set and the target opening of the valve are sent to the frequency converter and the electric actuator. The inner loop controller realizes the closed-loop tracking of the main pipe pressure and the closed-loop distribution of the zone flow, and provides real-time feedback of the actual value and compares it with the set value. When the wall temperature exceeds the limit or the vaporization margin is lower than the threshold, the strong cooling mode is triggered, which forcibly increases the main pipe pressure and the relevant zone flow until the risk is eliminated.

[0009] As a further improvement of the present invention, the process of sending the updated sewage discharge setting value and chemical dosing setting value to the water quality dynamic balance execution layer includes the following steps: The online collected conductivity of circulating water and the conductivity of makeup water are input into the concentration factor calculation module, and the real-time concentration factor is obtained through ratio calculation; the conductivity of circulating water is input into the conductivity-TDS calibration model, and the total dissolved solids concentration is obtained by converting the calibration coefficients based on periodic test regressions, forming the basic parameters of water quality status; The total dissolved solids concentration, online pH value, the most unfavorable zone water side wall temperature calculated by the thermal loop, and the calcium hardness and alkalinity of the makeup water are updated online or by laboratory testing and input into the Langier saturation index calculation module. The real-time scaling tendency index is calculated by the standard Langier formula and used as a water quality constraint criterion. The real-time scaling tendency index, the upper limit of total dissolved solids concentration, the upper limit of chloride ion concentration, and the makeup water quality parameters are input into the target concentration factor optimization solution module. Under the condition of meeting the scaling and water quality safety constraints, the maximum feasible concentration factor is searched to obtain the target concentration factor decision value. The target concentration factor decision value, along with the cooling tower evaporation rate and drift rate, are input into the wastewater discharge calculation module. Based on the mass conservation relationship, the wastewater discharge setpoint is obtained. The wastewater discharge setpoint, drift rate, and target inhibitor concentration are input into the dosage calculation module. Based on the effective component balance, the metering pump volume flow rate setpoint is obtained, forming a closed-loop control command for the wastewater discharge valve and the dosing pump.

[0010] As a further improvement of the present invention, it also includes anti-scaling water jacket branch pipes and key pipe networks, low-adhesion lining / high-gloss material pipe sections, long-radius elbows and downstream transition connectors, high-point gas collection and automatic exhaust structure of the system, and passive anti-scaling structure with physical field scale inhibition device arrangement at key inlets.

[0011] As a further improvement of the present invention, in critical areas with high temperature and easy scale formation, a polytetrafluoroethylene (PTFE) lined tube or a stainless steel tube is used. The low surface energy of PTFE tube and the high corrosion resistance of stainless steel are used to reduce the adhesion of crystal nuclei and the bonding strength of scale layer, and to inhibit the formation of corrosion products as a secondary nucleation substrate. Insulating gaskets / insulating joints are installed at the joints of dissimilar metals, and the inner wall of the weld is ground and rounded.

[0012] As a further improvement of this invention, the high-incidence locations of scaling correspond to low-velocity stagnation zones, localized backflow vortices, and reversal separation zones, imposing quantifiable geometric constraints on pipe fittings and connection methods: all elbows must be long-radius elbows and meet the following requirements. Meanwhile, at the branch junctions, downstream tees, oblique tees, or transition pieces with controlled diffusion angles are used.

[0013] To achieve the above objectives, the present invention also provides the following technical solution: A smart thermal-fluid coupled variable pressure circulating water system and its full life-cycle scale prevention system are applied to the aforementioned smart thermal-fluid coupled variable pressure circulating water system and its full life-cycle scale prevention method. The smart thermal-fluid coupled variable pressure circulating water system and its full life-cycle scale prevention system includes: The cooling circuit of the smelting furnace adopts a water jacket monitoring array, and thermocouples are embedded in the key cooling water jacket to monitor the copper wall temperature. High-precision temperature sensors, pressure transmitters and flow meters are installed in each branch pipeline; low roughness pipeline layout: the inner wall of the main circulating water pipeline and branch pipes is made of plastic-lined composite pipe or stainless steel pipe. The intelligent variable pressure water supply pump station is equipped with multiple parallel variable frequency centrifugal pumps; the main outlet pipe is equipped with an electric regulating valve and a bypass pressure relief valve to work with the frequency converter to achieve a wide range of pressure-flow decoupling regulation; it is equipped with pump motor current / power acquisition and pump body vibration monitoring for health diagnosis and efficiency evaluation. The online water quality control and scale prevention unit includes online monitoring instruments, automatic dosing and sewage discharge devices, and bypass physical treatment facilities. The online monitoring instruments include a conductivity meter, a circulating water and makeup water meter, a pH meter, and a turbidity meter. The automatic dosing and sewage discharge devices automatically add corrosion and scale inhibitors and acid according to controller commands; the sewage discharge valve automatically opens and closes according to the concentration ratio threshold. The bypass physical treatment facilities include a bypass filtration system with a treatment capacity of 3% to 5% of the total water volume, integrating a self-cleaning filter to remove suspended solids and a filter to change crystal morphology. The data acquisition and edge computing unit is an industrial gateway and edge computing terminal, an industrial PC / embedded system, which performs: data time alignment, outlier removal, missing value repair, and feature extraction, including features such as dT / dt, wall temperature gradient, and vaporization margin; it also interacts bidirectionally with PLC / DCS: outputting predicted setpoints, anomaly alarms, and optimized control quantities; and it supports industrial protocols. The modeling and inference unit can be deployed on in-plant servers or at the edge; it includes: heat load prediction model, water quality trend prediction model, anomaly detection model, and control strategy optimization model; it supports model version management, online calibration, and rollback strategies, and can roll back to rule / traditional control when the model confidence is insufficient; A digital twin and operation and maintenance decision-making platform is established to create a digital twin of the circulating water system: topological hydraulic network, heat-fluid mechanism, scaling / corrosion evolution, and data-driven agent; providing 3D / 2D visualization, simulation, maintenance plan output, remote diagnosis, and reports.

[0014] To achieve the above objectives, the present invention also provides the following technical solution: An electronic device includes a processor and a memory coupled to the processor, the memory storing program instructions executable by the processor; when the processor executes the program instructions stored in the memory, it implements the intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention method as described above.

[0015] To achieve the above objectives, the present invention also provides the following technical solution: A storage medium storing program instructions, which, when executed by a processor, enable the implementation of the intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention method described above.

[0016] This invention Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall structure and process flow of the intelligent heat-fluid coupled variable pressure circulating water system for heavy non-ferrous metal molten pool smelting according to the present invention. Figure 2 This is a schematic flowchart of one embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention; Figure 3 This is a diagram illustrating the process of outputting predicted values ​​of key thermal and water quality parameters within a future prediction window, as an embodiment of the intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention method of the present invention. Figure 4 This is a schematic diagram of the real-time data modeling and rolling prediction module architecture of an embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention. Figure 5 This is a schematic diagram illustrating the abnormal detection and early warning linkage of an embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention; Figure 6 This is a schematic diagram of the steps in which the updated sewage discharge setting value and chemical dosing setting value are sent to the thermal-fluid coupling execution layer in one embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention. Figure 7 This is a block diagram of the heat-fluid coupled variable pressure cooling control structure of an embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention; Figure 8 This is a schematic diagram of water quality dynamic balance scale prevention control in an embodiment of the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method of the present invention; Figure 9 This is a schematic diagram of the anti-scaling water jacket branch pipe and key pipeline structure of an embodiment of the intelligent heat-fluid coupling variable pressure circulating water and full life cycle anti-scaling method of the present invention; Figure 10 This is a functional module diagram of an embodiment of the intelligent heat-fluid coupling variable pressure circulating water and full life cycle scale prevention system of the present invention; Figure 11 This is a schematic diagram of the structure of an embodiment of the electronic device of the present invention; Figure 12This is a schematic diagram of the structure of one embodiment of the storage medium of the present invention. Detailed Implementation

[0018] 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 a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] The terms "first," "second," and "third" in this invention are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features. In the description of this invention, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this invention are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indication changes accordingly. Furthermore, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0020] References to embodiments herein mean that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0021] This invention specifically relates to a cooling water circulation system for large-scale smelting furnaces of heavy non-ferrous metals such as copper, lead, and nickel, including top-blown furnaces, side-blown furnaces, bottom-blown furnaces, and flash furnaces. It particularly relates to a pressure-flow coordinated intelligent control technology based on feedback from inlet and outlet water temperatures and water jacket wall temperatures, and a comprehensive scale prevention and predictive maintenance solution integrating online concentration ratio monitoring, electrochemical / electromagnetic physical water treatment, pipeline rheological optimization, artificial intelligence predictive control, anomaly diagnosis, and digital twin operation and maintenance. This invention provides an AI-enhanced intelligent heat-flow coupled variable pressure circulating water system and a full-lifecycle scale prevention and predictive maintenance method for heavy non-ferrous metal smelting. Its core lies in: heat-flow coupled variable pressure control: establishing a multivariable correlation model of water jacket wall temperature, inlet and outlet water temperature difference, and supply water pressure. Instead of simply adjusting the flow rate, it dynamically adjusts the system supply water pressure or saturation pressure margin and flow rate through the coordinated operation of variable frequency pump sets and electric regulating valves, achieving a balance between on-demand cooling and boiling suppression. Dual-Driven Scale Prevention Based on Water Quality and Operating Conditions: Upgrading traditional periodic blowdown to automated blowdown and chemical dosing based on online calculations of concentration ratio (COC) and Langerile saturation index (LSI); simultaneously introducing physical fields, such as electrochemical and magnetized water treatment, and eliminating scaling factors from equipment materials and flow field layout. AI Real-Time Modeling and Prediction: Utilizing multi-source sensor data to construct time-series predictive models of heat load, water concentration, and scaling trends, predicting wall temperature, vaporization margin, and scaling index within future rolling windows, providing feedforward setpoints and risk advances for the controller. AI Anomaly Detection and Early Warning, Adaptive Control, and Digital Twin Maintenance: Constructing self-learning detection models for anomalies such as vaporization risk, blockage, pump / valve failure, and sensor drift, forming early warning and fault-tolerant switching; introducing reinforcement learning / learning-based predictive control to achieve multi-objective adaptive optimization of pump frequency, valve position, blowdown, and chemical dosing, replacing fixed-parameter PID; constructing a mechanistic model and data-driven agent-based circulating water digital twin for online simulation, scenario drills, health assessment, and predictive maintenance decision-making. Figure 1 The diagram shows the overall structure and process flow of the intelligent heat-fluid coupled variable pressure circulating water system for heavy non-ferrous metal smelting. Figure 1The diagram illustrates the data and control connections between the smelting furnace cooling water jacket circuit, the variable frequency pump station, the zoned flow distribution valve group, the cooling tower, the bypass filtration and physical field scale inhibition unit, the online water quality analysis unit, the automatic sewage discharge and chemical dosing device, and the PLC / DCS and edge computing unit. The smelting furnace cooling water jacket circuit is connected to the variable frequency pump station, the PLC / DCS to the edge computing unit, and the cooling tower. The variable frequency pump station is connected to the zoned flow distribution valve group and the PLC / DCS to the edge computing unit. The zoned flow distribution valve group is connected to the PLC / DCS to the edge computing unit, the bypass filtration and physical field scale inhibition unit, and the online water quality analysis unit. The online water quality analysis unit is connected to the automatic sewage discharge and chemical dosing device. The automatic sewage discharge and chemical dosing device is connected to the bypass filtration and physical field scale inhibition unit. The bypass filtration and physical field scale inhibition unit is connected to the cooling tower. The cooling tower is connected to the PLC / DCS and the edge computing unit via data and control connections.

[0022] like Figure 2 As shown, this embodiment provides an example of an intelligent heat-fluid coupled variable pressure circulating water system and a full life-cycle scale prevention method. In this embodiment, the intelligent heat-fluid coupled variable pressure circulating water system and the full life-cycle scale prevention method specifically include the following steps: Step S1: After the system is powered on, the PLC / DCS completes the loading of basic parameters and boundary conditions, and loads engineering data such as pump and valve characteristic curves, pipeline geometric parameters and water quality calibration coefficients; real-time acquisition of thermal, water quality and operation and maintenance quantities, and after range, rate of change and consistency verification and abnormal data filtering and elimination, a real-time status vector with quality verification is formed. Step S2: The real-time state vector is pushed to the edge computing unit or PLC coprocessor, and rolling prediction is performed based on historical window data and exogenous disturbances. The predicted values ​​of key thermal and water quality quantities within the future prediction window are output. At the same time, rapid risk identification based on physical margin and anomaly diagnosis based on the consistency of prediction-measured residuals are executed in parallel to obtain risk prediction results and anomaly diagnosis conclusions including vaporization margin, deposition risk, equipment degradation, etc. Step S3: The risk prediction results and anomaly diagnosis conclusions are input into the integrated setting device, and are updated on a rolling basis according to constraint priority and safety coverage logic in each strategy cycle: the zonal flow setting value is determined by taking the maximum value of the minimum anti-deposition, on-demand heat dissipation and risk suppression compensation flow; the main pipe pressure setting value is dominated by the anti-vaporization requirement of the most unfavorable zonal and superimposed with the pipeline network loss and uncertainty margin; the target concentration factor is taken as the maximum feasible value that meets the water quality boundary conditions, and the sewage discharge setting value and chemical dosing setting value are calculated; when a high-level risk is determined, safety coverage is implemented on the sewage discharge setting value and chemical dosing setting value to prioritize the protection of intrinsic safety constraints; Step S4: The updated sewage discharge setpoints and chemical dosing setpoints are sent to the heat-flow coupling execution layer and the water quality dynamic balance execution layer: The heat-flow loop uses the main pipe pressure as the main control variable for variable frequency speed regulation closed loop and the zone flow rate as the main control variable for valve position adjustment closed loop. The water quality loop executes sewage discharge-water replenishment-chemical dosing closed loop control according to the target concentration ratio, sewage discharge setpoints and chemical dosing setpoints. At the same time, the bypass filtration and physical field treatment devices operate in coordination. Key variables, setpoints, alarm events and energy and chemical consumption indicators are continuously recorded, and the model retraining prompts or degradation strategies are triggered based on the statistical results of the prediction model residuals, forming a traceable operation file and a full life cycle maintenance closed loop.

[0023] Preferably, this embodiment ensures the accuracy and reliability of data by collecting real-time thermal, water quality, and operational data, and performing quality verification, providing accurate basis for prediction and risk assessment. Utilizing edge computing units or PLC coprocessors, combined with historical data and exogenous disturbances, rolling predictions are performed to forecast future key thermal and water quality quantities, while simultaneously identifying risks and diagnosing anomalies, improving the system's response speed and accuracy to potential problems. Within the strategy cycle, flow and pressure setpoints, as well as water quality management parameters, are dynamically updated according to constraint priorities and safety coverage logic, achieving intelligent control of zoned flow and main pipe pressure, effectively preventing sedimentation and vaporization risks, and optimizing water quality management. When the system determines a high-level risk, it automatically implements safety coverage for sewage discharge and chemical dosing to ensure that the system's inherent safety constraints are met, enhancing system safety. Closed-loop control of the heat-flow loop is achieved through variable frequency speed regulation and valve position adjustment. Closed-loop control of the water quality loop is executed according to the target concentration factor, sewage discharge setpoint, and chemical dosing setpoint. Simultaneously, the bypass filtration and physical field treatment devices operate in coordination, improving the system's control accuracy and stability. The system continuously records key variables, setpoints, alarm events, and energy and drug consumption indicators to form a traceable operation file. Based on the statistical results of model residuals, it triggers model retraining prompts or degradation strategies, realizing a closed-loop maintenance throughout the entire life cycle and improving the long-term stability and reliability of the system.

[0024] In this embodiment, the integrated control process and implementation steps employ a two-tiered structure: a strategy layer or predictor / diagnostic / setter layer, and an execution layer or pump / valve and water quality internal loop. The strategy layer updates the setpoints on a rolling basis at a fixed cycle, while the execution layer tracks the setpoints more frequently and executes mandatory safety actions when interlocking conditions are triggered. The specific process is as follows.

[0025] Step 1: System initialization and parameter loading. After the system is powered on, the PLC / DCS loads basic parameters and boundary conditions, including the upper limit of wall temperature. Upper limit of temperature difference Minimum anti-deposition flow velocity Anti-vaporization margin parameters and measurement uncertainty parameters, as well as water quality safety limits, such as Etc. Simultaneously load the pump performance curve fitting coefficients, valve... Calibration curves and pipeline geometry parameters Effective system volume Conductivity-TDS calibration coefficient Engineering parameters; all of the above parameters are derived from drawings, equipment data, commissioning calibration or laboratory regression, and have clear acquisition channels.

[0026] Step 2: Real-time data acquisition, alignment, and data quality verification; controller data acquisition. ,as well as Simultaneously collect thermal-fluid state parameters such as pump frequency / valve position feedback. Water quality and maintenance volume are monitored. After data collection, range / rate of change / consistency checks are performed, and outliers are filtered or removed. When key monitoring points are unavailable, a conservative mode is entered and an alarm is triggered.

[0027] Step 3: Online Derivation Calculation and State Vector Update, Calculation Equal derivation quantities are used to form the state vector for policy layer computation. At this point, the data link is closed, providing the input conditions for prediction, diagnosis, and setpoint calculation.

[0028] Step 4: Rolling predictive calculation, edge computing unit or PLC coprocessor based on historical window data and optional exogenous perturbations Rolling forecasts are performed on key thermal and water quality parameters, and the results within the future forecast window are output. The preferred prediction model adopts a gray box structure of mechanistic model + data-driven residual correction to ensure that the results are consistent with the conservation relationships and water property constraints; when the model drifts or becomes unavailable, the system automatically switches to mechanistic conservative prediction output and increases the safety margin coefficient.

[0029] Step 5: Anomaly Detection and Tiered Early Warning. The system performs parallel detection of two types of anomalies: one is rapid risk identification based on physical margin, for example, when anomalies occur within the prediction window... When the value approaches or exceeds the warning threshold, a risk of vaporization is determined; when When the pressure drop-flow relationship is abnormal, a risk of deposition / clogging is identified. Secondly, diagnosis is based on the consistency of predicted and measured residuals. For example, residual and distance indicators are constructed for key output quantities. When the residual continuously exceeds the limit, slow-changing anomalies such as pump performance degradation, valve jamming, and sensor drift are identified. The system outputs alarms according to a three-level strategy of early warning, alarm, and interlock, and pushes the anomaly type, location zone, and suggested actions to the DCS HMI and the operation and maintenance system.

[0030] Step 6: Rolling update of setpoints. The integrated heat-fluid and water quality setpoint system updates and distributes key setpoints based on prediction results and constraint priorities within each strategy cycle: zone flow setpoints. The maximum value of the minimum anti-deposition flow rate, the on-demand heat dissipation flow rate, and the risk suppression compensation flow rate is used to ensure that the requirements for heat dissipation, vaporization suppression, and anti-deposition are met simultaneously; the main pipe pressure setpoint. The prediction is primarily based on the anti-vaporization pressure requirement of the most unfavorable zone within the forecast window, while also incorporating pipeline losses and uncertainty margins to ensure that the most unfavorable zone still meets the anti-vaporization margin; target concentration factor. Take the condition that satisfies the prediction window Maximum feasible values ​​for water quality boundary conditions such as chloride ions, to improve reuse rates and control scaling / corrosion risks; wastewater discharge setpoints. Based on the evaporation and concentration mechanism and Calculated; Dosing setpoint Based on the principle of maintaining the target effective concentration, a compensatory calculation is performed according to the amount of wastewater and drift carried away. The required proportion and density of the effective components of the reagent are all derived from reagent data or metrological calibration.

[0031] When an anomaly detection determines a high-risk level, the system implements a safety overriding of the above-mentioned settings: prioritizing elevation. And improve related Lower if necessary And improve To quickly reduce the risk of scaling, ensuring that intrinsic safety constraints are always the top priority.

[0032] Step 7: Thermal-fluid coupling execution layer control, the execution layer controls... and Implement rapid closed-loop tracking: The main pump frequency converter uses the main pipe pressure as the primary control variable to achieve rapid pressure regulation, while the zoned electric regulating valves use the zoned flow rate as the primary control variable to achieve flow distribution. If the forced cooling interlock condition is triggered, such as when the wall temperature approaches... If the vaporization margin exceeds the limit, the system will immediately enter the strong cooling mode: prioritize increasing the water supply pressure, simultaneously increase the flow rate of the relevant zones, and restrict energy-saving frequency reduction actions according to the interlock logic until the risk is eliminated and the switchback conditions are met.

[0033] Step 8: Dynamic water quality balance execution layer control, water quality loop according to... Implement a closed-loop system for sewage discharge, water replenishment, and chemical dosing: The sewage discharge valve should be set according to... Flow closed-loop control is implemented, with the water supply valve / pump automatically compensating based on system water level or flow rate, and the metering pump... Add scale inhibitors / corrosion inhibitors; when If the conductivity prediction indicates that the concentration is about to exceed the limit, the system will perform small-scale multiple flushing and dosage fine-tuning in advance to avoid a sudden increase in scale caused by a sudden increase in concentration.

[0034] Step 9: Side-flow filtration and physical field treatment work in tandem; the side-flow pump maintains continuous side-flow treatment; when turbidity... The backwashing procedure is automatically started when the limit is exceeded or the filter differential pressure exceeds the limit; the electromagnetic / electrochemical anti-scaling device performs frequency / power adaptive adjustment according to the bypass flow rate or the main circuit flow rate, and works in conjunction with the water quality circuit to avoid the cost and environmental burden caused by relying solely on chemical dosing.

[0035] Step 10: Operation Recording, Performance Evaluation, and Model Self-Correction Triggering. The system continuously records key variables, setpoints, alarm events, and indicators such as pump work, sewage discharge, water replenishment, and chemical consumption, forming a traceable operation file. The residual statistics of the predictive model are used to monitor model drift online. When the drift indicator exceeds the threshold, a model retraining prompt / degradation strategy is triggered to ensure stability and maintainability throughout the entire lifecycle.

[0036] The specific implementation of this invention focuses on on-site measurable data, online derived quantity calculation, rolling prediction and anomaly diagnosis, setpoint updates, pump and valve and water quality execution, and closed-loop verification and recording. It adopts an integrated prediction-diagnosis-control-maintenance control process. The system is deployed on an architecture primarily based on PLC / DCS and supplemented by edge computing units: the PLC / DCS is responsible for real-time data acquisition, inner-loop control, and interlock execution; the edge computing units are responsible for rolling prediction, anomaly detection, and strategy-level setpoint calculation, and write the results back to the PLC / DCS in the form of setpoints, thereby ensuring the real-time performance of the control closed loop and its engineering feasibility.

[0037] For data acquisition, verification, and online derived quantity calculation, the system collects and integrates thermal, hydraulic, water quality, and actuator status data via industrial Ethernet / industrial fieldbus to form a unified time-scaled data stream. Data acquisition preferably uses a sampling period of 1–5 seconds, and timestamp alignment and quality verification are performed internally within the controller to ensure that multi-source data is available within the same control cycle.

[0038] Thermal and hydraulic data acquisition was conducted, and zoned inlet water temperatures were configured in slag line areas, spray gun areas, and hot spot areas of each cooling zone. Outlet water temperature Partition pressure Partition traffic Alternatively, valve differential pressure can be measured using a soft sensor; thermocouples can be installed in the critical water jacket to measure the temperature at the depth of the copper wall burial. Configure a main pipeline pressure in the pump station main pipeline. Total flow In addition, actuator status parameters such as pump inverter frequency / speed, valve opening degree, and feedback value are also measured. All of these variables are standard industrial measurement points, and the signals can be accessed via 4–20mA, RS485, or Ethernet protocols.

[0039] Water quality and operation and maintenance data collection, configuration of circulating water conductivity Water conductivity Online pH value turbidity And record the sewage flow rate. Water replenishment flow rate In addition, maintenance-related parameters such as bypass filter flow rate and backwash status are also recorded. If an online hardness / alkalinity analyzer is available on-site, calcium hardness data can be directly collected. With alkalinity If this is not currently available, a soft measurement calibration model will be established using online conductivity / temperature / pH testing and periodic analysis to ensure accuracy. , The acquisition of these resources has a clear source and an update mechanism.

[0040] Data quality verification and fault tolerance processing: The controller performs range verification or upper and lower limits, rate of change verification or abrupt changes / drifts, and consistency verification or redundant measurement point differences or model prediction-measured residual verification for each measurement point. When short-term noise or single-point anomalies occur, moving average / exponential filtering is used to suppress high-frequency interference; when missing or unavailable data occurs, the last valid value is held and a conservative margin strategy is adopted within the allowed holding time. If the holding time is exceeded, degraded operation is triggered to ensure that the system remains safe even in the event of sensor malfunctions.

[0041] Online derived quantity calculation: Within each control cycle, the system calculates key derived quantities online based on measurable data, such as zone temperature rise. Zoned heat load , water physical properties Depend on Online table lookup / function calculation; zoned waterside wall temperature ,Depend on The required burial depth is obtained by inversion with heat flux. Copper thermal conductivity Heat exchange area All are traceable from drawings / materials manuals; saturated vapor pressure Calculated online using the standard water vapor pressure equation; anti-vaporization margin Obtained by parameterization of design margin and instrument uncertainty; Concentration factor Conductivity-TDS conversion , Obtained through laboratory calibration regression and can be updated periodically; Scaling tendency index ,in Take the most unfavorable partition The maximum value. All of the above derived quantities consist of measurable data, traceable structural parameters, and calibrable parameters, satisfying the requirements for engineering implementation and audit trail.

[0042] In summary, this embodiment effectively improves the operating efficiency and safety of the circulating water system, reduces maintenance costs, and extends the service life of the equipment through intelligent monitoring, prediction, control, and maintenance.

[0043] This embodiment, focusing on the overall goals of prioritizing intrinsic safety, minimizing energy consumption, and controlling scaling / corrosion risks, constructs an integrated control system with dual closed-loop control, predictive feedforward, and anomaly protection. Its foundational layer consists of two mutually coupled but clearly defined closed-loop loops: In the heat-fluid coupled variable pressure cooling control model, the inlet and outlet water temperatures, wall temperatures, pressures, and flow rates of each zone are used as core state variables. Online heat load identification and anti-vaporization pressure margin constraints are established. Through the coordinated regulation of variable frequency pumps and regulating valves, the joint optimization of water supply pressure and flow rate is achieved, thus balancing on-demand cooling and vaporization suppression under conditions of drastic fluctuations in furnace heat load. In the water quality dynamic balance anti-scaling control loop, conductivity, pH, turbidity, and online or soft-measured hardness / alkalinity are used as core water quality state variables. A calculable criterion for concentration ratio and scaling tendency is constructed, driving the closed-loop coordination of wastewater discharge, water replenishment, chemical dosing, and bypass treatment. This allows the system to maximize the reuse rate of circulating water and extend equipment life while meeting scaling and corrosion boundary conditions. Further, an intelligent enhancement layer is introduced: the real-time data modeling and prediction model constructs a consistent gray-box modeling and rolling prediction framework for two key variables: thermal performance and water quality. Utilizing historical window data and optional furnace-side exogenous disturbance inputs, it proactively predicts short-term heat load, wall temperature / anti-vaporization pressure requirements, and short-to-medium-term concentration ratios and scaling trends. The prediction results are then transformed into feedforward corrections and rolling updates for water supply pressure setpoints, zone flow setpoints, target concentration ratios, wastewater discharge, and chemical dosage. Simultaneously, the anomaly detection and early warning model uses physical margin indicators, such as anti-vaporization margin, A multi-layered mechanism, including low flow margin, predictive residual consistency diagnosis, and unsupervised anomaly identification / change point detection, enables early identification and graded warning of anomalies such as vaporization risk, blockage and scaling, pump and valve deterioration, and sensor drift. It also links with the control loop to trigger a strong cooling mode, mandatory minimum flow rate constraints, backup equipment switching, and maintenance decision output. By sharing key state variables and unified constraint boundaries, such as upper limit of wall temperature, minimum flow rate, pressure margin, and water quality safety threshold, the above four parts together form a prediction-diagnosis-control-maintenance closed-loop system for all operating conditions, ensuring long-term stable, safe, and efficient operation of the system.

[0044] Furthermore, such as Figure 3 As shown, step S2, which outputs the predicted values ​​of key thermal and water quality quantities within the future prediction window, specifically includes the following steps: Step S21: Collect thermal hydraulic, water quality and actuator status data through industrial Ethernet or fieldbus, and after timestamp alignment and data quality verification, construct a historical window and status vector containing zone inlet and outlet temperature, wall temperature, pressure, flow rate, main pipe pressure, total flow rate, conductivity, pH, turbidity, sewage flow rate, makeup water flow rate and optional furnace side operation quantity. Step S22: Input the historical window and state vector into the gray box prediction module of mechanism model and data-driven residual correction. The thermal and water quality baseline prediction values ​​are obtained by mechanism calculation based on heat load autoregression and water quality conservation. Then, the unmodeled disturbances are corrected by data-driven residual compensation. The key prediction quantities in the future prediction window are output, such as zone heat load, wall temperature, anti-vaporization pressure requirement, circulating water conductivity, concentration factor and scaling tendency index. Step S23: Characterize the uncertainty of key prediction quantities to obtain the standard deviation or quantile interval of prediction error. Combine the configurable safety factor to form a conservative prediction result and pass it as a feedforward correction quantity to the set value rolling update interface to generate the set values ​​of main pipe pressure, zone flow rate, target concentration factor, sewage flow rate and dosing amount with safety margin. Step S24: The prediction model deployed on the edge computing unit or PLC coprocessor is put into online operation. At the same time, a sliding window is used to perform real-time statistics on the mean and variance of the prediction residuals. When the drift index exceeds the set threshold, the system automatically switches to the pure mechanism conservative prediction mode and triggers the model retraining prompt to ensure the stability and maintainability of the prediction module throughout its entire life cycle.

[0045] Preferably, in this embodiment, the data acquired via industrial Ethernet or fieldbus undergoes timestamp alignment and data quality verification to ensure the accuracy and reliability of the input data, providing a high-quality data foundation. Combining mechanistic models and data-driven residual correction, it considers both the mechanistic calculations of heat load autoregression and water quality conservation, and compensates for unmodeled disturbances using data-driven methods, improving prediction accuracy. Uncertainty characterization of prediction quantities: Uncertainty characterization of key prediction quantities yields the standard deviation or quantile interval of prediction errors, improving the reliability of prediction results. Enhanced safety: Combining a safety factor to form a conservative prediction result and using it as a feedforward correction quantity generates a setpoint with a safety margin, enhancing system safety. Online operation and adaptive adjustment: Deploying the prediction model on an edge computing unit or PLC coprocessor, and using a sliding window to statistically analyze prediction residuals in real time, enables online operation and adaptive adjustment of the model, improving system stability and maintainability. Model stability assurance: When the drift index exceeds a set threshold, the system automatically switches to a pure mechanistic conservative prediction mode and triggers a model retraining prompt, ensuring the stability of the prediction module throughout its entire lifecycle.

[0046] In summary, this embodiment achieves high-quality data acquisition and preprocessing, improved prediction accuracy, enhanced reliability of prediction results, improved system security, online model operation and adaptive adjustment, and guaranteed model stability, thereby improving the overall prediction performance and reliability of the system.

[0047] In this embodiment, to enable feedforward predictive capabilities for the heat-fluid coupled variable pressure control and water quality dynamic balance anti-scaling control, the present invention introduces a real-time data modeling and rolling prediction mechanism on top of the dual closed-loop control framework, forming a closed-loop link of measurement-identification-prediction-setup update. The prediction module uses feedback from water-side sensors and pump valve actuators as basic inputs, and can optionally incorporate key operating variables from the furnace-side DCS, such as injection intensity and feed rate, as exogenous disturbances. This allows for early prediction of short-term heat load mutations and short-to-medium-term water quality concentration evolution, thereby further reducing energy consumption and wastewater discharge while meeting anti-vaporization and anti-scaling constraints. Figure 4 A schematic diagram of the architecture for the real-time data modeling and rolling prediction module. Figure 4 The diagram illustrates state vector construction and data quality verification, gray-box prediction using mechanistic models and data-driven residual correction, prediction uncertainty assessment, and the... The interface for feedforward correction and rolling update of set values.

[0048] Data structure, sampling, and state vector definition, assuming a sampling period of . In engineering practice, the time interval is typically taken as 1–5 seconds. The system state vector is: in: , Temperature sensors are used to measure the inlet and outlet temperatures of the zones. This refers to the temperature of the embedded wall-mounted thermocouple. For partition pressure; For zoned flow, use flow meters or valves for soft measurement; , These are the main pipe pressure and the total flow rate, respectively. The conductivity of circulating water and makeup water; For online pH; Turbidity; For sewage discharge flow rate, use a sewage branch flow meter or valve position-differential pressure soft measurement. For makeup water flow, a makeup water flow meter is used. All the above quantities can be obtained from field instruments, frequency converter / valve feedback, or soft measurement calibration, ensuring a closed data source. If furnace-side DCS data is available on-site, an exogenous input vector can be defined: As perturbation information, it enhances prediction accuracy; however, the prediction module of this invention uses only water-side variables. They can also work independently under certain circumstances.

[0049] The prediction module outputs target quantities across multiple time scales, using both prediction objectives and rolling prediction methods. Short-term thermal forecast: Used for early pressure boosting / flow enhancement to suppress vaporization; short- to medium-term water quality forecasting: Used for pre-emptive sewage discharge / chemical dosing to prevent scaling inflection points; among which Indicates the prediction step size, and the prediction time domain is... Generally, the thermal prediction window is set to 30–300 seconds, and the water quality prediction window is set to 0.5–24 hours, which can be configured according to operational requirements.

[0050] In a unified mathematical expression, the prediction model can be written as: in For length A historical window Provide an external input history window, if available; It can be implemented as a gated recurrent network (GRU / LSTM), a temporal convolutional network (TCN), or a feature-based gradient boosting tree model (GBDT / XGBoost), etc., with parameters... It is obtained through offline training based on historical operational data and then deployed in a fixed manner, with a clear acquisition channel.

[0051] Physically Consistent Gray-Box Prediction: Mechanism Model + Data-Driven Residual Correction. To ensure that the prediction results are consistent with thermodynamic / hydraulic constraints, this invention preferably employs a gray-box structure. in To predict the mechanism based on conservation relationships and equipment curves, To compensate for data-driven residuals for unmodeled disturbances, such as sudden changes in furnace conditions, measurement deviations, and fluctuations in local heat transfer coefficients, thereby improving accuracy without introducing unmeasurable parameters.

[0052] The current value of the heat load for each zone is identified online based on the heat load and wall temperature mechanism prediction. Short-term prediction can be performed using the autoregressive exogenous (ARX) form. in Identified using historical data through least squares / recursive least squares (auditable). By default, it can still work by retaining only the autoregressive term; subsequently, we obtain... , then calculate The above calculations rely solely on measurable temperature, pressure, flow rate, and known structural parameters. Wait, to ensure it is fully feasible to implement.

[0053] Water concentration mechanism prediction, with conductivity / tracer concentration as the core, and circulating water conductivity... Considered as a soluble tracer that can be measured online, in the effective water volume of the system When the system volume is approximately constant based on the system volume drawings / water replenishment-sewage discharge test calibration, a discrete mass conservation model can be established: in For water replenishment flow, Sewage discharge flow can be measured online; Water drift coefficient of water separator From equipment data, Measured by the cooling tower circulation flow meter; Unmodeled disturbances can be absorbed by filters; those predicted... Compared with actual / predicted We can obtain: And combined with prediction Based on whether water quality constraints are about to be triggered, updates can be made in advance. .

[0054] To avoid the control strategy becoming overly sensitive to noise due to only outputting point predictions, this invention outputs uncertainty characterizations for key prediction quantities, such as standard deviation or quantile intervals. For example, for... Give the confidence margin: in The standard deviation of the prediction error is obtained from residual statistics or quantile regression. The parameterizable safety factor is configured by the risk level; similarly, the same applies to... Provide conservative forecasts to allow for advance wastewater discharge / chemical dosing, ensuring that the forecast channel can be used for engineering safety decisions.

[0055] With online deployment and self-calibration mechanisms, the prediction model runs in industrial edge computing units or PLC coprocessors, and the trained model parameters... The system is managed and traceable by version number; it also employs a sliding window to perform online statistical updates of the residual mean and variance to adapt to seasonal changes in water quality and distribution drift caused by equipment aging. When drift indicators, such as residual mean deviation, exceed a threshold trigger, the system can switch to mechanism prediction. The conservative mode is used, and the model is prompted to be retrained to ensure stable availability throughout its lifecycle.

[0056] Furthermore, step S2, which involves obtaining risk prediction results and anomaly diagnosis conclusions including types such as vaporization margin, deposition risk, and equipment degradation, specifically includes the following steps: Step S25: Input the online collected thermal hydraulic and water quality data and rolling prediction results into the physical margin calculation module, calculate the anti-vaporization margin and anti-deposition margin of each zone and their minimum values ​​within the prediction window in real time, and output an instant risk warning signal after two-level threshold judgment, including warning, alarm and strong cooling trigger command. Step S26: Construct a residual vector from the measured values ​​and the predicted values, standardize it, and calculate the Mahalanobis distance based on the covariance matrix of historical normal data statistics. Obtain the system anomaly detection result by comparing it with the preset threshold. Then, locate the source of the anomaly based on the contribution of the residual components, such as temperature, pressure, flow rate or water quality link, and output the anomaly type and severity. Step S27: Construct characteristic residuals for key rotating equipment and pipeline components, including pump head residuals and efficiency consistency indicators, valve command-feedback-flow consistency residuals, and equivalent roughness based on pressure drop inversion. After trend analysis or change point detection, output equipment deterioration trend warnings and maintenance suggestions. Step S28: The physical margin warning, residual diagnosis and location results, and equipment degradation trend information are summarized into the anomaly comprehensive assessment unit. After multi-source information fusion and hierarchical logic judgment, the system anomaly level, partition location and suggested actions are output, and the control layer is linked to execute safety coverage, force pressure increase and flow increase, switch to standby equipment, trigger backwashing and upload maintenance work order.

[0057] Preferably, this embodiment features real-time monitoring and early warning: combining real-time collected data with prediction results, calculating the anti-vaporization and deposition margins of each zone, and promptly outputting risk warning signals through threshold judgment, thereby achieving real-time monitoring and early warning of the system's operating status. Anomaly detection and location utilize residual vectors and Mahalanobis distance to detect system anomalies and locate the source of the anomaly based on the contribution of the residual components, accurately identifying and locating anomaly points in the system. Equipment status monitoring constructs characteristic residuals for key equipment and performs trend analysis or change point detection, enabling early warning of equipment degradation trends and providing maintenance suggestions, thus achieving effective monitoring of the status of key equipment. Comprehensive evaluation and decision-making summarizes various early warning, diagnostic, and equipment degradation information, and outputs the system anomaly level, zone location, and suggested actions through multi-source information fusion and logical judgment, improving the system's comprehensive evaluation capability and decision-making efficiency. Safety control linkage: Finally, the evaluation results are linked to the control layer to execute safety coverage measures, such as forced pressure boosting and switching to backup equipment, to ensure the safe and stable operation of the system.

[0058] In summary, this embodiment combines real-time monitoring and early warning, anomaly detection and location, equipment status monitoring, comprehensive evaluation and decision-making, and security control linkage to achieve comprehensive monitoring and analysis of the system's operating status, thereby improving the system's security, stability, and reliability.

[0059] In this embodiment, the anomaly detection and early warning model aims to shift from reactive maintenance to predictive maintenance and pre-accident prevention. It constructs a multi-level anomaly detection and early warning model, forming a combined solution of physical margin index early warning, model residual consistency diagnosis, and unsupervised anomaly identification. This solution can provide millisecond-second responses to rapid hazardous states such as vaporization risks, and also provide trend-based early warnings for slow-changing anomalies such as scaling, blockage, and pump / valve deterioration, thereby improving the inherent safety and operational controllability of the system. Figure 5 : Schematic diagram of anomaly detection and early warning linkage Figure 5 The document shows physical indicators such as anti-vaporization margin and low flow rate margin, a combined early warning mechanism of predictive residual consistency diagnosis and unsupervised anomaly identification / change point detection, and the linkage process of alarm classification, mandatory security coverage, backup equipment switching and operation and maintenance command output.

[0060] Physical margin warnings for safety-critical indicators, including vaporization / overheating / deposition: Boiling Margin: Defines the vaporization margin for each partition. Boiling Margin: Defines the vaporization margin for each partition. in Measured online, Calculated online using equation [1c], calculate, Sure.

[0061] The early warning logic uses a two-level threshold: when Triggering an early warning, feedforward voltage boosting / current increase; when The alarm is triggered and the system enters high-power cooling mode, prioritizing [efficiency / improvement]. Simultaneous improvement .

[0062] To prevent sedimentation margins and mitigate the risk of low flow rates / dead zones, flow rate margins are defined for each zone. in , Online measurement or soft measurement, Given geometric parameters. If a risk of sedimentation / sludge buildup is detected, a minimum flow rate constraint and a side-filter / backwashing strategy will be triggered.

[0063] Furthermore, a prediction margin can be constructed based on the prediction results: when When the system exceeds the prediction window, it will perform pressure boosting / flow enhancement in advance, significantly reducing the probability of vaporization bursts. Consistency diagnosis based on prediction model residuals involves constructing a residual vector using the predicted output and actual measurements: in Key observations can be selected, such as All of these are available online. The residuals are standardized and the Mahalanobis distance is calculated:

[0064] The residual covariance matrix under normal operating data is obtained offline and can be recursively updated online. , Desirable When the system enters an abnormal state, it is determined that the distribution quantile threshold or empirical threshold is reached, and the source of the abnormality is located based on the contribution of the residual components, such as temperature link, pressure link, flow link or water quality link.

[0065] The advantage of this method is: threshold ,matrix All data are derived from traceable historical normal data statistics and do not rely on unmeasurable parameters; moreover, residual diagnosis is naturally compatible with the combination structure of mechanism prediction and data-driven prediction, and is suitable for unified detection of multiple anomaly types.

[0066] Identifying characteristic parameters of equipment deterioration and blockage / scaling in pumps, valves, and piping networks; diagnosing pump performance degradation; ensuring head / efficiency consistency; calculating a given speed using pump curves and similarity laws. With traffic Theoretical rise The measured head was obtained using online suction / discharge pressure measurement. Define head residual: when A consistently negative value with increasing amplitude over time indicates impeller wear, scaling, or cavitation leading to performance degradation; this is combined with the inverter's electrical power... (VFD readable) Efficiency consistency metrics can be further developed for pre-maintenance decision-making.

[0067] Valve jamming / abnormal operation diagnosis, command-feedback-flow consistency; for electric valves with valve position feedback, the definition is:

[0068] And combined with the valve standard flow equation, from Calibration curve and differential pressure calculate Constructing flow consistency residuals: in The pressure is obtained from the pressure transmitters before and after the valve. If or If the limit is continuously exceeded, it is determined that the valve is stuck, the actuator is malfunctioning, or the differential pressure measurement is abnormal, and a bypass / redundant valve switching and maintenance prompts are triggered.

[0069] Pipeline blockage / scaling trend identification and online equivalent roughness inversion: If the pressure drop along the pipe continuously increases at a given flow rate, it often corresponds to internal scaling or deposit blockage. This is based on measured pressure drop. The equivalent friction coefficient can be calculated from geometric parameters and local drag coefficients. in It can be obtained from a two-point differential pressure transmitter or converted from the pump outlet / zone pressure difference. Depend on We obtain the equivalent roughness by using the Swamee–Jain explicit relation inversion: And The time trend serves as a health indicator for the slow variable of scaling / clogging. When When the initial baseline is exceeded by a set multiple or its growth rate exceeds a threshold, bypass filtration enhancement, flushing / acid washing maintenance plans, and zone flow redistribution strategies are triggered. The above inversion relies solely on differential pressure, flow rate, and temperature for... The pipe length, pipe diameter, and local resistance coefficient can all be obtained, thus avoiding unmeasurable variables.

[0070] Unsupervised anomaly identification and change point detection, as a supplementary layer: For complex anomalies that are difficult to enumerate, such as multi-source disturbance superposition, sensor drift, and unknown operating condition combinations, this invention introduces an unsupervised anomaly identification algorithm as a supplementary layer. Isolation Forest, One-Class SVM, or Autoencoder can be used to learn low-dimensional representations of normal operating condition data, and anomaly scores can be calculated online. .when When an unknown anomaly is triggered, an alarm is automatically generated for that time period, and the data from that period is automatically archived for subsequent model retraining and knowledge base updates. Threshold It is obtained by quantile calibration of normal data and has a clear acquisition path.

[0071] At the same time, for slowly changing health indicators, such as Long-term mean values ​​are used to implement trend early warning using CUSUM or sliding window change point detection: when ( It can determine the occurrence of degradation points (using parameterizable thresholds), output maintenance suggestions in advance, and automatically reduce the system's operating load or increase redundancy margins.

[0072] Early warning output is linked with control, actions are executable, and the anomaly detection module is linked with the control setpoint layer: For rapid risks such as vaporization / overheating: Immediately increase [efforts / measures]. And increase the size of the relevant partitions. Switch to strong cooling mode if necessary; For low flow rate / deposition risk: enforce increased minimum flow constraints in relevant zones and side-filter backwashing strategies; For pump and valve deterioration / pipeline scaling trends: output maintenance work orders and executable operation derating / switching strategies (standby pumps, bypass valves, zone flow restriction, etc.) and upload the anomaly level, location information and suggested actions to DCS / host computer to achieve closed-loop operation and maintenance.

[0073] Furthermore, such as Figure 6 As shown, the process of sending the updated sewage discharge setting value and chemical dosing setting value to the heat-fluid coupling execution layer in step S4 specifically includes the following steps: Step S41: Input the inlet and outlet water temperature, pressure, and flow rate of each zone collected online into the heat load identification module, and obtain the instantaneous heat load of the zone through energy conservation calculation; combine the zone heat load with the preset target temperature rise and minimum anti-deposition flow rate, and generate the target flow rate setting value of each zone after constraint take-maximum operation; Step S42: Input the zone heat load and the measured temperature of the wall temperature thermocouple into the heat conduction inversion module, and calculate the water-side wall temperature by combining the known burial depth and thermal conductivity; input the water-side wall temperature into the saturated vapor pressure calculation module, obtain the saturated pressure through the standard physical property equation, and superimpose it with the safety margin to form the zone minimum target pressure; superimpose the zone minimum target pressure with the pressure drop and elevation difference along the pipeline, and generate the main pipeline target pressure setpoint after taking the largest value after the most unfavorable zone calculation; Step S43: Input the target pressure of the main pipe, the target flow rate of each zone, the performance curve of the pump set, and the resistance characteristics of the pipeline network into the optimization solver. With the goal of minimizing the total power of the pump station, and under the conditions of meeting the upper limit of wall temperature, temperature difference constraint, anti-vaporization constraint, and minimum flow velocity constraint, the target speed of the parallel pump set and the target opening of the regulating valve of each zone are obtained through nonlinear programming, forming the pump-valve coordinated setpoint. Step S44: The target speed of the pump set and the target opening of the valve are sent to the frequency converter and the electric actuator. The inner loop controller realizes the closed-loop tracking of the main pipe pressure and the closed-loop distribution of the zone flow, and provides real-time feedback of the actual value and compares it with the set value. When the wall temperature exceeds the limit or the vaporization margin is lower than the threshold, the strong cooling mode is immediately triggered to forcibly increase the main pipe pressure and the relevant zone flow until the risk is eliminated.

[0074] Preferably, this embodiment precisely controls heat load and flow rate. Based on real-time collected data such as temperature, pressure, and flow rate, it accurately calculates the instantaneous heat load of each zone and, combined with preset target temperature rise and minimum anti-deposition flow rate, determines the target flow rate setpoint for each zone to achieve precise matching of heat supply and demand. Precise control of wall temperature and pressure is achieved by calculating the water-side wall temperature through a heat conduction inversion module and calculating the saturated vapor pressure based on this temperature, adding a safety margin to form the minimum target pressure for each zone. Then, combined with the pressure drop along the pipeline and the elevation difference, a target pressure setpoint for the main pipeline is formed, effectively controlling the system's wall temperature and pressure to prevent overheating and vaporization. Optimized pump and valve coordination is achieved by solving nonlinear programming problems, combining pump performance curves and pipeline resistance characteristics to derive the target speed of the pump set and the target opening degree of the regulating valves in each zone. This minimizes the total power consumption of the pump station while meeting system safety and efficiency requirements. Real-time adjustment and feedback control transmit the pump and valve setpoints to the control equipment, and the inner loop controller achieves closed-loop tracking of the main pipeline pressure and zone flow rate. The system provides real-time feedback comparing actual values ​​with set values ​​and takes immediate action when anomalies are detected to ensure the safe and stable operation of the system.

[0075] In summary, this embodiment achieves precise control, optimized operation, and real-time feedback of the system, improving the energy efficiency and safety of the heating system, and ensuring stable operation and reliable heating quality.

[0076] In this embodiment, as Figure 7 Block diagram of thermal-fluid coupled variable voltage cooling control structure. Figure 7 The diagram shows the online identification of zone heat load, wall temperature inversion and saturated vapor pressure calculation, anti-vaporization target pressure and zone target flow generation, inner loop tracking control with pump frequency and valve position coordinated execution, and strong cooling / energy saving mode switching and safety interlock coverage logic.

[0077] Online identification of zoned heat load and calculation of cooling demand; the cooling system is divided into sections according to water jackets / branch circuits. One cooling zone. Instantaneous heat load of each zone. The following results were obtained from online calculations based on the law of conservation of energy: in, , The inlet and outlet water temperatures for this zone are measured by the branch temperature sensor. The zone pressure is measured by the branch pressure transmitter. For zoned volumetric flow rate, it is preferable to measure it using a branch electromagnetic / ultrasonic flow meter; if valve flow rate soft measurement is used, it is determined by the valve flow coefficient. Pressure difference across the valve Calculated using the standard valve flow equation. Measured by a differential pressure transmitter; These are water property parameters, which can be obtained from the controller's built-in IAPWS-IF97 water property function or by lookup table interpolation. Their input depends only on the measurable... and . Desirable As the average temperature of the partitioned volume, it ensures the closure of the online calculation.

[0078] To further map the heat load into executable flow setpoints, a zoned target temperature rise is introduced. The allowable value of material thermal stress and operating experience are determined, and the following conditions are met. The target traffic for a partition can be given by the following formula: In the formula For the current heat load identified by equation [1], a short-time filter can be selected to suppress noise; Minimum flow rate required to prevent sedimentation From the inner diameter of the branch pipe Direct calculation , Setting constants for engineering parameters, such as 1.5 m / s, can be configured and traced in the PLC parameter table.

[0079] Water-side vaporization suppression: wall temperature, saturation pressure, and target pressure. To prevent localized nucleus boiling from developing into film boiling, a pressure constraint needs to be established based on the saturated vapor pressure corresponding to the water-side wall temperature. Since in engineering, thermocouples are typically embedded in the copper wall, and their measured temperatures... Temperature of the water-side wall A definite thermal conductivity relationship exists. Let the depth of the thermocouple from the inner surface of the water side (i.e., the interface between the water jacket and the water) be... As specified in the installation process and drawings, it is traceable, and the thermal conductivity of copper is [value missing]. As provided in the materials handbook / inspection report, the zoned heat exchange area is... The water jacket structure dimensions determine the water-side wall temperature, which can then be derived from the online heat load: Based on this, water at temperature Saturated vapor pressure The calculation is performed using the Wagner form, which can be directly implemented in engineering. Let... ,but: in , The above coefficients are standard thermophysical constants of water, which can be embedded in the controller program to satisfy the feasibility and auditability of online calculations.

[0080] Based on this, the minimum target pressure for zoned anti-vaporization is defined as follows: Among them, safety margin Instead of using uninterpretable constants, they are composed of design margins and measurement uncertainties: In the formula The basic margin set for the process can be given by enterprise standards / HAZOP results and parameterized in the PLC; These are the standard uncertainties for pressure and temperature measurements, respectively, provided by the instrument accuracy class and calibration report. This is the coverage factor; for example, a value of 3 corresponds to approximately 99.7% confidence coverage. It can be achieved through analytical differentiation or by using numerical difference within the controller, both of which are feasible computational methods.

[0081] Further considering the pressure drop along the pipeline and in certain areas, to ensure that the formula is still satisfied in the most unfavorable zone, the target pressure of the main water supply pipe of the system is taken as: in The elevation difference between the pump outlet and the zone is determined by the site elevation. Online calculations using the Darcy-Weisbach equations:

[0082] coefficient of friction Avoid iteration by using the explicit Swamee–Jain form:

[0083] in All are determined by the pipeline network structure and material parameters. Dynamic viscosity can be determined from water property tables / functions according to... get.

[0084] Pressure-flow coordinated optimization and execution layer implementation: Under the premise of satisfying heat dissipation and anti-vaporization constraints, the controller aims to minimize the pump station's electrical power by solving for the coordinated setpoint values ​​of pump speed or frequency converter and valve opening. The instantaneous electrical power of the pump station can be expressed as:

[0085] in , The first The flow rate and speed of the parallel pumps are fed back by the frequency converter; and Determined by the pump performance curve. To ensure engineering feasibility, the pump curve uses a quadratic polynomial obtained from factory / field performance tests at rated speed. Downfit:

[0086] And based on the similarity law, it is converted to any rotational speed. :

[0087] coefficient All data originates from pump curve fitting results and can be stored in the PLC / edge controller parameter table, providing a clear data source and acquisition path.

[0088] The optimization solution needs to satisfy the following set of constraints:

[0089] The execution layer adopts a two-level structure: an outer layer for optimizing setpoints and an inner layer for fast loop closure. The outer layer calculates... With each partition The inner layer uses a frequency converter to control the pump speed to track... And by adjusting the zone flow through an electric regulating valve to track When a sudden change in furnace conditions causes a rise in wall temperature or a decrease in differential pressure margin in a certain zone, the outer layer will first increase... Prioritize raising the boiling point to suppress vaporization, and simultaneously increase the boiling point. When entering the low heat load stage, while meeting the requirements... and Under the premise of lowering This reduces pump power and saves water and energy compared to the total flow rate.

[0090] Furthermore, the process of sending the updated sewage discharge setpoints and chemical dosing setpoints to the water quality dynamic balance execution layer in step S4 specifically includes the following steps: Step S45: Input the online collected conductivity of circulating water and conductivity of makeup water into the concentration factor calculation module, and obtain the real-time concentration factor through ratio calculation; input the conductivity of circulating water into the conductivity-TDS calibration model, and calculate the total dissolved solids concentration based on the calibration coefficients of periodic test regression, forming the basic parameters of water quality status; Step S46: Input the total dissolved solids concentration, online pH value, the most unfavorable zone water side wall temperature calculated by the thermal loop, and the calcium hardness and alkalinity of the makeup water into the Langerier saturation index calculation module online or through laboratory updates. The real-time scaling tendency index is obtained by calculation using the standard Langerier formula and used as a water quality constraint criterion. Step S47: Input the real-time scaling tendency index, the upper limit of total dissolved solids concentration, the upper limit of chloride ion concentration, and the makeup water quality parameters into the target concentration factor optimization solution module. Under the condition of satisfying scaling and water quality safety constraints, search for the maximum feasible concentration factor and obtain the target concentration factor decision value. Step S48: Input the target concentration factor decision value, cooling tower evaporation rate, and drift rate into the wastewater discharge calculation module, and obtain the wastewater discharge setpoint based on the mass conservation relationship; input the wastewater discharge setpoint, drift rate, and target inhibitor concentration into the dosage calculation module, and obtain the metering pump volume flow rate setpoint based on the effective component balance, forming a closed-loop control command for the wastewater discharge valve and the dosing pump.

[0091] Preferably, this embodiment precisely controls water quality by acquiring and calculating data online. It allows for real-time monitoring and evaluation of fundamental parameters such as conductivity and total dissolved solids (TDS) of circulating water, thus accurately understanding the water quality status. Real-time water quality analysis utilizes a conductivity-TDS calibration model to convert circulating water conductivity into TDS, enabling real-time conversion and analysis of water quality parameters. Scaling tendency detection, combining parameters such as circulating water conductivity, pH value, water-side wall temperature, and makeup water calcium hardness and alkalinity, calculates a real-time scaling tendency index, providing crucial constraints for water quality control. Optimizing the target concentration factor, while ensuring water quality safety and scaling control, calculates the maximum feasible concentration factor through a target concentration factor optimization solution module, achieving efficient water resource utilization and conservation. Precise wastewater discharge and chemical dosing control, through wastewater discharge and chemical dosing calculation modules, combined with parameters such as the target concentration factor and cooling tower evaporation rate, accurately calculates wastewater discharge setpoints and chemical dosing, achieving closed-loop control of the wastewater discharge valve and chemical dosing pump, optimizing the water treatment process. The dynamic balancing process achieves dynamic balance control of water quality, ensuring stability in changing environments while reducing the use of chemicals and minimizing environmental pollution. The system's collaborative optimization combines various modules and technical features to form a cohesive system capable of dynamically adjusting based on real-time data to optimize water treatment efficiency and effectiveness.

[0092] In summary, this embodiment achieves accurate monitoring, real-time analysis, dynamic control, and optimized management of water quality, thereby improving water resource utilization efficiency and reducing environmental risks.

[0093] In this embodiment, a dynamic water quality balance anti-scaling control loop aims to control the concentration factor, scale formation tendency, and corrosion risk. It establishes a calculable criterion and action quantity mapping based on online conductivity and key water quality indicators. This loop elevates wastewater discharge and chemical dosing from empirically determined values ​​to closed-loop decisions based on online estimation and constraint solving, thereby maximizing water reuse while ensuring heat exchange and equipment lifespan. Figure 8 Schematic diagram of water quality dynamic balance scale prevention control. Figure 8 The concentration factor is shown in the figure. Online calculation and rolling evaluation of TDS and scaling propensity indices, such as LSI, and target concentration factor. The dynamic solution and the amount of sewage discharged Dosage The synergistic closed-loop relationship between the bypass filtration / backwashing and the physical field scale inhibition device.

[0094] Concentration factor and TDS online estimation: The real-time concentration factor of circulating water is calculated online using the conductivity ratio. in The conductivity measurements were obtained using conductivity meters for both circulating water and makeup water. This was used to obtain the total dissolved solids (TDS) in LSI calculations. The conductivity-TDS calibration relationship is adopted: coefficient The conductivity of the water sample – the dry residue / total ion content is obtained through calibration. It is derived from the regression of periodic test data and is a traceable parameter that can be updated regularly according to seasonal and water source changes.

[0095] A calculable index of fouling tendency, based on LSI membrane temperature correction, is expressed as the Langerile saturation index. As a criterion, the temperature input is taken as the water-side wall temperature of the most unfavorable zone to reflect the membrane temperature effect, thereby avoiding the introduction of unexplained temperature difference correction coefficients. Definition: in Measured by an online pH meter; get; ,by The calcium hardness of the sample ,by The total alkalinity is preferably obtained using an online titration analyzer; if such an analyzer is not available on-site, a soft measurement model can be established by periodically testing online conductivity, temperature, and pH. and The model coefficients are estimated by fitting historical test data, and have a clear acquisition path.

[0096] The engineering formula for saturated pH adopts the standard Langelier form and makes all coefficients explicit to avoid the inconvenience of looking up tables:

[0097] The units for all the above quantities are standardized as follows: Input in °C, All in mg / L, Inputs are calculated; all inputs can be obtained through online instruments or soft measurement / analysis calibration, meeting the requirements for formula parameter feasibility.

[0098] Quantitative decision-making and target concentration factor calculation for wastewater discharge-makeup water in cooling tower evaporation and concentration systems. No longer fixed as an empirical constant, but by satisfying The maximum feasible concentration factor for the constraints and upper limits of water quality is determined, namely: in Parameterize the upper limit of the tolerance of enterprise standards or equipment materials; The results can be obtained directly from online or laboratory tests of the replenishing water quality, thus reducing the dimensionality of the problem to a single-variable search. It can be implemented using a PLC / edge controller with a finite step size traversal, and the calculation path is clear.

[0099] get Then, based on the conservation of water volume and dissolved solids, the discharge volume... The setting can be determined by the evaporation rate. With drift volume calculate:

[0100] The evaporation rate is inferred from the heat dissipation of the cooling tower:

[0101] In the formula The flow rate of the cooling tower is measured by the flow meter. The temperature of the water entering and exiting the tower was measured by a thermometer. The latent heat of vaporization is obtained from the water property table / function based on temperature. The drift rate can be determined according to the tower type and desiccant specifications. As specified in the equipment manual, this is auditable. Accordingly, the drain valve is installed according to... Flow closed-loop control is implemented, and the water replenishment is automatically matched by the water level / flow closed loop to achieve dynamic stability of the concentration ratio.

[0102] A calculable dosing model is used to ensure that scale inhibitors / corrosion inhibitors maintain the target effective concentration in circulating water. The dosage is mg / L, calculated based on the effective ingredient content. The mass flow rate is determined by compensating for the effective ingredient carried away by the wastewater and drift. If the mass fraction of the active ingredient in the drug concentrate is As specified in the pharmaceutical technical instructions, the density of the stock solution is... The flow rate can be specified by metering or the instruction manual; therefore, the volumetric flow rate setting for the metering pump is: All of the above parameters have a clear source: Obtained from online loop calculation and flow meter closed-loop. As set by the operating standards, Based on pharmaceutical data or actual measurements, this allows for the calculation and auditing of dosages. For pH adjustments, whether adding acid or alkali, a pH-pesticide consumption coefficient calibration table can be established on-site using alkalinity titration curves of makeup water / circulating water. This converts pH deviations into dosage setpoints, avoiding the introduction of unpredictable parameters.

[0103] In addition to active regulation based on a closed-loop system of heat-fluid and water quality, this embodiment further constructs a passive anti-scaling system that does not rely on chemical agents, based on two pathways: low nucleation / low adhesion on the material surface and no retention / self-flushing in the hydraulic structure. This inherently weakens the risk of scaling at the hardware level, thereby significantly reducing the scaling rate and maintenance frequency during long-term operation and improving the system's robustness under extreme conditions. This passive anti-scaling design emphasizes that, under the same water quality and temperature conditions, surface roughness, flow retention area, air pockets / drying points, localized high film temperatures, and material electrochemical properties are key factors determining whether scale forms rapidly and adheres firmly. Therefore, quantifiable structural and material indicators are needed to systematically constrain these factors. Figure 9 Schematic diagram of anti-scaling water jacket branch pipe and key pipeline structure. Figure 9 The diagram shows key passive anti-scaling structural features, including low-adhesion lining / high-gloss material pipe section 5 (such as PTFE lining or stainless steel polished pipe), long-radius elbow and downstream transition connector 4, high-point gas collection and automatic exhaust structure 3, and physical field scale inhibition device arrangement at key inlets 6.

[0104] In critical areas with high temperature and easy scale formation (such as the water jacket outlet branch pipe, heat exchanger inlet section and local high film temperature section), it is preferable to use PTFE (polytetrafluoroethylene) lined pipe or high-gloss 316L stainless steel pipe. The material properties of low surface energy of PTFE and high corrosion resistance of stainless steel are used to reduce the adhesion of crystal nuclei and the bonding strength of scale layer, and to inhibit the formation of corrosion products as a secondary nucleation substrate.

[0105] To ensure that the material advantages can be realized in engineering, the surface quality of the pipeline is constrained by a verifiable roughness index: Let the absolute roughness of the inner wall of the pipe be... The roughness is obtained from the material manufacturer's certificate / endoscopic inspection or sampling roughness measurement, and the critical sections are required to meet the following requirements. ,For example This can be used as a design and acceptance parameter; for stainless steel sections, internal wall polishing / electropolishing processes can be used to adjust the surface roughness. Portable roughness testers can measure and serve as delivery and acceptance indicators, thereby transforming low adhesion from a descriptive goal into an actionable and acceptable engineering constraint.

[0106] In addition, to reduce galvanic corrosion and corrosion product deposition at the joints of dissimilar metals, insulating gaskets / insulating joints can be installed at the joints of dissimilar metals, and the inner wall of the weld can be ground and rounded to prevent weld beads and steps from forming micro-vortices and low-speed retention zones, thereby reducing the probability of local deposition from the structural source.

[0107] High-incidence areas of scaling typically correspond to low-velocity stagnation zones, localized backflow vortices, and reversal separation zones. Therefore, this invention adopts a design principle of full-loop turbulence and self-flushing of critical sections in the pipe network and branch structure: ensuring the Reynolds number of all critical pipe sections is maintained even under minimum load conditions. With flow rate The flow rate should not be lower than the design minimum to suppress suspended solids settling, reduce crystal nucleus residence time, and improve wall shearing and peeling capabilities. The flow velocity in each pipe section is determined by the online flow rate and pipe diameter. in It can be measured by a flow meter or by a valve. Calculated by converting with pressure difference, Determined by the dimensions on the drawing. The corresponding Reynolds number is: in The water property function is obtained by inputting online temperature / pressure and then looking up a table or using a built-in formula. In engineering, the sufficient turbulence threshold can be explicitly written as... ,For example and the minimum flow rate in the control constraints Together they form a rigid boundary against deposition and scaling.

[0108] To reduce localized separation and backflow, this invention proposes quantifiable geometric constraints on pipe fittings and connection types: all elbows are preferably long-radius elbows and meet the following requirements. To avoid strong separation and dead zones caused by right-angle connections, and at branch junctions, it is preferable to use downstream tees / oblique tees or transition components with controlled diffusion angles to reduce local energy dissipation and hot spots caused by sudden expansion / contraction. Long-radius elbows and exhaust structures, as important components of the dead-zone-free design, have been clearly proposed in the original scheme.

[0109] In critical scaling-prone sections, such as near the water jacket outlet and the straight pipe section at the heat exchanger inlet, this invention introduces engineering constraints on wall shear stress to further reduce self-flushing to calculable parameters. Wall shear stress can be expressed using the Darcy friction factor. Represented as: in It can be derived from an explicit formula that is feasible, such as Swamee–Jain. and Calculated. Can be set in engineering. As a criterion for self-flushing of critical segments, among which For calibrable parameters: the corresponding minimum stripping shear threshold can be determined through the step-by-step upflow-turbidity / particle surge response test or the side filter interception change test in the initial stage of operation, and then solidified as the design / operation parameter to achieve a verifiable, reproducible, and auditable passive scale prevention hydraulic boundary.

[0110] Air pockets can lead to localized heat exchange deterioration and water film drying, creating a vicious cycle of high film temperature, rapid scaling, and subsequent drying and consolidation. To eliminate air pockets, this invention installs an air collection tank at a high point in the system and equips it with an automatic exhaust valve to ensure continuous gas discharge during operation, preventing localized overheating and scaling. Simultaneously, the pipeline slope and high-point layout are engineered to ensure a continuous gas migration path to the exhaust device, reducing the geometric probability of gas stagnation.

[0111] Without increasing the burden on chemical reagents, this invention integrates physical field scale inhibition as an inherent hardware capability: variable frequency electromagnetic descaling devices are installed at the heat exchanger inlet and key water jacket inlet, subjecting the water to an electromagnetic field before it enters high-film-temperature, scale-prone areas. This alters the morphology and aggregation behavior of some carbonate crystals, reducing the firm deposition of hard scale on the wall and increasing the likelihood of it being discharged with the flow. The installation location, pipe diameter, and straight pipe length of this device can be adjusted according to the pipe diameter. The on-site space can be quantified and determined, such as retaining straight pipe sections with a diameter of at least several times before and after the inlet to ensure the stability of the field effect, and together with the bypass filtration unit, forming a passive scale prevention closed-loop basis of physical scale inhibition and particle removal. To ensure the continuous effectiveness of the passive scale prevention design throughout its entire life cycle, this invention reserves maintainable and verifiable interfaces at key locations, including detachable short sections, online / offline sampling ports, endoscopic inspection ports, and flushing ports, etc., so that the degradation of inner wall roughness, deposition growth, and device failure can be quantitatively verified through periodic inspections, and supports rapid replacement and local maintenance in zones, thereby transforming passive scale prevention from a one-time design into a sustainable long-term operating capability.

[0112] like Figure 10 As shown, this embodiment also provides an embodiment of an intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention system. In this embodiment, the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention system is applied to the intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention method as described in the above embodiment. The intelligent heat-fluid coupled variable pressure circulating water and full life cycle scale prevention system includes a smelting furnace cooling circuit, an intelligent variable pressure water supply pump station, an online water quality control unit, a physical scale prevention and precision filtration unit, a data acquisition and edge computing unit, a modeling and inference unit, a digital twin and operation and maintenance decision platform, and a central control system based on PLC / DCS.

[0113] The smelting furnace cooling circuit employs a water jacket monitoring array, with thermocouples embedded in key cooling water jackets, such as the slag line area and the lance area, to monitor the copper wall temperature. High-precision temperature sensors were installed in each branch pipeline. Pressure transmitter Flow meter Low-roughness piping layout: The inner walls of the main circulating water pipes and branch pipes are made of plastic-lined composite pipes or stainless steel pipes, with an absolute inner wall roughness of [missing information]. Pipeline design follows the principle of constant flow velocity to ensure that all pipe sections achieve the lowest possible Reynolds number under minimum load. Fully turbulent flow eliminates dead zones.

[0114] The intelligent variable pressure water supply pump station is equipped with multiple parallel variable frequency centrifugal pumps; the main outlet pipe is equipped with an electric regulating valve and a bypass pressure relief valve to work with the frequency converter to achieve a wide range of pressure-flow (PQ) decoupled regulation; it is equipped with pump motor current / power acquisition and pump body vibration monitoring for health diagnosis and efficiency evaluation.

[0115] The online water quality control and scale prevention unit includes online monitoring instruments, automatic dosing and sewage discharge devices, and bypass physical treatment facilities. The online monitoring instruments include a conductivity meter, circulating water ECcirc and makeup water ECfresh, pH meter, and turbidity meter. The automatic dosing and sewage discharge devices automatically add corrosion and scale inhibitors and acid according to controller instructions. The sewage discharge valve automatically opens and closes according to the concentration ratio threshold. The bypass physical treatment facilities include a bypass filtration system with a treatment capacity of 3% to 5% of the total water volume, integrating a self-cleaning filter to remove suspended solids and a high-frequency electromagnetic descaling device that changes the crystal morphology to prevent the adhesion of calcite-type scale.

[0116] The data acquisition and edge computing unit is an industrial gateway and edge computing terminal, an industrial PC / embedded system, which performs: data time alignment, outlier removal, missing value repair, and feature extraction. Feature extraction includes dT / dt, wall temperature gradient, vaporization margin, etc. It interacts bidirectionally with PLC / DCS: outputting predicted setpoints, anomaly alarms, and optimized control quantities. It supports industrial protocols such as OPCUA / ModbusTCP / Profinet, and the implementation method is not limited.

[0117] The modeling and inference unit can be deployed on in-plant servers or at the edge; it includes: heat load prediction model, water quality trend prediction model, anomaly detection model, and control strategy optimization model; it supports model version management, online calibration, and rollback strategies, and can roll back to rule / traditional control when the model confidence is insufficient.

[0118] A digital twin and operation and maintenance decision-making platform is established to create a digital twin of the circulating water system: topological hydraulic network, heat-fluid mechanism, scaling / corrosion evolution, and data-driven agent; providing 3D / 2D visualization, simulation, maintenance plan output, remote diagnosis, and reports.

[0119] Preferably, this embodiment utilizes a water jacket monitoring array and high-precision sensors to achieve precise monitoring of temperature, pressure, and flow rate in key components of the smelting furnace cooling circuit, enhancing cooling efficiency and safety. The selection of low-roughness pipe materials and a rational pipeline design reduces fluid resistance and energy loss, improving the energy efficiency of the circulating water system. Utilizing frequency conversion technology in conjunction with regulating valves enables decoupled pressure-flow regulation of the pump station, optimizing its operating efficiency and adaptability. Combined with online monitoring, automatic chemical dosing and sewage discharge, and bypass physical treatment, water quality is effectively controlled and scaling is prevented, extending equipment life and reducing maintenance costs. Data acquisition and edge computing enable real-time processing and analysis of system operating data, improving the system's predictive capabilities and response speed. Optimization using predictive models and control strategies enhances the system's intelligence level, improving its adaptability and stability. The construction of a digital twin enables system visualization, simulation, and remote maintenance, improving operational efficiency and decision-making quality.

[0120] In summary, this embodiment achieves more efficient, intelligent and environmentally friendly management of the circulating water system, improves the stability and economy of system operation, and reduces energy consumption and maintenance costs.

[0121] This embodiment uses an oxygen-enriched side-blown smelting furnace with an annual production capacity of 100,000 tons of crude copper as a typical industrial application scenario. The cooling system includes a key water jacket loop, a circulating pump station, a cooling tower, and online water quality control and bypass physical treatment devices. During implementation, approximately 30 wall temperature measuring points are arranged in key water jackets such as the slag line area and the spray gun area, and temperature, pressure, and flow measuring points are arranged in the main branches. The pump station is upgraded to a 3-in-1 standby high-pressure variable frequency pump set, with a head adjustable within the range of 0.6 MPa–1.0 MPa. A PLC control cabinet, online conductivity / pH / turbidity meters, and automatic sewage discharge, chemical dosing, and bypass filtration devices are also deployed. The system operates according to the process described in section 5.2 above: when furnace conditions fluctuate, the heat-fluid coupling loop can increase the water supply pressure and coordinate flow distribution based on the wall temperature and anti-vaporization margin; when water quality fluctuates, the water quality loop can continuously adjust the sewage discharge and chemical dosing strategies and link with the physical treatment unit based on the concentration ratio and scaling index. Operational statistics show that the average concentration ratio increased from 2.5 to 4.0, fresh water consumption decreased by 35%, and the overall power saving rate of the water pump reached 20%. After 12 months of continuous operation and inspection, there was no obvious hard scale on the inner wall of the water jacket, only a small amount of loose soft scale, and the heat exchange efficiency did not show a significant decline. Furthermore, the system achieved a rapid response to the overheating trend caused by multiple jet fluctuations, and the system can adjust the pressure to the safe range within seconds, significantly reducing the risk of vaporization and burn-through.

[0122] Compared with the prior art, the present invention has the following significant advantages: 1. Active intrinsic safety protection to prevent burn-through: By introducing the linkage control of water jacket wall temperature and pressure, this invention achieves active suppression of film boiling in an industrial circulating water system for the first time. Unlike traditional methods that rely solely on increasing flow rate for cooling, this invention raises the boiling point of the cooling water by rapidly increasing system pressure, eliminating the formation of a vapor layer on the inner wall of the water jacket from a physical and thermodynamic perspective, and greatly reducing the risk of burn-through and leakage in the smelting furnace.

[0123] 2. Significant energy and water conservation benefits: Energy saving: It breaks away from the extensive operation mode of long-term high-flow, high-pressure operation. Under low-load conditions, it significantly reduces pump station energy consumption while meeting the minimum anti-scaling flow rate. Water saving: Based on dynamic concentration ratio control of LSI index, the system can operate at the limit of water quality tolerance, maximizing the reuse of circulating water and reducing sewage discharge and fresh water replenishment.

[0124] 3. Full life cycle scale prevention, extending equipment life, integrating three lines of defense: chemical regulation or dosing, physical scale prevention or electromagnetic / flow rate control, and material optimization or PTFE lining; it not only prevents chemical scaling caused by water concentration, but also effectively prevents suspended solids or sludge deposits by controlling the flow rate v≥vmin; the comprehensive treatment solution extends the acid washing maintenance cycle of water jackets and pipelines by 2 to 3 times.

[0125] 4. Digital and intelligent management: The system provides a wealth of hydraulic and thermal data, which can be used not only for closed-loop control but also as a basis for heat balance analysis of the smelting furnace. Operators can intuitively see the heat flux density and scaling risk trend of each cooling zone, realizing the transformation from post-maintenance to predictive maintenance.

[0126] like Figure 11 As shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 1 includes a processor 11 and a memory 12 coupled to the processor 11.

[0127] The memory 12 stores program instructions for implementing the intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention method of any of the above embodiments.

[0128] The processor 11 is used to execute program instructions stored in the memory 12 to lay out the chemical pump body processing equipment.

[0129] The processor 11 can also be referred to as a CPU (Central Processing Unit). The processor 11 may be an integrated circuit chip with signal processing capabilities. The processor 11 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor.

[0130] Furthermore, Figure 12 This is a schematic diagram of the structure of a storage medium according to an embodiment of this application. The storage medium 2 in this embodiment stores program instructions 21 capable of implementing all the methods described above. These program instructions 21 can be stored in the storage medium in the form of a software product, including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0131] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0132] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

[0133] The specific embodiments of the invention have been described in detail above, but these are merely examples, and the invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this invention. Therefore, all equivalent transformations, modifications, and improvements made without departing from the spirit and principles of this invention should be included within the scope of this invention.

Claims

1. A smart thermal-fluid coupled variable pressure circulating water system and a full life-cycle scale prevention method, characterized in that, The intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention method includes: After the system is powered on, the PLC / DCS loads the basic parameters, boundary conditions and engineering data, and collects thermal, water quality and operation and maintenance data in real time to form a real-time status vector that has been verified by quality. The real-time state vector is pushed to the edge computing unit or PLC coprocessor, and rolling prediction is performed based on historical window data and exogenous disturbances. The predicted values ​​of key thermal and water quality quantities within the future prediction window are output. At the same time, rapid risk identification based on physical margin and anomaly diagnosis based on the consistency of prediction-measured residuals are performed in parallel to obtain risk prediction results and anomaly diagnosis conclusions. Risk prediction results and anomaly diagnosis conclusions are input into the integrated setter, which is rolled up in each strategy cycle according to constraint priority and safety coverage logic: determine the zonal flow setpoint, main pipe pressure setpoint and target concentration factor to take the maximum feasible value that meets the water quality boundary conditions, and calculate the sewage discharge setpoint and chemical dosing setpoint. The updated sewage discharge and chemical dosing settings are sent to the heat-fluid coupling execution layer and the water quality dynamic balance execution layer.

2. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 1, characterized in that, The process of generating a quality-verified real-time state vector includes the following steps: The system collects thermal hydraulic, water quality and actuator status data via industrial Ethernet or fieldbus, including zone temperature, pressure, flow rate, wall temperature, main pipe pressure, total flow rate, pump and valve frequency or opening feedback, as well as conductivity, pH, turbidity, sewage flow rate and makeup water flow rate. After being timestamped, the data is formed into a multi-source data stream with a unified time scale. Perform range verification, rate of change verification, and consistency verification on the data stream with unified time scale; use moving average or exponential filtering to process short-term noise; and adopt a strategy of preserving the last valid value and conservative margin for missing or unavailable data to obtain a valid dataset that has undergone quality verification and fault tolerance processing. Intermediate derived quantities are calculated online based on effective datasets, including zone temperature rise, zone heat load, water-side wall temperature and saturated vapor pressure. The physical property parameters are obtained by online table lookup or function calculation, and the structural parameters are traced from drawings or material manuals. Based on intermediate derived quantities and calibrable parameters, the final derived quantities are further calculated, including anti-vaporization margin, concentration factor, TDS value and scaling tendency index LSI, forming a real-time state vector for prediction, diagnosis and setpoint calculation.

3. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 1, characterized in that, The process of outputting the predicted values ​​of key thermal and water quality quantities within the future prediction window includes the following steps: The system collects thermal hydraulic, water quality and actuator status data through industrial Ethernet or fieldbus. After timestamp alignment and data quality verification, it constructs a historical window and status vector containing zone inlet and outlet temperatures, wall temperatures, pressures, flow rates, main pipe pressures, total flow rates, conductivity, pH, turbidity, sewage flow rates, makeup water flow rates and optional furnace-side operation quantities. The historical window and state vector are input into the gray box prediction module of the mechanism model and data-driven residual correction. The thermal and water quality baseline prediction values ​​are obtained by the mechanism calculation of heat load autoregression and water quality conservation. Then, the unmodeled disturbances are corrected by data-driven residual compensation. The key prediction quantities of the zone heat load, wall temperature, anti-vaporization pressure requirement, circulating water conductivity, concentration factor and scaling tendency index in the future prediction window are output. The uncertainty of key forecast quantities is characterized to obtain the standard deviation or quantile interval of the forecast error. Combined with the configurable safety factor, a conservative forecast result is formed and passed to the setpoint rolling update interface as a feedforward correction quantity to generate the setpoints for main pipe pressure, zone flow, target concentration factor, sewage flow and chemical dosage with safety margin. The predictive model deployed on the edge computing unit or PLC coprocessor is put into online operation, and a sliding window is used to perform real-time statistics on the mean and variance of the prediction residuals. When the drift index exceeds the set threshold, the system automatically switches to the pure mechanism conservative prediction mode and triggers the model retraining prompt.

4. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 1, characterized in that, The process of obtaining risk prediction results and anomaly diagnosis conclusions, including vaporization margin, deposition risk, and equipment degradation type, includes the following steps: The online collected thermal hydraulic and water quality data and rolling prediction results are input into the physical margin calculation module to calculate the anti-vaporization margin, anti-deposition margin and minimum value within the prediction window of each zone in real time. After two-level threshold judgment, the module outputs an instant risk warning signal, including warning, alarm and strong cooling trigger command. The measured and predicted values ​​are used to construct a residual vector. After standardization, the Mahalanobis distance is calculated based on the covariance matrix of historical normal data. The result of the system anomaly detection is obtained by comparing it with a preset threshold. Then, the source of the anomaly is located based on the contribution of the residual components, such as temperature, pressure, flow rate or water quality. The anomaly type and severity are output. Characteristic residuals are constructed for key rotating equipment and pipeline components, including pump head residuals and efficiency consistency indicators, valve command-feedback-flow consistency residuals, and equivalent roughness based on pressure drop inversion. After trend analysis or change point detection, equipment deterioration trend warnings and maintenance suggestions are output. The physical margin warning, residual diagnosis and location results, and equipment degradation trend information are summarized into the anomaly comprehensive assessment unit. After multi-source information fusion and hierarchical logic judgment, the system anomaly level, partition location and suggested actions are output, and the control layer is linked to execute safety coverage, force pressure increase and flow increase, switch to standby equipment, trigger backwashing and upload maintenance work orders.

5. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 1, characterized in that, The process of sending the updated sewage discharge setpoints and chemical dosing setpoints to the heat-fluid coupling execution layer and the water quality dynamic balance execution layer includes the following steps: The online collected inlet and outlet water temperatures, pressures, and flow rates of each zone are input into the heat load identification module. The instantaneous heat load of each zone is obtained through energy conservation calculation. The heat load of each zone is combined with the preset target temperature rise and minimum anti-deposition flow velocity. After constraint take-maximum operation, the target flow rate set value of each zone is generated. The zone heat load and the measured temperature of the wall temperature thermocouple are input into the heat conduction inversion module. The water-side wall temperature is calculated by combining the known burial depth and thermal conductivity. The water-side wall temperature is input into the saturated vapor pressure calculation module. The saturated pressure is obtained by the standard physical property equation and superimposed with the safety margin to form the minimum target pressure of the zone. The minimum target pressure of the zone is superimposed with the pressure drop and elevation difference along the pipeline network. After the maximum value is calculated for the most unfavorable zone, the target pressure set value of the main pipeline is generated. The target pressure of the main pipeline, the target flow rate of each zone, the performance curve of the pump set, and the resistance characteristics of the pipeline network are input into the optimization solver. With the goal of minimizing the total power of the pump station, under the conditions of meeting the upper limit of wall temperature, temperature difference constraint, anti-vaporization constraint, and minimum flow velocity constraint, the target speed of the parallel pump set and the target opening of the regulating valve of each zone are obtained through nonlinear programming, forming the pump-valve coordinated setpoint. The target speed of the pump set and the target opening of the valve are sent to the frequency converter and the electric actuator. The inner loop controller realizes the closed-loop tracking of the main pipe pressure and the closed-loop distribution of the zone flow, and provides real-time feedback of the actual value and compares it with the set value. When the wall temperature exceeds the limit or the vaporization margin is lower than the threshold, the strong cooling mode is triggered, which forcibly increases the main pipe pressure and the relevant zone flow until the risk is eliminated.

6. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 5, characterized in that, The process of issuing updated sewage discharge and chemical dosing settings to the water quality dynamic balance execution layer includes the following steps: The online collected conductivity of circulating water and the conductivity of makeup water are input into the concentration factor calculation module, and the real-time concentration factor is obtained through ratio calculation; the conductivity of circulating water is input into the conductivity-TDS calibration model, and the total dissolved solids concentration is obtained by converting the calibration coefficients based on periodic test regressions, forming the basic parameters of water quality status; The total dissolved solids concentration, online pH value, the most unfavorable zone water side wall temperature calculated by the thermal loop, and the calcium hardness and alkalinity of the makeup water are updated online or by laboratory testing and input into the Langier saturation index calculation module. The real-time scaling tendency index is calculated by the standard Langier formula and used as a water quality constraint criterion. The real-time scaling tendency index, the upper limit of total dissolved solids concentration, the upper limit of chloride ion concentration, and the makeup water quality parameters are input into the target concentration factor optimization solution module. Under the condition of meeting the scaling and water quality safety constraints, the maximum feasible concentration factor is searched to obtain the target concentration factor decision value. The target concentration factor decision value, along with the cooling tower evaporation rate and drift rate, are input into the wastewater discharge calculation module. Based on the mass conservation relationship, the wastewater discharge setpoint is obtained. The wastewater discharge setpoint, drift rate, and target inhibitor concentration are input into the dosage calculation module. Based on the effective component balance, the metering pump volume flow rate setpoint is obtained, forming a closed-loop control command for the wastewater discharge valve and the dosing pump.

7. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 6, characterized in that, It also features anti-scaling water jacket branch pipes and key pipe networks, low-adhesion lining / high-gloss material pipe sections, long-radius elbows and downstream transition connectors, high-point air collection and automatic exhaust structure, and passive anti-scaling structure with physical field scale inhibition devices at key inlets.

8. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 7, characterized in that, In critical areas with high temperature and easy scale formation, PTFE-lined tubes or stainless steel tubes are used. The low surface energy of PTFE tubes and the high corrosion resistance of stainless steel reduce the adhesion of crystal nuclei and the bonding strength of scale layers, and inhibit the formation of corrosion products as a secondary nucleation substrate. Insulating gaskets / insulating joints are installed at the joints of dissimilar metals, and the inner wall of the weld is ground and rounded.

9. The intelligent heat-fluid coupled variable pressure circulating water and full life-cycle scale prevention method according to claim 8, characterized in that, High-incidence areas of scaling correspond to low-velocity stagnation zones, localized backflow vortices, and reversal separation zones, imposing quantifiable geometric constraints on pipe fittings and connection types: all elbows must be long-radius elbows and meet the following requirements. Meanwhile, at the branch junctions, downstream tees, oblique tees, or transition pieces with controlled diffusion angles are used.

10. A smart heat-fluid coupled variable pressure circulating water system and a full life-cycle scale prevention system, applied to the smart heat-fluid coupled variable pressure circulating water system and the full life-cycle scale prevention method as described in any one of claims 1 to 9, characterized in that, The intelligent thermal-fluid coupled variable pressure circulating water and full life-cycle scale prevention system includes: The cooling circuit of the smelting furnace adopts a water jacket monitoring array, and thermocouples are embedded in the key cooling water jacket to monitor the copper wall temperature. High-precision temperature sensors, pressure transmitters and flow meters are installed in each branch pipeline; low roughness pipeline layout: the inner wall of the main circulating water pipeline and branch pipes is made of plastic-lined composite pipe or stainless steel pipe. The intelligent variable pressure water supply pump station is equipped with multiple parallel variable frequency centrifugal pumps; the main outlet pipe is equipped with an electric regulating valve and a bypass pressure relief valve to work with the frequency converter to achieve a wide range of pressure-flow decoupling regulation; it is equipped with pump motor current / power acquisition and pump body vibration monitoring for health diagnosis and efficiency evaluation. The online water quality control and scale prevention unit includes online monitoring instruments, automatic dosing and sewage discharge devices, and bypass physical treatment facilities. The online monitoring instruments include conductivity meters, circulating water and makeup water meters, pH meters, and turbidity meters. The automatic dosing and sewage discharge devices automatically add corrosion and scale inhibitors and acid according to controller instructions. The sewage discharge valve automatically opens and closes according to the concentration ratio threshold. The bypass physical treatment facilities include a bypass filtration system with a treatment capacity of 3% to 5% of the total water volume, integrating a self-cleaning filter to remove suspended solids and a filter to change crystal morphology. The data acquisition and edge computing unit is an industrial gateway and edge computing terminal, an industrial PC / embedded system, which performs: data time alignment, outlier removal, missing value repair, and feature extraction, including features such as dT / dt, wall temperature gradient, and vaporization margin; it also interacts bidirectionally with PLC / DCS: outputting predicted setpoints, anomaly alarms, and optimized control quantities; and it supports industrial protocols. The modeling and inference unit can be deployed on in-plant servers or at the edge; it includes: heat load prediction model, water quality trend prediction model, anomaly detection model, and control strategy optimization model; it supports model version management, online calibration, and rollback strategies, and can roll back to rule / traditional control when the model confidence is insufficient; A digital twin and operation and maintenance decision-making platform is established to create a digital twin of the circulating water system: topological hydraulic network, heat-fluid mechanism, scaling / corrosion evolution, and data-driven agent; providing 3D / 2D visualization, simulation, maintenance plan output, remote diagnosis, and reports.

Images