Condenser system based on self-adaptive adjustment control

Abstract

The invention relates to the technical field of control adjustment, in particular to a condenser system based on adaptive adjustment control, which comprises a data acquisition module, a data processing module, a game decision control module, a field collaborative execution module and a dynamic knowledge base module. The data acquisition module is used for acquiring tube bundle surface temperature gradient distribution, a steam phase change rate image, condensed water ion migration flux and tube wall crystal nucleus adhesion density; the game decision control module is used for converting genotypes into space equilibrium, transient response and long-acting protection control instructions and outputting dynamic control parameters including valve opening and voltage strategies; the field cooperative execution module responds to the parameters to regulate and control the heat absorption strength of the phase change material and the wettability state of the pipe wall; the dynamic knowledge base module associates the thermal stress parameters with the equipment life mapping relation and feeds variation intensity constraints back to the data processing module. The problems of condensation water degradation and equipment reliability reduction caused by thermal load imbalance in the tube bundle switching process are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of control adjustment technology, in particular to a condenser system based on adaptive adjustment control. BACKGROUND

[0002] In power plants, the condenser system cools the low-pressure high-temperature steam discharged by the steam turbine through a heat exchange process, so that it condenses into liquid water; after the cooling water flows through the pipe bundle to absorb the steam heat, the steam undergoes phase change, the steam pressure is reduced, and the required vacuum degree is established to provide qualified feed water for the boiler, which not only maintains the high-efficiency operation of the steam turbine unit, but also optimizes the resource recycling of the thermal cycle, avoiding the loss of heat and water resources caused by direct steam discharge.

[0003] In the partition control scene of dynamic pipe bundle switching of the condenser system, the main technical pain points are the response lag of partition actuators and the mismatch of heat load distribution under transient conditions. Specifically, when the control logic of pipe bundle partition deactivation or activation is triggered by the fluctuation of the steam turbine power generation load, the action delay of the mechanical actuators (such as valves or baffles) superimposed with the dynamic change of the steam flow will cause an enlarged local heat exchange temperature difference in the switching transition period, causing non-uniform thermal stress between pipe bundle units and deterioration of the condensate water quality. For example, when the load drops suddenly, the control system closes a pipe bundle partition, and if the actual flow rate decays lags behind the set command due to the time lag of the actuator, the residual steam continues to condense in the area to be deactivated, forming local supercooling, which will accelerate the crystallization and precipitation of salt impurities when the supercooling degree exceeds the design margin, causing an increased risk of pipe fouling and exceeding the conductivity of the condensate water. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides a condenser system based on adaptive adjustment control, which solves the problem of condensate water deterioration and equipment reliability decline caused by partition heat load imbalance in the dynamic pipe bundle switching process.

[0005] To solve the above technical problems, the specific technical solutions of the present application are as follows:

[0006] The present application provides a condenser system based on adaptive adjustment control, which comprises:

[0007] A data acquisition module acquires the temperature gradient distribution of the condenser pipe bundle surface, the dynamic image of the steam phase change rate, the ion transmembrane flux of the condensate water, and the crystal nucleus adhesion density of the pipe wall;

[0008] A data processing module inputs the collected data of the data acquisition module, constructs the pipe bundle partition topology structure, the cooling water distribution scheme, and the genotype representation of the supercooling inhibition factor, takes the temperature difference mean square error, the fouling risk index, and the thermal stress distribution entropy as the thermodynamic constraint parameters, and performs multi-objective evolutionary calculation to generate an optimized genotype;

[0009] A game decision control module receives the optimized genotype of the data processing module, converts into a non-dominated solution set of space balance control, transient response control and long-term protection control, and outputs dynamic control parameters including valve opening degree instructions and voltage application strategies through an arbitration mechanism;

[0010] A field synergy execution module responds to the dynamic control parameters, triggers phase change material heat absorption intensity adjustment and pipe wall interface wettability state conversion;

[0011] A dynamic knowledge base module associates the pipe wall crystal nucleus attachment density and thermal stress fluctuation amplitude of the data acquisition module, and generates variation intensity constraint parameters for equipment life attenuation prediction;

[0012] The dynamic knowledge base module feeds back the variation intensity constraint parameters to the data processing module, and the electric heating wire power data and coating contact angle state data of the field synergy execution module are simultaneously input into the dynamic knowledge base module to update the preset life prediction model.

[0013] Further, the condenser system based on adaptive adjustment control provided by the application comprises:

[0014] A flexible thermocouple array for temperature field monitoring covers the surface of the tube bundle at a spacing of ≤10 cm, and outputs temperature gradient distribution data;

[0015] A laser-induced fluorescence probe for phase change process capture is distributed along the steam flow direction, and outputs local phase change rate dynamic images;

[0016] A microfluidic electrochemical sensor for impurity migration analysis is arranged at the outlet of the condensed water, and outputs calcium and magnesium ion migration flux;

[0017] An ultrasonic wave diffractometer for crystallization state detection outputs quantitative values of pipe wall crystal nucleus attachment density;

[0018] The temperature gradient distribution data, local phase change rate dynamic images, local phase change rate dynamic images and quantitative values of pipe wall crystal nucleus attachment density are synchronously transmitted to the data processing module through an industrial bus.

[0019] Further, the condenser system based on adaptive adjustment control provided by the application comprises:

[0020] The data processing module receives the pipe wall crystal nucleus attachment density and temperature gradient distribution data of the data acquisition module, and constructs a genotype representation including the following: a three-dimensional binary topological matrix of tube bundle partition, a cooling water flow rate adjustment gene sequence and a phase change material activation temperature threshold floating point gene;

[0021] When the local temperature difference monitoring value extracted from the temperature gradient distribution data exceeds the preset threshold, the elite retention crossover and directed mutation calculation are triggered, and the optimized genotype is output to the game decision control module.

[0022] Furthermore, in the condenser system based on adaptive regulation control according to the present invention, the multi-objective fitness function includes:

[0023] Convert the temperature gradient distribution data output by the data acquisition module into the mean square error parameter of the temperature difference on the tube bundle surface;

[0024] The scaling risk index parameter is generated based on the calcium and magnesium ion migration flux detected by the microfluidic electrochemical sensor;

[0025] The entropy parameter of thermal stress distribution is calculated based on the crystal nucleus attachment density quantified by ultrasonic diffractometer;

[0026] The output gene selection pressure parameter drives the elite retention crossover calculation and complies with: condensate water conductivity ≤ 0.5μS / cm, tube bundle deformation ≤ 0.1mm, and the parameters are updated in real time through the industrial bus.

[0027] Furthermore, in the condenser system based on adaptive regulation control according to the present invention, in response to a trigger condition in which a local temperature difference monitoring value extracted from the temperature gradient distribution data exceeds a preset threshold, the directional variation calculation includes:

[0028] The current thermal field state data in the data processing module is called to construct the initial population, the optimal configuration of adjacent partitions stored in the elite gene library is read for gene exchange, and random temperature perturbations within the range of ±5°C are applied to the gene fragments in the supercooled area. The output perturbed gene sequence overwrites the original cooling water flow rate regulation gene sequence and the phase change material activation temperature threshold floating-point gene.

[0029] Furthermore, in the condenser system based on adaptive regulation control according to the present invention, the game decision control module is configured as follows:

[0030] The optimized genotype output by the input data processing module is converted into the following control dimension instructions through the Pareto solution generator:

[0031] Spatial balance control instructions are used to control the shape memory alloy variable stiffness network;

[0032] Transient response control instructions are used for piezoelectric ceramic microfluidic array control;

[0033] Long-term protection control instructions are used for intelligent wettability coating regulation;

[0034] After receiving the valve response delay parameter feedback from the field cooperative execution module, the Shapley value arbiter calculates the marginal contribution of each execution unit:

[0035] Φvalve = ΔTreduction / τresponse;

[0036] Φcoating = (1 - Kcrystallization) x tduration;

[0037] Output the dynamic control parameter set to the execution interface, including valve opening degree code, voltage application coordinates and duration instruction.

[0038] Further, the condenser system based on adaptive adjustment control disclosed by the application is characterized in that the field coordination execution module is configured to:

[0039] Receive the dynamic control parameter set output by the game decision control module, and execute:

[0040] The phase change array regulation layer analyzes the valve opening degree code, positions the microencapsulated phase change material at the pipe bundle support frame, and triggers the stage-by-stage heat absorption of the electric heating wire according to the real-time feedback of the local supercooling state of the data acquisition module.

[0041] The intelligent interface protection layer analyzes the voltage application coordinates and duration instruction, applies the voltage-responsive polymer coating to the target area of the pipe wall, and dynamically adjusts the contact angle to realize the gradient distribution of wettability.

[0042] The actual power curve of the electric heating wire, the contact angle change curve of the coating and the valve action delay parameter are fed back to the dynamic knowledge base module.

[0043] Further, the condenser system based on adaptive adjustment control disclosed by the application is characterized in that in response to the voltage application coordinates and duration instruction in the dynamic control parameter set, the dynamic regulation of the intelligent interface protection layer includes:

[0044] The supercooling region is positioned by analyzing the voltage application coordinates, and +5V positive voltage is applied to make the contact angle <30° to enhance hydrophilicity.

[0045] The normal region not covered by the coordinates is maintained in a hydrophobic state with a contact angle >90°;

[0046] The contact angle change curve is generated and fed back to the dynamic knowledge base module to form a non-uniform wetting interface that blocks the attachment of impurities;

[0047] The coverage range of the hydrophilicity enhanced region coincides with the spatial coordinates of the supercooling region gene fragment.

[0048] Further, the condenser system based on adaptive adjustment control disclosed by the application is characterized in that the dynamic knowledge base module is configured to:

[0049] Receive the feedback of the actual power curve of the electric heating wire and the contact angle change curve of the coating, and establish:

[0050] mapping relationship between the power fluctuation of the electric heating wire and the nucleation growth rate of the tube wall;

[0051] correlation between the contact angle change gradient and the fluctuation amplitude of thermal stress;

[0052] When the life attenuation probability output by the life prediction model exceeds 15%, a variation intensity constraint parameter is generated and transmitted to the data processing module of claim 3 through an industrial bus.

[0053] Further, the condenser system based on adaptive adjustment control comprises the following:

[0054] The real-time crystal nucleus attachment density data of the integrated ultrasonic wave diffractometer and the electric heating wire power fluctuation data are convolved through a long short-term memory network:

[0055] Input the current working condition gene code and the historical failure cases;

[0056] Fusion of thermal stress fluctuation amplitude change gradient parameters;

[0057] Output the tube bundle residual life probability distribution to the variation intensity constraint generator;

[0058] When the tube bundle strain data detected by the flexible thermocouple array reaches a threshold, the elite reservation crossover mechanism is triggered to reconstruct the initial population.

[0059] Advantages of the present application

[0060] The present application captures the thermal state and impurity migration state of the tube bundle in real time through a multi-dimensional perception module, establishes a thermal load distribution panorama; the data processing module encodes physical parameters into evolvable genotypes, generates pre-control strategies through multi-objective optimization, and reconstructs the cooling water distribution scheme and phase change material activation threshold before the tube bundle switching instruction is triggered; the game decision control module dynamically schedules the priority of the execution unit according to the Shapley value arbitration mechanism, the spatial balance instruction compensates for the structural deformation, the transient response instruction suppresses the flow lag, and the long-acting protection instruction blocks the crystallization path; the field cooperative execution module completes the phase change material heat absorption and interface wettability regulation in milliseconds; the dynamic knowledge base module associates the execution effect with the equipment degradation model, and feedbacks and corrects the genetic calculation strategy through life prediction. The closed-loop control system balances the partition thermal load of the tube bundle from the thermodynamic source, significantly reduces the risk of condensate water quality degradation caused by local supercooling, and delays the thermal fatigue damage by dynamically optimizing the tube bundle topology structure. BRIEF DESCRIPTION OF DRAWINGS

[0061] In order to more clearly illustrate the technical solutions of the present application, the following will briefly introduce the drawings needed in the embodiments. Obviously, other drawings can also be obtained by those skilled in the art without creative labor.

[0062] Figure 1 The system architecture diagram of the condenser system based on adaptive adjustment control is provided for the embodiments of the present application. DETAILED DESCRIPTION

[0063] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions of the present application will be described below in detail with the embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present application. The technical solutions provided by the embodiments of the present application will be described in detail below with the drawings. In order to better understand the objects of the present application, the present application will be further described in detail.

[0064] Please refer to Figure 1 The present application provides a condenser system based on adaptive adjustment control, comprising:

[0065] A data acquisition module acquires the temperature gradient distribution of the condenser tube bundle surface, the dynamic image of the steam phase change rate, the condensate ion migration flux and the tube wall crystal nucleus attachment density.

[0066] A data processing module inputs the collected data of the data acquisition module, constructs the tube bundle partition topology structure, the cooling water distribution scheme and the genotype representation of the supercooling inhibition factor, takes the temperature difference mean square error, the fouling risk index and the thermal stress distribution entropy as the thermodynamic constraint parameters, and executes the multi-objective evolution calculation to generate the optimized genotype.

[0067] A game decision control module receives the optimized genotype of the data processing module, converts into the non-dominated solution set of the spatial equilibrium control, the transient response control and the long-acting protection control, and outputs the dynamic control parameters including the valve opening degree instruction and the voltage application strategy through the arbitration mechanism.

[0068] A field synergy execution module responds to the dynamic control parameters, triggers the phase change material heat absorption intensity adjustment and the tube wall interface wettability state conversion.

[0069] A dynamic knowledge base module associates the tube wall crystal nucleus attachment density of the data acquisition module with the thermal stress fluctuation amplitude, and generates the variation intensity constraint parameter of the equipment life attenuation prediction.

[0070] The dynamic knowledge base module feeds back the variation intensity constraint parameter to the data processing module, and the electric heating wire power data and the coating contact angle state data of the field synergy execution module are synchronously input into the dynamic knowledge base module to update the preset life prediction model.

[0071] The data acquisition module collects the key operating parameters of the condenser through a stereoscopic sensing network. A flexible thermocouple array uniformly covers the surface of the tube bundle at a spacing of not more than 10 cm, and outputs real-time temperature gradient distribution data; a laser-induced fluorescence probe is distributed along the steam flow path to capture local phase change rate dynamic images; a microfluidic electrochemical sensor is deployed at the condensate outlet to continuously detect calcium and magnesium ion flux; an ultrasonic wave diffraction instrument is installed at key areas of the tube wall to quantify the crystal nucleus attachment density. Each sensing unit transmits raw monitoring data synchronously through an industrial bus.

[0072] The data processing module performs genetic modeling after receiving multi-source sensing data. The crystal nucleus attachment density of the tube wall and the temperature gradient distribution data are input into the genetic code engine to construct a complex genotype representation including a three-dimensional binary topology matrix of tube bundle partition, a cooling water flow rate adjustment gene sequence, and a phase change material activation temperature threshold floating point gene. When the local temperature difference monitoring value extracted from the temperature gradient data exceeds the preset threshold, a multi-objective evolutionary calculation process is triggered: the temperature gradient data is converted into a tube bundle surface temperature difference mean square error parameter, the ion flux generates a fouling risk index parameter, the crystal nucleus attachment density calculates a thermal stress distribution entropy parameter, and the elite reservation crossover and directional variation calculation is driven to output the optimized genotype that meets the condensate conductivity threshold and the tube bundle deformation threshold.

[0073] The game decision control module analyzes the optimized genotype to generate execution strategies. The Pareto solution set generator converts the genotype into space equilibrium control, transient response control, and long-term protection control instructions, corresponding to shape memory alloy variable stiffness network regulation, piezoelectric ceramic micro-jet array regulation, and intelligent wetting coating regulation, respectively. The Shapley value arbitrator calculates the marginal contribution of each execution unit based on the valve response delay parameter feedback from the field synergy execution module, and outputs a dynamic control parameter set including valve opening code, voltage application coordinates, and duration instructions.

[0074] The field synergy execution module executes physical regulation in response to dynamic control parameters. The phase change array regulation layer analyzes the valve opening code, positions the microencapsulated phase change material at the action point of the tube bundle support frame, and triggers the graded heat absorption of the heating wire according to real-time temperature monitoring data; the intelligent interface protection layer analyzes the voltage application coordinates and duration instructions, applies a voltage-responsive polymer coating to the target area of the tube wall, and dynamically generates a hydrophilic-hydrophobic gradient distribution interface. The heating wire power curve, coating contact angle change curve, and valve action delay parameter generated during the execution process are fed back to the backend module in real time.

[0075] The dynamic knowledge base module constructs a device life prediction model. The mapping relationship between the power fluctuation of the electric heating wire and the nucleation growth rate of the tube wall is associated, and the correlation between the contact angle change gradient and the fluctuation amplitude of the thermal stress is associated. When the life attenuation probability output by the convolution long short-term memory network exceeds the set threshold, the variation intensity constraint parameter is generated and transmitted to the data processing module; when the strain data detected by the flexible thermocouple array reaches the critical value, the genetic model initial population reconstruction mechanism is triggered.

[0076] Each module forms a closed-loop control system: multi-dimensional perception data drives genetic evolution calculation, optimization decision instruction guides physical execution device action, and execution effect data feedback to knowledge base generates optimization constraints, and finally the genetic model calculation strategy is iteratively corrected through constraint parameters. This system realizes the dynamic balance control of the tube bundle heat load, and from the source, inhibits the condensate deterioration and equipment reliability decline caused by local subcooling.

[0077] The sensing units of the data acquisition module are deployed according to the function positioning. The flexible thermocouple array is uniformly arranged on the surface of the tube bundle with a spacing of not more than ten centimeters, generating temperature gradient distribution data; the laser-induced fluorescence probe is distributed along the steam flow axis direction, outputting phase change rate dynamic image sequence; the microfluidic electrochemical sensor is integrated in the inner wall of the condensate outlet pipeline, continuously capturing calcium and magnesium ion migration flux; the ultrasonic wave diffraction instrument is installed at the key nodes of the tube bundle support structure, quantifying the crystal nucleus adhesion density value. All sensing units realize millisecond-level time synchronization through industrial bus protocol, and package the original monitoring data for transmission to the data processing module.

[0078] After receiving the synchronously transmitted multi-source data, the data processing module performs gene coding. The tube wall crystal nucleus adhesion density value and the temperature gradient distribution data are input into the genotype construction engine to generate a composite coding structure including a tube bundle partition three-dimensional binary topology matrix, a cooling water flow rate adjustment gene sequence, and a phase change material activation temperature threshold floating point gene. When the local temperature difference monitoring value generated by the difference calculation of the temperature gradient data exceeds the preset threshold, the elite reservation crossover operation is triggered: the initial population is constructed by calling the current thermal field state snapshot, the gene fragments are exchanged with the optimal configuration of the adjacent partition in the elite gene library, and the basic optimization scheme is formed.

[0079] The multi-objective fitness function establishes a parameter conversion channel. The temperature gradient distribution data is converted into a tube bundle surface temperature difference mean square error parameter through spatial interpolation calculation; the ion migration flux collected by the microfluidic electrochemical sensor is converted into a fouling risk index parameter through concentration integration; the crystal nucleus adhesion density value output by the ultrasonic wave diffraction instrument is converted into a thermal stress distribution entropy parameter through stress field simulation calculation. The above parameters are updated in real time through the industrial bus, which together drive the elite reservation crossover calculation process, and output the gene selection pressure parameter, which is forced to meet the upper limit of the condensate conductivity and the deformation tolerance constraint of the tube bundle.

[0080] The directional variation mechanism is activated on the basis of elitist crossover. For the supercooled region gene fragments in the initial population, random temperature perturbation operations are applied within a range of ±5 degrees Celsius. The perturbation operation generates random seeds using the Monte Carlo method, modifies the flow rate distribution coefficient of the cooling water flow rate adjustment gene sequence, and synchronously adjusts the trigger threshold of the phase change material activation temperature threshold floating point gene. The output perturbed gene sequence covers the original encoding structure, which is verified by the data verification module and then transmitted to the game decision control module.

[0081] The game decision control module performs multi-objective decision analysis. The Pareto solution set generator decomposes the optimized genotype into spatial equilibrium control instructions, transient response control instructions, and long-term protection control instructions, which are mapped to the spatial deformation regulation of the shape memory alloy variable stiffness net, the transient flow regulation of the piezoelectric ceramic micro-jet array, and the interface property regulation of the intelligent wetting coating, respectively. The Shapley value arbitrator receives the valve response delay time parameters fed back by the field synergy execution module, calculates the marginal contribution weight of the execution unit, and generates a dynamic control parameter set including valve opening binary encoding, voltage application spatial coordinates, and action duration.

[0082] The field synergy execution module implements physical layer regulation. The phase change array regulation layer analyzes the valve opening binary encoding, locates the microencapsulated phase change material action points in the pipe bundle support frame coordinate grid, and triggers the heating power of the heating wire according to real-time temperature monitoring data. The intelligent interface protection layer analyzes the voltage application spatial coordinates, deposits a voltage-responsive polymer coating in the target area of the pipe wall, and dynamically adjusts the contact angle by changing the surface charge density. The execution process records the heating wire power fluctuation curve, the coating contact angle change trajectory, and the valve action delay parameter in real time, and feeds them back to the dynamic knowledge base module through the data interface.

[0083] The intelligent interface protection layer establishes a hydrophilic-hydrophobic gradient control. In response to the voltage application spatial coordinate instructions, a positive five-volt voltage is applied to the supercooled region located by the coordinates, causing the surface energy of the polymer coating to increase to a hydrophilic state with a contact angle less than thirty degrees; the non-coordinate covered area maintains the original hydrophobic characteristics of the coating, with a contact angle greater than ninety degrees. The generated contact angle change curve includes time-space two-dimensional distribution data, which establishes a grid mapping relationship with the spatial coordinates of the supercooled region gene fragments in claim 5, forming a dynamically updated non-uniform wetting interface database.

[0084] The dynamic knowledge base module constructs a device degradation model. The received heating wire power fluctuation curve and contact angle change trajectory data establish an exponential correlation model between the heating wire power fluctuation amplitude and the pipe wall crystal nucleus growth rate; a linear regression relationship between the contact angle change gradient and the thermal stress fluctuation amplitude is also constructed. When the remaining life probability distribution output by the life prediction model shows that the decay rate has broken through the set threshold, a gene mutation intensity constraint parameter is generated and transmitted to the evolutionary computation engine of the data processing module through the industrial bus.

[0085] The life prediction model performs multi-source data fusion. The convolutional long short-term memory network simultaneously inputs the crystal nucleus adhesion density data collected by the ultrasonic diffractometer in real time, the power fluctuation spectrum data of the electric heating wire, and the historical failure case feature vector. The network hidden layer fuses the change gradient parameters of the thermal stress fluctuation amplitude, and outputs the residual life probability distribution matrix of the tube bundle. When the tube bundle strain data detected by the flexible thermocouple array exceeds the fatigue threshold after Fourier transform, the elite reservation crossover mechanism is triggered to reconstruct the initial population, and the tube bundle partition three-dimensional binary topology matrix basic configuration is reloaded.

[0086] The physical parameters obtained by the sensing layer are encoded into evolvable genotypes by the data processing module, and execution instructions are generated through multi-objective decision making; the field coordination execution module converts the instructions into physical regulation actions, and the running state parameters thereof are fed back to the knowledge base; the degradation model established by the knowledge base generates an optimization constraint callback correction gene calculation strategy, forming a closed-loop control system. This system dynamically adjusts the tube bundle heat load distribution, essentially inhibiting the condensate water quality degradation and equipment life attenuation problems caused by local supercooling.

[0087] In the dynamic tube bundle switching scene of the power plant condenser, the system collects operating parameters through a multi-level sensing network. The flexible thermocouple array covers the tube bundle surface with a spacing of not more than ten centimeters, and outputs temperature gradient distribution data; the laser-induced fluorescence probe is distributed along the steam flow direction, and captures local phase change rate dynamic images; the microfluidic electrochemical sensor at the condensate water outlet detects calcium and magnesium ion flux; the ultrasonic diffractometer quantifies the tube wall crystal nucleus adhesion density in real time. Industrial bus protocol synchronously transmits each sensing data to the processing unit.

[0088] The data processing module receives multi-source sensing data and constructs a genotype representation. The tube wall crystal nucleus adhesion density and temperature gradient distribution data generate a tube bundle partition three-dimensional binary topology matrix, a cooling water flow rate adjustment gene sequence, and a phase change material activation temperature threshold floating point gene. When the local temperature difference monitoring value exceeds the preset threshold, the elite reservation crossover operation is triggered: the initial population is constructed by calling the current thermal field state data, and the optimal configuration of the adjacent partition is read from the elite gene library for exchange. The multi-objective fitness function converts the temperature gradient data into a temperature difference mean square error parameter, the ion flux generates a fouling risk index parameter, the crystal nucleus adhesion density calculates a thermal stress distribution entropy parameter, and outputs a gene selection pressure parameter while satisfying the condensate water conductivity threshold and the tube bundle deformation threshold constraints.

[0089] The game decision control module analyzes and optimizes the genotype to generate execution instructions. The Pareto solution set generator converts the genotype into spatial equilibrium control instructions to drive the shape memory alloy variable stiffness net regulation, transient response control instructions to drive the piezoelectric micro jet array regulation, and long-term protection control instructions to drive the intelligent wetting coating regulation. The Shapley value arbitrator calculates the marginal contribution of the execution unit based on the valve response delay parameter, and outputs a dynamic control parameter set including valve opening coding, voltage application coordinates, and duration.

[0090] The field coordination execution module responds to the instruction to execute physical regulation. The phase change array regulation layer analyzes the valve opening coding, positions the microencapsulated phase change material at the action point of the tube bundle support frame, and triggers the graded heat absorption of the electric heating wire according to real-time temperature data; the intelligent interface protection layer analyzes the voltage application coordinates, deposits a voltage-responsive polymer coating on the target area of the tube wall, and dynamically adjusts the contact angle to form a hydrophilic-hydrophobic gradient interface. Applying a positive five-volt voltage to the supercooled region achieves a hydrophilic state with a contact angle less than thirty degrees, and the non-covered region maintains a hydrophobic characteristic with a contact angle greater than ninety degrees. The execution process feedbacks the electric heating wire power curve, the contact angle change trajectory, and the valve delay parameter.

[0091] The dynamic knowledge base module establishes a device degradation model. The electric heating wire power fluctuation curve is related to the tube wall crystal nucleus growth rate, and the contact angle change trajectory is related to the thermal stress fluctuation amplitude. The convolutional long short-term memory network fuses crystal nucleus adhesion density data, electric heating wire power spectrum, and historical failure cases to output the tube bundle remaining life probability distribution. When the life decay probability exceeds the set threshold, the variation intensity constraint parameter is generated to return to the genetic model; when the tube bundle strain data detected by the flexible thermocouple array exceeds the fatigue threshold, the elite reservation crossover mechanism is triggered to reconstruct the initial population.

[0092] The system predicts the thermal load imbalance risk in advance through real-time sensing and genetic modeling. The multi-dimensional sensing module collects the tube bundle surface temperature gradient distribution and steam phase change rate dynamic image, combines the condensate ion migration flux and tube wall crystal nucleus adhesion density data, and constructs a thermodynamic state panorama. The data processing module encodes the physical parameters into the tube bundle partition topology genotype, generates an optimized control strategy through multi-objective evolutionary calculation, and completes the thermal load balance pre-configuration before the tube bundle switching instruction is issued.

[0093] The game decision control module converts the genotype into space-time dual-dimensional control instructions. The spatial equilibrium instruction dynamically adjusts the tube bundle support structure through the shape memory alloy variable stiffness net, the transient response instruction drives the piezoelectric micro jet array to compensate for the flow lag, and the long-term protection instruction controls the intelligent coating to form a hydrophilic-hydrophobic gradient interface. When a local temperature difference expansion trend is detected, the directional variation calculation immediately corrects the cooling water distribution scheme, and the Shapley value arbitration mechanism preferentially schedules the execution unit with a large response delay to complete the thermal load rebalance in milliseconds.

[0094] The physical regulation effect of the field synergy execution module is fed back to the dynamic knowledge base in real time. The power curve of the heating wire reflects the heat absorption efficiency of the phase change material, and the change trajectory of the coating contact angle represents the interface protection state. These parameters establish a mapping model with the pipe wall crystal nucleus growth rate and thermal stress fluctuation. When the prediction model shows a life attenuation risk, the system automatically reconstructs the initial population of the genetic algorithm, optimizes the gene coding of the pipe bundle topology structure, and suppresses the reliability decline trend of the equipment from the source. The closed-loop system stabilizes the condensate water conductivity within the design threshold, and reduces the risk of crystallization and scaling.

Claims

1. The condenser system based on adaptive regulation control is characterized by: include: Data acquisition module, which obtains the temperature gradient distribution on the condenser tube bundle surface, dynamic images of steam phase change rate, condensate ion migration flux and tube wall crystal nucleus attachment density; a data processing module, which inputs the collected data from the data acquisition module, constructs a genotype representation of the tube bundle partition topology, cooling water distribution scheme, and supercooling inhibition factor, uses the mean square error of the temperature difference, the scaling risk index, and the thermal stress distribution entropy as thermodynamic constraint parameters, and performs multi-objective evolutionary calculation to generate an optimized genotype; A game decision control module receives the optimized genotype of the data processing module, converts it into a non-dominated solution set of spatial equilibrium control, transient response control and long-term protection control, and outputs dynamic control parameters including valve opening instructions and voltage application strategy through an arbitration mechanism; A field synergy execution module, in response to the dynamic control parameters, triggers the adjustment of the heat absorption intensity of the phase change material and the wettability state conversion of the tube wall interface; A dynamic knowledge base module is configured to associate the tube wall crystal nucleus attachment density and the thermal stress fluctuation amplitude obtained from the data acquisition module to generate a variation strength constraint parameter for equipment life attenuation prediction; The dynamic knowledge base module feeds back the variation intensity constraint parameters to the data processing module, and at the same time, the heating wire power data and coating contact angle state data of the field synergy execution module are synchronously input into the dynamic knowledge base module to update the preset life prediction model.

2. The condenser system based on adaptive regulation control according to claim 1, characterized in that: The data acquisition module includes: A flexible thermocouple array for temperature field monitoring covers the tube bundle surface with a spacing of ≤10cm and outputs temperature gradient distribution data; The laser-induced fluorescence probes that capture the phase change process are distributed along the steam flow direction and output dynamic images of the local phase change rate; A microfluidic electrochemical sensor for impurity migration analysis is located at the condensate outlet and outputs calcium and magnesium ion migration flux; An ultrasonic diffractometer performs crystallization state detection and outputs a quantitative value of the crystal nucleus attachment density on the tube wall; The temperature gradient distribution data, the dynamic image of the local phase change rate, the dynamic image of the local phase change rate and the quantitative value of the tube wall crystal nucleus attachment density are synchronously transmitted to the data processing module through the industrial bus.

3. The condenser system based on adaptive regulation control according to claim 2, characterized in that: The data processing module: Receive the tube wall crystal nucleus attachment density and temperature gradient distribution data from the data acquisition module, and construct the following genotype representations: a three-dimensional binary topological matrix of the tube bundle partition, a cooling water flow rate regulation gene sequence, and a phase change material activation temperature threshold floating point gene; When the local temperature difference monitoring value extracted from the temperature gradient distribution data exceeds the preset threshold, the elite retention crossover and directed mutation calculation are triggered, and the optimized genotype is output to the game decision control module.

4. The condenser system based on adaptive regulation control according to claim 3, characterized in that: The multi-objective fitness function includes: Convert the temperature gradient distribution data output by the data acquisition module into the mean square error parameter of the temperature difference on the tube bundle surface; The scaling risk index parameter is generated based on the calcium and magnesium ion migration flux detected by the microfluidic electrochemical sensor; The entropy parameter of thermal stress distribution is calculated based on the crystal nucleus attachment density quantified by ultrasonic diffractometer; The output gene selection pressure parameter drives the elite retention crossover calculation and complies with: condensate water conductivity ≤ 0.5μS / cm, tube bundle deformation ≤ 0.1mm, and the parameters are updated in real time through the industrial bus.

5. The condenser system based on adaptive regulation control according to claim 4, characterized in that: In response to a trigger condition in which a local temperature difference monitoring value extracted from the temperature gradient distribution data exceeds a preset threshold, the directed variation calculation includes: The current thermal field state data in the data processing module is called to construct the initial population, the optimal configuration of adjacent partitions stored in the elite gene library is read for gene exchange, and random temperature perturbations within the range of ±5°C are applied to the gene fragments in the supercooled area. The output perturbed gene sequence overwrites the original cooling water flow rate regulation gene sequence and the phase change material activation temperature threshold floating-point gene.

6. The condenser system based on adaptive regulation control according to claim 5, characterized in that: The game decision control module is configured to: The optimized genotype output by the input data processing module is converted into the following control dimension instructions through the Pareto solution generator: Spatial balance control instructions are used to control the shape memory alloy variable stiffness network; Transient response control instructions are used for piezoelectric ceramic microfluidic array control; Long-term protection control instructions are used for intelligent wettability coating regulation; After receiving the valve response delay parameter feedback from the field cooperative execution module, the Shapley value arbiter calculates the marginal contribution of each execution unit: Φvalve=ΔTreduction / τresponse; Φcoating=(1-Kcrystallization)×tduration; Output dynamic control parameter set to the execution interface, including valve opening code, voltage application coordinates and duration instruction.

7. The condenser system based on adaptive regulation control according to claim 6, characterized in that: The field coordination execution module is configured to: Receive the dynamic control parameter set output by the game decision control module and execute: The phase change array control layer analyzes the valve opening code, locates the microencapsulated phase change material on the tube bundle support frame, and triggers the heating wire to absorb heat in stages based on the local supercooling state fed back in real time by the data acquisition module. Intelligent interface protection layer analyzes voltage application coordinates and duration instructions, applies a voltage-responsive polymer coating to the target area of ​​the tube wall, and dynamically adjusts the contact angle to achieve a wettability gradient distribution; The actual power curve of the heating wire, the coating contact angle change curve and the valve action delay parameters are fed back to the dynamic knowledge base module.

8. The condenser system based on adaptive regulation control according to claim 7, characterized in that: In response to the voltage application coordinates and duration instructions in the dynamic control parameter set, the dynamic regulation of the intelligent interface protection layer includes: The analytical voltage application coordinates the supercooled area, and a positive voltage of +5 V is applied to make the contact angle <30° to enhance the hydrophilicity; Maintain a hydrophobic state with a contact angle > 90° for the normal area not covered by the coordinates; Generate a contact angle change curve and feed it back to the dynamic knowledge base module to form a non-uniform wetting interface that blocks the adhesion of impurities; The coverage of the hydrophilicity-enhanced region coincides with the spatial coordinates of the gene fragments in the supercooling region.

9. The condenser system based on adaptive regulation control according to claim 8, characterized in that: The dynamic knowledge base module is configured as follows: The actual power curve of the heating wire and the contact angle change curve of the coating received feedback are used to establish: The mapping relationship between the heating wire power fluctuation and the tube wall crystal nucleus growth rate; Correlation between the contact angle gradient and the thermal stress fluctuation amplitude; When the life attenuation probability output by the life prediction model exceeds 15%, a variation strength constraint parameter is generated and transmitted to the data processing module of claim 3 via the industrial bus.

10. The condenser system based on adaptive regulation control according to claim 9, characterized in that: The lifespan prediction includes: The real-time crystal nucleus attachment density data from the integrated ultrasonic diffractometer and the heating wire power fluctuation data are used through the convolutional long short-term memory network: Input the current working condition gene code and historical failure cases; Fusion thermal stress fluctuation amplitude change gradient parameter; Output the probability distribution of the remaining life of the tube bundle to the mutation intensity constraint generator; When the tube bundle strain data detected by the flexible thermocouple array reaches a threshold, the elite-preserving crossover mechanism is triggered to reconstruct the initial population.

Images