A method and system for correcting a valve opening

Abstract

This invention discloses a method and system for correcting the opening degree of a one-way valve, relating to the field of liquid cooling pipeline control. The invention collects full operational data of the one-way valve, constructs an initial benchmark opening degree parameter set through correlation fitting and benchmark mapping, and performs equivalent opening degree extrapolation to obtain preliminary equivalent opening degree calculation results. It then integrates the residual pressure and flow rate of the liquid circuit with the residual heat exchange of the branch circuit to perform thermo-liquid coupling residual correction. Through deviation feature decomposition, fault cause matching, and safety situation assessment, it completes deviation source tracing and backflow risk assessment. Through adaptation screening, interval correction, and closed-loop convergence verification, it outputs the final opening degree correction result. This invention also configures a correction system composed of corresponding functional units, which can correct one-way valve opening degree drift online, automatically trace fault causes, and quantify backflow risk, improving the control accuracy and operating efficiency of the liquid cooling system, reducing maintenance costs, and adapting to the stable operation requirements of direct liquid cooling pipelines in AI data centers.

Description

Technical Field

[0001] This invention relates to the field of liquid cooling heat dissipation pipeline control, and in particular to a method and system for correcting the opening degree of a one-way valve. Background Technology

[0002] As the computing power density of AI servers and high-power GPUs continues to increase, data centers are gradually upgrading from traditional air cooling to direct liquid cooling systems with server cold plate branches and cabinet manifold architectures. One-way valves, as the core passive valves in closed-loop liquid cooling systems, are deployed on the supply side, return side, and branch protection locations, playing a crucial role in preventing backflow, suppressing crossflow, blocking leakage diffusion, and facilitating pump shutdown and recirculation. The actual opening state of the one-way valve directly determines the branch flow rate, cold plate heat exchange efficiency, pump energy consumption, and server temperature control stability. However, conventional one-way valves lack built-in opening detection sensors, making them susceptible to opening drift due to operating conditions, which restricts the reliable operation of the liquid cooling system.

[0003] Existing liquid cooling systems rely on CDUs, pump sets, differential pressure sensors, flow meters, and temperature sensors to achieve flow control and temperature closed-loop at the cabinet and pipeline levels. When branch circuits experience temperature rise, differential pressure disturbances, or flow attenuation, conventional methods often involve increasing pump speed, adjusting control valves, triggering abnormal alarms, or manually venting air, cleaning filter components, and on-site disassembly and inspection of valve components. Check valves are only calibrated for differential pressure and flow characteristics during factory and commissioning phases, and their flow resistance and opening parameters are kept constant by default during operation. Only when fault symptoms such as backflow, water hammer, or significant temperature changes occur do manual troubleshooting of jamming, blockage, and valve core aging issues become necessary.

[0004] However, existing technologies have significant shortcomings in practical applications, making it difficult to meet the high-precision operation requirements of direct liquid cooling systems in AI data centers. Firstly, existing technologies treat the check valve as a fixed flow resistance element, ignoring the impact of factors such as coolant temperature changes, branch bubble accumulation, particle blockage, valve core wear, and spring fatigue on the check valve's opening characteristics during operation. This makes it impossible to correct the true opening degree of the check valve online and accurately identify whether abnormal branch flow is caused by check valve opening drift. Secondly, existing systems lack a heat-liquid coupling correction mechanism centered on the check valve opening. They treat temperature anomalies, flow anomalies, and backflow risks as independent alarm events, failing to establish a unified judgment chain for changes in heat load, changes in liquid resistance, check valve opening deviation, and backflow risk. This results in the inability to simultaneously output the check valve opening correction value, deviation source, backflow risk, flow compensation amount, and maintenance priority, thus hindering accurate online correction of the check valve opening and making it difficult to ensure the stable and efficient operation of the direct liquid cooling system. Therefore, this invention proposes a check valve opening correction method and system. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for correcting the opening degree of a one-way valve, so as to solve the above-mentioned technical problems.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: a one-way valve opening correction method, comprising the following steps: Collect and organize the full operational data of the check valve to obtain the basic dataset for opening correction; The basic dataset for aperture correction is correlated and fitted with a benchmark mapping to obtain an initial benchmark aperture parameter set. The initial reference opening parameter set is subjected to opening equivalent derivation processing to obtain the initial calculation result of the equivalent opening; The initial equivalent opening result is subjected to hydrothermal coupling residual correction processing to obtain the preliminary opening correction result. The preliminary opening correction results are subjected to deviation tracing and risk assessment to obtain the opening correction risk assessment results. The preliminary opening correction results and the opening correction risk assessment results are integrated and optimized in a closed loop to obtain the final opening correction results of the check valve.

[0007] In a preferred embodiment, the aperture correction baseline dataset is correlated and fitted with a benchmark mapping to obtain an initial benchmark aperture parameter set, including: The target operating condition parameters are extracted from the opening correction basic dataset to obtain the basic operating characteristic parameters of the target one-way valve. The basic operating characteristic parameters of the target check valve are calibrated using key performance parameters to obtain the inherent characteristic parameter set of the check valve. The inherent characteristic parameter set of the check valve is correlated and fitted with the basic operating characteristic parameters of the target check valve to obtain the valve port pressure difference flow benchmark mapping relationship. The inherent characteristic parameter set of the one-way valve and the valve port pressure difference flow reference mapping relationship are normalized and fused to obtain the initial reference opening parameter set.

[0008] As a preferred embodiment, the inherent characteristic parameter set of the check valve is correlated and fitted with the basic operating characteristic parameters of the target check valve to obtain the valve orifice pressure differential flow rate benchmark mapping relationship, including: The basic operating characteristic parameters of the target one-way valve are subjected to steady-state condition screening to obtain an effective fitting condition sample set. The inherent characteristic parameter set of the one-way valve is matched and normalized with the effective fitting working condition sample set to obtain the modeling input parameter sequence; The modeling input parameter sequence is subjected to nonlinear correlation fitting to obtain the valve orifice pressure differential flow rate benchmark mapping relationship.

[0009] As a preferred embodiment, the initial reference opening parameter set is subjected to opening equivalent derivation processing to obtain the initial calculation result of the equivalent opening, including: The initial reference opening parameter set is processed by feature parameter extraction to obtain the one-way valve reference feature parameters. The reference characteristic parameters of the one-way valve are coupled and converted with the real-time operating parameters of the opening correction basic dataset to obtain the theoretical calculation value of the opening. The theoretically calculated opening value is subjected to boundary constraints and amplitude limiting to obtain the initial calculation result of the equivalent opening.

[0010] In a preferred embodiment, the initial equivalent aperture calculation result is subjected to hydrothermal coupling residual correction processing to obtain a preliminary aperture correction result, including: The initial calculation results of the equivalent opening are processed by working condition theory deduction based on the valve port pressure difference and flow rate reference mapping relationship to obtain the theoretical pressure difference and theoretical flow rate parameters; The theoretical pressure difference and theoretical flow parameters are compared with the measured operating condition parameters in the opening correction basic dataset to obtain the pressure and flow residuals. The branch operating thermal parameters in the opening correction basic dataset are subjected to supply and demand matching analysis to obtain the branch heat transfer residual. The pressure and flow residuals are coupled and correlated with the heat transfer residuals of the branch to obtain the hydrothermal coupling residual characteristics. Based on the hydrothermal coupling residual characteristics, the initial equivalent opening result is adaptively compensated and corrected to obtain the preliminary opening correction result.

[0011] As a preferred embodiment, the pressure and flow residuals are coupled and correlated with the branch heat transfer residuals to obtain hydrothermal coupling residual characteristics, including: The pressure and flow residuals are subjected to consistency matching and screening processing to obtain the joint deviation characteristics of the liquid circuit. The combined deviation characteristics of the liquid path and the heat transfer residual of the branch are dimensionally correlated and fused to obtain a multi-parameter coupled deviation set. The multi-parameter coupling deviation set is subjected to feature normalization calibration to obtain the hydrothermal coupling residual features.

[0012] As a preferred embodiment, the preliminary opening correction results are subjected to deviation tracing and risk assessment to obtain the opening correction risk assessment results, including: Based on the preliminary opening correction results and the time-series operating condition parameters of the opening correction basic dataset, deviation feature decomposition processing is performed to obtain opening deviation characterization parameters. By combining the opening deviation characterization parameters with the operating condition evolution parameters of the opening correction basic dataset, fault cause matching processing is performed to obtain the opening deviation source type. The safety situation assessment is performed by combining the source type of the opening deviation with the flow direction pressure difference parameter of the opening correction basic dataset to obtain the backflow risk level. The source type of the opening deviation and the backflow risk level are normalized and aggregated to obtain the opening correction risk assessment result.

[0013] In a preferred embodiment, the preliminary opening correction result and the opening correction risk assessment result are integrated and optimized in a closed loop to obtain the final opening correction result of the check valve, including: The preliminary opening correction results and the opening correction risk assessment results are adapted and screened to obtain the opening benchmark parameters to be optimized. The optimized opening reference parameter is matched and corrected with the steady-state constraint conditions of the opening correction base dataset to obtain the optimized opening parameter. Based on the optimized opening parameters, closed-loop convergence verification is performed to obtain the final opening correction result of the check valve.

[0014] In a preferred embodiment, the benchmark parameter of the opening degree to be optimized is matched and corrected with the steady-state constraints of the basic dataset of the opening degree correction to obtain the optimized opening degree parameter, including: The reference parameter of the opening degree to be optimized is subjected to boundary interval calibration to obtain the reference interval range of the parameter; The parameter reference range is compared with the steady-state constraint condition, and the difference is processed to obtain the parameter correction offset. The parameter correction offset is used to compensate and calibrate the opening reference parameter to be optimized to obtain the optimized opening parameter.

[0015] Secondly, the technical solution adopted by the present invention is: a one-way valve opening correction system, comprising: The data collection and construction unit is used to collect and organize the full operating data of the one-way valve to obtain the basic dataset for opening correction, and to perform correlation fitting and benchmark mapping on the basic dataset for opening correction to obtain the initial benchmark opening parameter set. The equivalent deduction calculation unit is used to perform equivalent deduction processing on the initial reference opening parameter set to obtain the initial calculation result of the equivalent opening. The hydrothermal residual correction unit is used to perform hydrothermal coupling residual correction processing on the equivalent opening initial calculation result to obtain the preliminary opening correction result. The deviation risk assessment unit is used to perform deviation source tracing and risk assessment on the preliminary opening correction results to obtain the opening correction risk assessment results. The closed-loop integration optimization unit is used to perform closed-loop integration optimization processing on the preliminary opening correction result and the opening correction risk assessment result to obtain the final opening correction result of the one-way valve.

[0016] The beneficial effects of this invention are: 1. This invention constructs a basic dataset for opening correction by collecting full data of one-way valve operation. Combined with equivalent deduction, thermo-liquid coupling residual correction and other steps, it achieves accurate online correction of the one-way valve opening. It can accurately identify whether abnormal branch flow is caused by opening drift, effectively avoid problems such as branch flow imbalance and reduced cold plate heat exchange efficiency caused by opening drift, and adapt to the high-precision operation requirements of direct liquid cooling system in AI data center.

[0017] 2. This invention overcomes the deficiency of existing systems that lack a heat-liquid coupling correction mechanism centered on the opening degree of the one-way valve. By constructing a coupled correlation judgment logic of pressure and flow residuals and branch heat exchange residuals, a unified judgment chain is established for changes in heat load, changes in liquid resistance, deviation of the one-way valve opening degree, and backflow risk. The opening correction value, source of deviation, and backflow risk level can be output simultaneously, breaking the limitation of existing technologies that handle temperature and flow abnormalities independently, and improving the stability and reliability of the liquid cooling system.

[0018] 3. Through processes such as deviation tracing and risk assessment, closed-loop integration and optimization, the entire process of valve opening correction is automated. Valve opening correction and risk assessment can be completed without manual disassembly and inspection of valve components. At the same time, conservative parameter configurations and temperature zone refresh management strategies are generated to reduce pump energy consumption, reduce manual maintenance costs, ensure long-term stable and efficient operation of the direct liquid cooling system, and adapt to the temperature control requirements of high-computing-power servers.

[0019] 4. This invention eliminates the need for additional dedicated opening sensors for each branch check valve. Relying on existing multi-source data such as pressure, flow, temperature, and server power consumption, combined with controllable pump speed perturbation response and hydrothermal dual residual joint correction, it can simultaneously output the corrected opening value, deviation cause type, backflow risk level, branch flow compensation amount, and maintenance priority. At the same time, it can automatically distinguish four types of abnormal root causes: server heat load fluctuation, cold plate heat exchange attenuation, increased pipeline flow resistance, and check valve opening drift. This upgrades the traditional monitoring mode that only performs over-limit alarms to an integrated decision output that includes quantitative opening correction, fault tracing, flow compensation, and maintenance classification. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation on the scope of this application.

[0021] Figure 1This is a flowchart illustrating the correction method in the embodiment.

[0022] Figure 2 This is a block diagram of the system functional units in the embodiment.

[0023] Figure 3 This is a bar chart comparing the effects of the present invention with those of existing technologies. Detailed Implementation

[0024] To make the technical means, creative features, and achieved objectives and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention. Unless otherwise specified, the experimental methods in the following embodiments are conventional methods, and the materials and reagents used in the following embodiments are commercially available unless otherwise specified.

[0025] like Figure 1 and Figure 2 As shown in the figure, this embodiment provides a method for correcting the opening degree of a one-way valve.

[0026] In this embodiment, the method includes the following steps: Collect and organize the full operational data of the check valve to obtain the basic dataset for opening correction; This step initiates the multi-source data acquisition module from the data collection and construction unit, uniformly collecting five types of structured numerical data, including: inherent valve parameters, such as rated opening differential pressure, spring stiffness, maximum valve core stroke, and valve core force area, which do not change with operating conditions; pipeline structural parameters, such as branch pipe length, pipe diameter, and pipeline reference flow resistance; real-time liquid circuit operating parameters, such as inlet pressure, outlet pressure, real-time flow rate, and supply liquid temperature; branch thermal operating parameters, such as server heat load power consumption, actual heat exchange power of cold plate, and ambient temperature; and time-series operating status parameters, such as running time, number of start-stop cycles, and sampling timestamps.

[0027] Simultaneously, coolant type and concentration data are collected to determine coolant density and viscosity; target branch number and target check valve number are also collected to distinguish calibration data for different branches and valves, avoiding confusion. Coolant density and viscosity are preferentially obtained from the coolant manufacturer's datasheet; when on-site coolant mixing or aging occurs, the data is updated based on sampling and testing results during the maintenance phase. The sampling period for branch pressure and flow is consistent with the CDU control period by default; when the CDU control period is greater than 1 second, the data sampling period for the check valve calibration phase is set to 1 second to ensure that response changes caused by pump speed perturbations are captured.

[0028] The above data was continuously collected at a fixed sampling period of 100ms. The data collection and construction unit preprocessed the collected raw time-series data using 3... The criteria include: eliminating outliers to prevent extreme data from interfering with subsequent calculations; using linear interpolation to fill in short-term missing sampling points to ensure data continuity; and standardizing the dimensions of all physical quantities to ensure the rationality of parameter calculations. After preprocessing, the data collection and construction unit uses branch number, valve number, and timestamp as joint primary keys to organize multi-source heterogeneous data into a structured table format, generating a basic dataset for opening correction. A standard data interface is also reserved for subsequent units to call this dataset for related processing.

[0029] The basic dataset for aperture correction is correlated and fitted with a benchmark mapping to obtain an initial benchmark aperture parameter set. This step is performed by the data collection and construction unit. Based on the basic dataset for aperture correction, a benchmark parameter system for subsequent aperture extrapolation is constructed through hierarchical processing. From weight 2, the basic dataset for aperture correction is correlated, fitted, and mapped to the benchmark to obtain an initial benchmark aperture parameter set, including: The target operating condition parameters are extracted from the opening correction basic dataset to obtain the basic operating characteristic parameters of the target one-way valve. Specifically, the data collection and construction unit calls the basic dataset for opening correction. First, it sets a clear steady-state operating condition screening threshold to filter out periods when the system is running continuously and stably. It focuses on eliminating disturbance data such as system start-up and shutdown, pump frequency adjustment, and sudden load changes to ensure the stability and representativeness of the extracted operating parameters. Then, it extracts the core operating parameters within this steady-state range and calculates their average characteristics, including average inlet and outlet pressure difference, average real-time flow rate, average liquid supply temperature, average heat load power consumption, and average cold plate heat exchange power. These average characteristics are integrated to obtain the basic operating characteristic parameters of the target one-way valve, providing a steady-state basis for subsequent parameter calibration and modeling.

[0030] The basic operating characteristic parameters of the target check valve are calibrated using key performance parameters to obtain the inherent characteristic parameter set of the check valve. Furthermore, the data collection and construction unit, based on the fundamental operating characteristic parameters of the target check valve, eliminates interference from external factors such as pipeline flow resistance, temperature disturbances, and thermal load fluctuations. Through inversion calibration, it determines the intrinsic performance parameters of the valve itself. Specifically, it inverts and calibrates the intrinsic flow resistance coefficient of the valve by combining differential pressure and flow data under steady-state conditions; it calibrates parameters such as spring stiffness and inherent opening differential pressure by combining the principle of spring force balance; and it determines the correspondence coefficient between valve core stroke and opening degree by combining the valve core mechanical structure. These inherent parameters, which do not change with operating conditions, are encapsulated into a set of inherent characteristic parameters for the check valve, ensuring the uniqueness and stability of the parameters.

[0031] The inherent characteristic parameter set of the check valve is correlated and fitted with the basic operating characteristic parameters of the target check valve to obtain the valve orifice pressure differential flow rate benchmark mapping relationship, including: The basic operating characteristic parameters of the target one-way valve are subjected to steady-state condition screening to obtain an effective fitting condition sample set. The tolerance thresholds for fluctuations were further tightened, narrowing the allowable ranges for pressure fluctuations, flow rate fluctuations, and temperature fluctuations. Samples were required to exhibit long-term continuous stability without transient pulse interference, ensuring the selected samples have good representativeness and consistency. Specifically, pressure fluctuations were ≤ ±2000 Pa, flow rate fluctuations were ≤ ±1.5%, and temperature fluctuations were ≤ ±0.8℃. Samples meeting these conditions were aggregated to form a sufficient and evenly distributed effective set of fitting operating conditions, providing high-quality data support for subsequent fitting modeling.

[0032] The inherent characteristic parameter set of the one-way valve is matched and normalized with the effective fitting working condition sample set to obtain the modeling input parameter sequence; The data collection and construction unit performs a one-to-one temporal matching between each parameter in the inherent characteristic parameter group of the one-way valve and each working condition sample in the effective fitting working condition sample set, ensuring that each sample can be associated with the corresponding inherent parameters of the valve. Since the dimensions and numerical magnitudes of different physical quantities vary significantly, to avoid affecting the fitting accuracy, the data collection and construction unit adopts a Min-Max normalization method, uniformly mapping all parameters to the [0,1] standard interval to eliminate differences in dimensions and magnitudes. Subsequently, the normalized parameters are arranged according to the sample temporal sequence to form a standardized and usable sequence of modeling input parameters.

[0033] The modeling input parameter sequence is subjected to nonlinear correlation fitting to obtain the valve orifice pressure differential flow rate benchmark mapping relationship; The data collection and construction unit, based on the modeling input parameter sequence, employs a multivariate nonlinear regression fitting method to establish a continuous interpolable correlation mapping model between branch inlet and outlet pressure difference, medium temperature, and real-time pipeline flow. Through fitting calculations, the coefficients of each element in the model are determined, the model expression is solidified, and a baseline mapping relationship between valve orifice pressure difference and flow is obtained. This mapping relationship is used to subsequently deduce the corresponding theoretical operating parameters based on the valve opening degree, enabling comparative analysis between theoretical and measured values.

[0034] The inherent characteristic parameter set of the one-way valve and the valve port pressure differential flow reference mapping relationship are normalized and fused to obtain the initial reference opening parameter set; The data collection and construction unit takes the calibrated inherent characteristic parameter set of the check valve, the fitted valve orifice pressure differential flow benchmark mapping relationship, and combines it with the check valve's opening physical constraint range (0-100%) and rated operating condition benchmark point, and uniformly organizes and encapsulates them to form a structured, directly callable initial benchmark opening parameter set. The data collection and construction unit then distributes this parameter set to the equivalent derivation calculation unit, providing core parameter support for subsequent opening equivalent derivation steps.

[0035] The initial reference opening parameter set is processed by equivalent opening derivation to obtain the initial equivalent opening result. This step is executed by the equivalent derivation calculation unit, which realizes the inverse derivation of the one-way valve opening under sensorless conditions based on the initial reference opening parameter set; including: The initial reference opening parameter set is processed by feature parameter extraction to obtain the one-way valve reference feature parameters. In this step, the equivalent deduction calculation unit receives the initial benchmark opening parameter set issued by the data collection and construction unit, filters and extracts the parameter set, removes redundant auxiliary parameters, and focuses on extracting the core parameters used for opening deduction, including the rated opening differential pressure, spring stiffness, valve core force area, and valve core maximum stroke in the one-way valve inherent characteristic parameter group, as well as the valve port differential pressure-flow benchmark mapping relationship and the opening physical constraint range. The core parameters are integrated to obtain the one-way valve benchmark characteristic parameters, simplifying the parameter dimensions and improving deduction efficiency.

[0036] The reference characteristic parameters of the one-way valve are coupled and converted with the real-time operating parameters of the opening correction basic dataset to obtain the theoretical calculation value of the opening. Specifically, the equivalent derivation calculation unit retrieves real-time operating parameters from the opening correction basic dataset generated by the data collection and construction unit through a data interface, focusing on extracting the real-time inlet pressure and real-time outlet pressure to calculate the real-time inlet and outlet pressure difference of the branch. Furthermore, using a sensorless equivalent opening derivation formula, the reference characteristic parameters of the one-way valve are coupled and converted with the real-time pressure difference to deduce the theoretical opening value under sensorless conditions.

[0037] Specifically, the formula for calculating the initial value of the equivalent aperture is as follows: ; in, The initial value of the equivalent opening (%); The real-time inlet and outlet pressure difference (Pa) of the branch line; The inherent opening pressure differential of the valve (Pa); Spring stiffness (N / m); The effective force-bearing area of ​​the valve core (m²) 2 ); This represents the maximum stroke of the valve core (m).

[0038] The theoretically calculated opening value is subjected to boundary constraints and amplitude limiting to obtain the initial calculation result of the equivalent opening. Furthermore, the equivalent derivation calculation unit sets reasonable upper and lower boundaries for the opening degree based on the mechanical and physical limits of the check valve. It then performs boundary truncation on theoretically calculated opening degrees exceeding these boundaries to ensure that the calculation results conform to the actual operating range of the check valve. Simultaneously, to suppress opening degree fluctuations caused by instantaneous operating condition pulse disturbances, a sliding mean filtering method is used to calculate the average of the most recent 10 sets of theoretically calculated opening degrees, smoothing out fluctuations. Finally, a stable and reliable preliminary calculation result for the equivalent opening degree is output, and this result is sent to the hydrothermal residual correction unit for subsequent correction processing.

[0039] The initial equivalent opening calculation result is subjected to hydrothermal coupling residual correction processing to obtain a preliminary opening correction result. This step is performed by the hydrothermal residual correction unit to eliminate the opening deviation caused by the coupling of thermal and liquid circuit conditions, and to accurately correct the initial equivalent opening calculation result; including: The initial calculation results of the equivalent opening degree are processed by working condition theory deduction based on the valve port pressure difference and flow rate reference mapping relationship to obtain the theoretical pressure difference and theoretical flow rate parameters; Specifically, the hydrothermal residual correction unit receives the preliminary calculation result of the equivalent opening from the equivalent derivation calculation unit, and simultaneously retrieves the valve orifice pressure differential flow rate benchmark mapping relationship generated by the data collection and construction unit through the data interface. Using the preliminary calculation result of the equivalent opening as a benchmark, and combined with the real-time supply temperature, the unit performs a forward derivation through the valve orifice pressure differential flow rate benchmark mapping relationship to obtain the corresponding standard theoretical pressure differential and theoretical flow rate under this opening condition, forming a theoretical operating condition parameter benchmark, which serves as a reference standard for subsequent comparison with measured values.

[0040] The theoretical pressure difference and theoretical flow parameters are compared with the measured operating condition parameters in the opening correction basic dataset to obtain the pressure and flow residuals. Furthermore, the hydrothermal residual correction unit retrieves the measured operating parameters at the current moment from the opening correction basic dataset, focusing on extracting the measured inlet and outlet pressure difference and the measured flow rate. These are then compared with the theoretical pressure difference and theoretical flow rate, and the absolute differences are calculated to obtain the pressure residual and flow rate residual, respectively. The pressure residual and flow rate residual are used to visually characterize the deviation between the theoretical and actual operating values ​​of the liquid circuit, providing basic data for subsequent residual coupling analysis.

[0041] The branch operating thermal parameters in the opening correction basic dataset are subjected to supply and demand matching analysis to obtain the branch heat transfer residual. In this embodiment, the hydrothermal residual correction unit continues to retrieve the branch operating thermal parameters from the opening correction basic dataset, extracts the server thermal load power consumption and the actual heat exchange power of the cold plate, and performs a heat dissipation supply and demand matching analysis on the two. The server thermal load power consumption represents the heat dissipation demand of the branch, and the actual heat exchange power of the cold plate represents the heat dissipation capacity of the branch. The absolute difference between the two is calculated to obtain the branch heat exchange thermal residual, which quantitatively characterizes the degree of mismatch between the thermal load demand and the actual heat exchange capacity, thereby quantifying the thermal condition deviation.

[0042] Based on server power consumption, supply liquid temperature, return liquid temperature, and branch flow rate, branch thermal load matching results are further generated to distinguish the causes of temperature anomalies and avoid misjudging one-way valve opening deviations: when server power consumption increases and branch return liquid temperature increases simultaneously, while branch flow rate does not decrease, this change is marked as a thermal load-dominant change; when server power consumption does not increase significantly but branch flow rate decreases and return liquid temperature increases, this change is marked as a liquid supply insufficiency change; when branch flow rate decreases and differential pressure increases, this change is marked as an increase in resistance change; when branch flow rate decreases but differential pressure does not increase simultaneously, this change is marked as a suspected one-way valve opening insufficiency change. The power consumption change threshold is set by default to the 75th percentile of the target server's load fluctuation for similar tasks over the past 7 days; if the system deployment is less than 7 days old initially, the data from the most recent 24 hours is used, and the data is automatically replaced after 7 days of operation.

[0043] The pressure and flow residuals are coupled and correlated with the branch heat transfer residuals to obtain hydrothermal coupling residual characteristics, including: The pressure and flow residuals are subjected to consistency matching and screening processing to obtain the joint deviation characteristics of the liquid circuit. Specifically, the hydrothermal residual correction unit compares the deviation direction and fluctuation amplitude of the pressure residual and the flow residual, identifies the occasional outliers and effective system deviations, eliminates meaningless instantaneous disturbance deviations, and integrates the core characteristics of the two residuals: amplitude, deviation direction, and stability to form a joint deviation characteristic of the liquid circuit that can comprehensively characterize the liquid circuit offset state and fully reflect the deviation of the liquid circuit operating conditions.

[0044] The combined deviation characteristics of the liquid path and the heat transfer residual of the branch are dimensionally correlated and fused to obtain a multi-parameter coupled deviation set. In this embodiment, the hydrothermal residual correction unit introduces fixed weighting coefficients: the total weighting coefficients are 1, where the pressure residual weight is 0.4, the flow residual weight is 0.3, and the thermal residual weight is 0.3. The pressure residual and flow residual in the joint deviation characteristics of the liquid circuit are cross-dimensionally correlated and fused with the heat transfer residual of the branch circuit. Through weighted calculation and integration, a multi-parameter coupled deviation set containing multi-dimensional deviation information is formed, realizing the collaborative characterization of thermal condition and liquid circuit condition deviation.

[0045] The multi-parameter coupling deviation set is subjected to feature normalization calibration to obtain the hydrothermal coupling residual features; Specifically, the hydrothermal residual correction unit performs dimensionless normalization on the set of multi-parameter coupling deviations, mapping it to the [0,1] interval to eliminate the dimensional differences between different residuals. Subsequently, based on the normalized deviation values, it classifies the deviations into three levels: slight, moderate, and severe, and identifies the specific direction of the deviation, such as low flow rate or insufficient heat exchange, thus solidifying them into hydrothermal coupling residual characteristics that can be directly used for compensation and correction.

[0046] When a suspected change in the one-way valve opening is observed in the target branch, and the server temperature does not exceed the safety limit, the hydrothermal residual correction unit sends a low-amplitude perturbation command to the CDU pump speed control terminal to execute a low-amplitude pump speed perturbation. Simultaneously, it collects the pressure difference response, flow response, and temperature response data after the perturbation. The pump speed perturbation amplitude is set to 3% to 5% of the current pump speed by default; the perturbation duration is set to 10 to 30 seconds by default. If the server temperature is less than 5°C from the safety limit, the perturbation is not executed, and only existing operating data is used for conservative correction. The perturbation amplitude and duration are constrained to not trigger server frequency reduction, not cause branch pressure to exceed limits, and not cause sudden changes in return liquid temperature, selecting the minimum pump speed change that can cause identifiable pressure difference and flow response.

[0047] Meanwhile, the hydrothermal residual correction unit uses the hydrothermal coupling integrated residual calculation formula to quantify the coupling residual, providing a quantitative basis for deviation level calibration; The formula for calculating the hydrothermal coupling integrated residual is as follows: ; in, This is the residual of hydrothermal coupling; , , These are weighting coefficients that sum to 1; The measured pressure difference between the inlet and outlet (Pa); The theoretical inlet and outlet pressure difference (Pa); Measured flow rate (m³) 3 / s); Theoretical flow rate (m 3 / s); The branch thermal load power consumption (W); The actual heat exchange power of the cold plate (W); Based on the hydrothermal coupling residual characteristics, the initial calculation result of the equivalent opening is adaptively compensated and corrected to obtain the preliminary opening correction result. In this embodiment, before compensation correction, a logical judgment is first made based on the consistency of pressure residual, flow residual, and thermal residual: pressure residual represents the deviation of the measured differential pressure from the theoretical differential pressure under the reference opening relationship; flow residual represents the deviation of the measured flow rate from the theoretical flow rate; thermal residual represents the deviation of the actual heat exchange capacity of the current branch from the server power consumption requirements. When both pressure residual and flow residual indicate insufficient flow capacity, and thermal residual indicates insufficient heat exchange, it is determined that the equivalent opening needs to be reduced for correction; when the pressure residual is abnormal but the thermal residual is not abnormal, it is marked as short-term liquid path fluctuation, and no opening correction is performed; when the thermal residual is abnormal but the pressure and flow residual are not abnormal, it is determined that the server thermal load or cold plate heat exchange is abnormal, and it is not attributed to the one-way valve opening deviation. The thermal residual judgment threshold is determined by historical normal data under the same model of server, the same type of cold plate, and the same type of load. The default value is the 95th percentile of the absolute value of the normal thermal residual, and it can be iteratively updated by fault maintenance samples.

[0048] The hydrothermal residual correction unit matches corresponding adaptive compensation correction coefficients based on the deviation level and direction of the hydrothermal coupling residual characteristics: minor deviations correspond to smaller correction coefficients, such as 0.99; moderate deviations correspond to medium correction coefficients, such as 0.95; and severe deviations correspond to larger correction coefficients, such as 0.90. The compensation correction coefficients are multiplied by the initial equivalent opening calculation result to perform adaptive incremental correction on the initial equivalent opening result, eliminating the system deviation caused by hydrothermal operating condition coupling, obtaining the corrected preliminary opening correction result, and then sending this result to the deviation risk assessment unit for subsequent deviation source tracing and risk assessment.

[0049] The preliminary opening correction results are then subjected to deviation source tracing and risk assessment to obtain the opening correction risk assessment results. This step, performed by the deviation risk assessment unit, is used to clarify the causes of deviations in the preliminary opening correction results and assess backflow risk, providing a constraint basis for subsequent closed-loop optimization; including: Based on the preliminary opening correction results and the time-series operating condition parameters of the opening correction basic dataset, deviation feature decomposition processing is performed to obtain opening deviation characterization parameters. In this step, the deviation risk assessment unit receives the preliminary opening correction results from the hydrothermal residual correction unit, and simultaneously retrieves long-term operating condition and opening data from the opening correction basic dataset, specifically the most recent 100 consecutive samples with a time span of 10 seconds. The deviation characteristics of the preliminary opening correction results are decomposed and quantified to obtain three core characterization parameters: first, the deviation amplitude, i.e., the difference between the preliminary opening correction result and the opening at the rated operating condition benchmark; second, the fluctuation range, i.e., the maximum fluctuation range of the most recent 100 opening values; and third, the drift rate, i.e., the rate of change of opening within the most recent 10 seconds. These quantified parameters comprehensively characterize the specific situation of the opening deviation.

[0050] By combining the opening deviation characterization parameters with the operating condition evolution parameters of the opening correction basic dataset, fault cause matching processing is performed to obtain the opening deviation source type. Specifically, the deviation risk assessment unit pre-constructs a rule base for deviation causes. This rule base contains characteristic thresholds for different deviation types, such as valve core wear, spring fatigue, particle blockage, and temperature drift disturbances, including drift rate thresholds and fluctuation amplitude thresholds. The opening deviation characteristic parameter is matched with the operating condition evolution parameters in the opening correction base dataset, such as the changing trends of differential pressure, flow rate, and temperature, and historical maintenance records, to automatically trace and determine the specific cause of the opening deviation, obtain the opening deviation source type, and clarify the root cause of the deviation.

[0051] Further precise source tracing is achieved by combining the response characteristics after pump speed perturbation: when the differential pressure changes significantly after perturbation but the flow response is lagging and recovers slowly, it is judged as suspected valve core jamming; when the corrected opening value continues to decrease for a long time under the same differential pressure, it is judged as suspected spring fatigue or impurity deposition blockage; when the opening value fluctuates irregularly for a short period of time and returns to normal after branch venting, it is judged as suspected pipeline bubble interference; when there are no obvious abnormalities in pressure and flow, but the thermal residual is continuously high, it is judged as suspected abnormal cold plate heat transfer attenuation. The perturbation response delay threshold is set by default to twice the normal response time during the commissioning phase. The normal response time during commissioning is calibrated by taking the median of three standard pump speed perturbation tests.

[0052] The safety situation assessment is performed by combining the source type of the opening deviation with the flow direction pressure difference parameter of the opening correction basic dataset to obtain the backflow risk level. In this step, the deviation risk assessment unit retrieves the branch flow direction pressure difference parameters from the opening correction basic dataset, focusing on analyzing the distribution characteristics of the branch forward and reverse pressure differences. Combined with the type of opening deviation tracing, it assesses the trend of deviation deterioration and evolution. Based on the degree of insufficient forward pressure difference, the magnitude of the reverse pressure difference, and the severity of the deviation tracing type, the backflow risk is divided into four levels: no risk, low backflow risk, medium backflow risk, and high backflow risk. The current backflow risk level is determined, and the potential safety hazards caused by the opening deviation are assessed.

[0053] The source type of the opening deviation and the backflow risk level are normalized and aggregated to obtain the opening correction risk assessment result. In this step, the deviation risk assessment unit will structurally integrate the type of opening deviation source, the opening deviation characterization parameters, the time series drift trend, and the backflow risk level to form a complete opening correction risk assessment result. This will clarify the root cause, severity, and safety hazard level of the deviation, providing a clear constraint basis for the subsequent closed-loop integration and optimization steps. At the same time, the assessment result will be sent to the closed-loop integration and optimization unit.

[0054] The preliminary opening correction result and the opening correction risk assessment result are integrated and optimized in a closed loop to obtain the final opening correction result of the one-way valve. This step is executed by the closed-loop integration and optimization unit to combine the opening correction risk assessment result with the preliminary opening correction result to ensure the accuracy and safety of the final opening correction result; specifically, it includes the following steps: The preliminary opening correction results and the opening correction risk assessment results are adapted and screened to obtain the opening benchmark parameters to be optimized. Specifically, the closed-loop integration optimization unit receives the preliminary opening correction results from the hydrothermal residual correction unit and the opening correction risk assessment results from the deviation risk assessment unit. Based on the backflow risk level and deviation severity in the opening correction risk assessment results, it performs adaptation screening: correction points corresponding to high backflow risk and severe deviation are eliminated, and preliminary opening correction results with no risk, low risk, slight to moderate deviation and meeting the system safety operation constraints are retained. These are used as the opening benchmark parameters to be optimized to ensure that the basic parameters for subsequent optimization have safety and availability.

[0055] The optimized opening parameter is obtained by matching and correcting the benchmark parameter to be optimized with the steady-state constraints of the opening correction dataset, including: The reference parameter of the opening degree to be optimized is subjected to boundary interval calibration to obtain the reference interval range of the parameter; Furthermore, the closed-loop integrated optimization unit combines the mechanical and physical limits of the check valve with the steady-state operation specifications of the liquid cooling system to calibrate the upper and lower limits of the safe operating range of the opening degree reference parameter to be optimized, clarify the range of the parameter reference range, and ensure that the optimized opening degree is within a safe and reasonable range.

[0056] The parameter reference range is compared with the steady-state constraint condition, and the difference is processed to obtain the parameter correction offset. Furthermore, the closed-loop integration optimization unit compares the parameter reference range with the system steady-state constraints in the opening correction basic dataset to quantify the degree to which the current opening reference parameter deviates from the safe steady-state range, calculates the parameter correction offset, and clarifies the direction and magnitude of the opening correction.

[0057] The parameter correction offset is used to compensate and calibrate the opening reference parameter to be optimized to obtain the optimized opening parameter. In this embodiment, the closed-loop integration optimization unit compensates and calibrates the opening reference parameter to be optimized within the parameter reference range based on the parameter correction offset, adjusts the opening value to simultaneously meet the mechanical limit of the one-way valve and the steady-state operation constraint of the liquid cooling system, eliminates the problem of the opening deviating from the safe steady-state range, and obtains the optimized opening parameter.

[0058] Based on the optimized opening parameters, a closed-loop convergence verification process is performed to obtain the final opening correction result of the check valve. The closed-loop integrated optimization unit adopts an iterative approximation method to make small adjustments to the optimized opening parameter one by one. After each iteration, the relative error of the opening between two adjacent iterations is calculated, and the iterative convergence criterion formula is used to determine whether the iteration is complete.

[0059] The iterative convergence criterion formula is: ; in, The relative error of the opening degree between adjacent iterations; For the first Opening degree (%) after the second iteration of optimization; For the first Opening degree (%) after the second iteration of optimization; To preset the convergence threshold, this embodiment uses 0.005, corresponding to 0.5% error.

[0060] When the relative error between two consecutive iterations of the opening degree is less than the set convergence threshold, the iteration is considered complete, optimization stops, and the final opening degree correction result of the one-way valve is fixed and output. Subsequently, the closed-loop integrated optimization unit synchronously pushes the final opening degree correction result to the liquid cooling system management platform to complete the parameter update. At the same time, based on the opening degree correction risk assessment result, alarm information such as the cause of deviation and risk level is synchronously pushed to achieve closed-loop control of the entire opening degree correction process.

[0061] Simultaneously, based on the corrected opening value, opening deviation, backflow risk level, perturbation response delay, and thermal residual duration, a unified flow compensation amount and equipment maintenance priority are generated. The opening deviation is the difference between the benchmark nominal opening and the corrected actual opening. Backflow risk is comprehensively determined by the frequency of reverse pressure difference occurrences in the low differential pressure range, the reverse flow trend of the branch flow, and the one-way valve closing response delay. The flow compensation amount is generated by converting the branch's rated heat exchange demand with the current actual flow difference. For ordinary passive one-way valves, the flow compensation amount is converted into pump speed fine-tuning and bypass branch adjustment commands; for one-way valves with adjustable preload or electronically assisted structures, it can be directly converted into an opening fine-tuning control amount.

[0062] The closed-loop integration and optimization unit receives on-site operation and maintenance feedback data, including branch venting records, filter cleaning records, check valve disassembly and replacement records, valve core wear detection results, spring deformation status, cold plate descaling and cleaning records, and branch steady-state operation data after maintenance. If maintenance confirms that the fault is caused by check valve jamming, blockage, or spring aging, the corresponding operating condition sample is classified into the valve component abnormality sample set; if maintenance confirms that the check valve is normal and the fault originates from the cold plate or pipeline filter, it is classified into the non-valve component abnormality sample set. Based on the two types of sample sets, the system adaptively updates the check valve opening differential pressure, flow coefficient, thermal residual judgment threshold, perturbation response delay threshold, and maintenance priority judgment rules, and sends the updated benchmark parameters back to the data collection and construction unit, equivalent deduction calculation unit, and hot liquid residual correction unit to achieve a full-process self-learning closed loop, adapting to different coolant aging degrees, different branch pipeline characteristics, and long-term aging conditions of valve components.

[0063] This embodiment achieves precise online correction of the one-way valve opening through the above steps. To intuitively demonstrate the technical advantages of this invention compared to existing technologies, the following is set up: Figure 3 The results are shown in the bar chart.

[0064] Figure 3 In this study, using the existing one-way valve control method of liquid cooling systems as a benchmark, three core indicators—opening degree correction accuracy, correction efficiency, and operation and maintenance cost—were selected for normalized comparison. Opening accuracy correction: Existing technologies are limited by the inability to correct opening drift online, and can only achieve about 95% control accuracy. This invention improves the correction accuracy to over 99.5% through hydrothermal coupling residual correction, deviation tracing and closed-loop optimization, which significantly reduces the risk of flow deviation and insufficient heat exchange.

[0065] Calibration efficiency: Existing technologies rely on manual inspection and valve disassembly, resulting in long calibration cycles and low efficiency. This invention achieves fully automated online calibration through sensorless equivalent deduction and pump speed perturbation-assisted calibration, improving calibration efficiency by more than 50% compared to existing technologies.

[0066] Maintenance costs: Existing technologies require frequent manual inspections and disassembly of check valves, resulting in high maintenance costs; this invention can automatically identify the causes of deviations and output maintenance priorities, significantly reducing ineffective maintenance actions and lowering maintenance costs by more than 60% compared to existing technologies.

[0067] In summary, this invention achieves higher calibration accuracy, higher operating efficiency, and lower maintenance costs, solves the core pain points of existing technologies, and possesses significant technical advantages and inventiveness.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0069] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this application and form different embodiments. For example, all the embodiments above can be used in any combination. The information disclosed in this background section is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art. The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method of correcting for opening of a check valve, characterized by, Includes the following steps: Collect and organize the full operational data of the check valve to obtain the basic dataset for opening correction; The basic dataset for aperture correction is correlated and fitted with a benchmark mapping to obtain an initial benchmark aperture parameter set. The initial reference opening parameter set is subjected to opening equivalent derivation processing to obtain the initial calculation result of the equivalent opening; The initial equivalent opening result is subjected to hydrothermal coupling residual correction processing to obtain the preliminary opening correction result. The preliminary opening correction results are subjected to deviation tracing and risk assessment to obtain the opening correction risk assessment results. The preliminary opening correction results and the opening correction risk assessment results are integrated and optimized in a closed loop to obtain the final opening correction results of the check valve.

2. The correction method of claim 1, wherein The aperture correction baseline dataset is correlated and fitted with a benchmark mapping to obtain an initial benchmark aperture parameter set, including: The target operating condition parameters are extracted from the opening correction basic dataset to obtain the basic operating characteristic parameters of the target one-way valve. The basic operating characteristic parameters of the target check valve are calibrated using key performance parameters to obtain the inherent characteristic parameter set of the check valve. The inherent characteristic parameter set of the check valve is correlated and fitted with the basic operating characteristic parameters of the target check valve to obtain the valve port pressure difference flow benchmark mapping relationship. The inherent characteristic parameter set of the one-way valve and the valve port pressure difference flow reference mapping relationship are normalized and fused to obtain the initial reference opening parameter set.

3. The correction method of claim 2, wherein The inherent characteristic parameter set of the check valve is correlated and fitted with the basic operating characteristic parameters of the target check valve to obtain the valve orifice pressure differential flow rate benchmark mapping relationship, including: The basic operating characteristic parameters of the target one-way valve are subjected to steady-state condition screening to obtain an effective fitting condition sample set. The inherent characteristic parameter set of the one-way valve is matched and normalized with the effective fitting working condition sample set to obtain the modeling input parameter sequence; The modeling input parameter sequence is subjected to nonlinear correlation fitting to obtain the valve orifice pressure differential flow rate benchmark mapping relationship.

4. The correction method of claim 1, wherein The initial reference aperture parameter set is subjected to aperture equivalence derivation processing to obtain the initial equivalent aperture calculation results, including: The initial reference opening parameter set is processed by feature parameter extraction to obtain the one-way valve reference feature parameters. The reference characteristic parameters of the one-way valve are coupled and converted with the real-time operating parameters of the opening correction basic dataset to obtain the theoretical calculation value of the opening. The theoretically calculated opening value is subjected to boundary constraints and amplitude limiting to obtain the initial calculation result of the equivalent opening.

5. The correction method of claim 3, wherein The initial equivalent aperture calculation results are subjected to hydrothermal coupling residual correction processing to obtain preliminary aperture correction results, including: The initial calculation results of the equivalent opening are processed by working condition theory deduction based on the valve port pressure difference and flow rate reference mapping relationship to obtain the theoretical pressure difference and theoretical flow rate parameters; The theoretical pressure difference and theoretical flow parameters are compared with the measured operating condition parameters in the opening correction basic dataset to obtain the pressure and flow residuals. The branch operating thermal parameters in the opening correction basic dataset are subjected to supply and demand matching analysis to obtain the branch heat transfer residual. The pressure and flow residuals are coupled and correlated with the heat transfer residuals of the branch to obtain the hydrothermal coupling residual characteristics. Based on the hydrothermal coupling residual characteristics, the initial equivalent opening result is adaptively compensated and corrected to obtain the preliminary opening correction result.

6. The correction method of claim 5, wherein, The pressure and flow residuals are coupled and correlated with the branch heat transfer residuals to obtain hydrothermal coupling residual characteristics, including: The pressure and flow residuals are subjected to consistency matching and screening processing to obtain the joint deviation characteristics of the liquid circuit. The combined deviation characteristics of the liquid path and the heat transfer residual of the branch are dimensionally correlated and fused to obtain a multi-parameter coupled deviation set. The multi-parameter coupling deviation set is subjected to feature normalization calibration to obtain the hydrothermal coupling residual features.

7. The correction method of claim 1, wherein The preliminary opening correction results are then subjected to deviation tracing and risk assessment to obtain the opening correction risk assessment results, including: Based on the preliminary opening correction results and the time-series operating condition parameters of the opening correction basic dataset, deviation feature decomposition processing is performed to obtain opening deviation characterization parameters. By combining the opening deviation characterization parameters with the operating condition evolution parameters of the opening correction basic dataset, fault cause matching processing is performed to obtain the opening deviation source type. The safety situation assessment is performed by combining the source type of the opening deviation with the flow direction pressure difference parameter of the opening correction basic dataset to obtain the backflow risk level. The source type of the opening deviation and the backflow risk level are normalized and aggregated to obtain the opening correction risk assessment result.

8. The correction method of claim 1, wherein The preliminary opening correction results and the opening correction risk assessment results are integrated and optimized in a closed loop to obtain the final opening correction results of the one-way valve, including: The preliminary opening correction results and the opening correction risk assessment results are adapted and screened to obtain the opening benchmark parameters to be optimized. The optimized opening reference parameter is matched and corrected with the steady-state constraint conditions of the opening correction base dataset to obtain the optimized opening parameter. Based on the optimized opening parameters, closed-loop convergence verification is performed to obtain the final opening correction result of the check valve.

9. The correction method of claim 8, wherein, The optimized opening parameter is obtained by matching and correcting the benchmark parameter to be optimized with the steady-state constraints of the opening correction dataset, including: The reference parameter of the opening degree to be optimized is subjected to boundary interval calibration to obtain the reference interval range of the parameter; The parameter reference range is compared with the steady-state constraint condition, and the difference is processed to obtain the parameter correction offset. The parameter correction offset is used to compensate and calibrate the opening reference parameter to be optimized to obtain the optimized opening parameter.

10. A system for correcting the opening of a check valve, suitable for use in the method of correcting according to any one of claims 1 to 9, characterized in that, include: The data collection and construction unit is used to collect and organize the full operating data of the one-way valve to obtain the basic dataset for opening correction, and to perform correlation fitting and benchmark mapping on the basic dataset for opening correction to obtain the initial benchmark opening parameter set. The equivalent deduction calculation unit is used to perform equivalent deduction processing on the initial reference opening parameter set to obtain the initial calculation result of the equivalent opening. The hydrothermal residual correction unit is used to perform hydrothermal coupling residual correction processing on the equivalent opening initial calculation result to obtain the preliminary opening correction result. The deviation risk assessment unit is used to perform deviation source tracing and risk assessment on the preliminary opening correction results to obtain the opening correction risk assessment results. The closed-loop integration optimization unit is used to perform closed-loop integration optimization processing on the preliminary opening correction result and the opening correction risk assessment result to obtain the final opening correction result of the one-way valve.

Images