A liquid cooling system self-optimization control method, system, computer equipment and medium

Abstract

The present invention provides a liquid cooling system self-optimization control method, system, computer equipment and medium, wherein the method is to analyze the current working status of the liquid cooling system and determine it to be in an abnormal state, identify the key pipeline nodes with abnormal flow phenomena as pipeline nodes to be analyzed according to the coolant state data at each preset key pipeline node, and perform blockage simulation analysis and blockage process prediction based on the node blockage prediction results obtained by flow vector analysis of the pipeline node to be analyzed. After obtaining the node blockage process prediction results, a blockage warning signal is generated, and when the blockage warning signal is received, a blockage response strategy is executed according to the preset adjustment control logic and the blockage response strategy is optimized and adjusted based on the simulation analysis, and the liquid cooling system performance and anti-blocking strategy are optimized based on the adjustment effect. The present invention can effectively prevent and solve the blockage problem of the liquid cooling system, significantly improve the cooling efficiency and overall performance of the system, and ensure the reliable operation of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of liquid cooling system control, and in particular to a liquid cooling system self-optimization control method, system, computer equipment and medium. Background Art

[0002] Blockage in the liquid cooling system can seriously affect the cooling efficiency of the system and the stable operation of the equipment. Blockage usually occurs in narrow pipes or places where impurities are deposited, resulting in blockage of the coolant flow, which in turn affects the heat dissipation performance of various components in the system. In order to ensure the effective operation of the liquid cooling system, it is necessary to carefully monitor the flow situation and promptly detect and deal with blockage problems to avoid overheating of the system or even damage to sensitive electronic components.

[0003] Although the existing proposed blocking vector analysis method can determine whether there is a blockage based on the flow direction and size of the coolant in the system, it needs to rely on real-time and accurate monitoring of the flow state of the coolant in the liquid cooling system. However, in practical applications, there are the following challenges in accurately monitoring and analyzing the vector state of blockage: how to accurately collect coolant state data, how to process and analyze massive real-time data to achieve fast and accurate judgment, and how to automatically adjust system settings to avoid damage when abnormal flow is detected; among them, the more challenging problem is that blockage may be a dynamically changing process, and its development speed and impact range are not fixed, and may vary with time and system conditions. If the self-optimization mechanism of the liquid cooling system only relies on the setting of relevant static thresholds to predict and control abnormalities such as blockage, it will inevitably be unable to adapt to the actual changes in the coolant flow state, and cannot guarantee the reliability and rationality of the self-optimization control of the liquid cooling system, and thus it is difficult to ensure the normal and stable operation of the liquid cooling system. Summary of the invention

[0004] The purpose of the present invention is to provide a self-optimizing control method for a liquid cooling system. By comprehensively applying various cross-disciplinary knowledge such as fluid mechanics, sensing technology, data processing and intelligent algorithms, the flow vector data under different working conditions are dynamically analyzed, the changes in flow state are predicted and adapted to in real time, and the global temperature control and self-optimizing state prediction of the liquid cooling system are designed and optimized. The application defects that the flow state of the coolant in the existing liquid cooling system is difficult to monitor accurately in real time and effective self-optimizing control cannot be achieved. Based on the real-time and reliable monitoring of the flow state of the liquid cooling system, the blockage problem of the liquid cooling system is effectively prevented and solved, the true self-optimizing control of the liquid cooling system is achieved, and the cooling efficiency and overall performance of the system are significantly improved, thereby ensuring the reliable and stable operation of the system.

[0005] In order to achieve the above objectives, it is necessary to provide a liquid cooling system self-optimization control method, system, computer equipment and medium in response to the above technical problems.

[0006] In a first aspect, an embodiment of the present invention provides a method for self-optimizing control of a liquid cooling system, the method comprising the following steps:

[0007] Analyze the collected data on the flow status of the coolant in the liquid cooling system to obtain the current working status of the liquid cooling system;

[0008] When the current working condition of the liquid cooling system is abnormal, judging whether there is abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and taking the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed;

[0009] Perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain the corresponding node blockage prediction result;

[0010] Performing congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node to obtain the corresponding node congestion process prediction results, and generating a congestion warning signal according to the node congestion process prediction results;

[0011] In response to receiving the blocking warning signal, acquiring and executing a blocking response strategy according to a preset adjustment control logic, and performing simulation analysis on the execution of the blocking response strategy to optimize and adjust the blocking response strategy;

[0012] The adjustment effect analysis is performed based on the obtained system operation data after adjustment to optimize the liquid cooling system performance and anti-blocking strategy.

[0013] Furthermore, the step of judging whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node includes:

[0014] Compare and analyze the coolant status data with a pre-built historical status database to obtain corresponding comparison results;

[0015] If the comparison result is that the flow rate is abnormal and the pressure is normal, the flow rate is predicted according to the obtained liquid viscosity and liquid density, and the corresponding liquid temperature to obtain the corresponding theoretical flow rate, and when the theoretical flow rate is greater than the actual flow rate, the flow pattern is identified based on the flow fluctuation analysis and the pressure fluctuation analysis to obtain the corresponding coolant flow pattern;

[0016] If the coolant flow pattern suddenly changes from laminar flow to turbulent flow, an isolation forest algorithm is used to perform anomaly detection on the coolant state data;

[0017] When an abnormality is detected, a corresponding abnormal alarm is issued, and abnormal flow phenomena are classified in combination with the acquired pump working status data; the abnormal flow phenomena include system leakage, filter blockage, internal pipe blockage and valve closure.

[0018] Furthermore, the real-time monitoring data includes real-time flow data and real-time pressure data; the node congestion prediction result includes the congestion location and the corresponding congestion impact degree;

[0019] The step of performing flow vector analysis according to the real-time monitoring data to obtain a corresponding node blocking prediction result comprises:

[0020] The coolant fluid characteristics and pipeline specifications of the pipeline node to be analyzed are obtained, and a theoretical fluidity prediction is performed according to the coolant fluid characteristics, the pipeline specifications and a pre-built fluidity prediction model to obtain a corresponding unobstructed flow state; the pipeline specifications include pipeline size and pipeline shape; the unobstructed flow state includes flow direction and flow velocity;

[0021] Performing fluidity analysis on the real-time monitoring data by using a preset computational fluid dynamics analysis tool to generate a corresponding real-time flow vector diagram of the coolant, and identifying abnormal flow areas based on the real-time flow vector diagram of the coolant and the unobstructed flow state to obtain corresponding abnormal flow areas to be analyzed;

[0022] According to the obtained pipeline material data and roughness parameters in the abnormal flow area to be analyzed, the corresponding fluid resistance influence is evaluated, and when it is determined that there is an influence, the coolant flow vector distribution is obtained through vector analysis of streamline distribution and flow field distribution; the coolant flow vector distribution includes flow direction distribution and flow velocity distribution;

[0023] Identify the area with abnormally increased pressure drop according to the coolant flow vector distribution, and use the obtained area with abnormally increased pressure drop as the blocking position;

[0024] According to the pre-built recurrent neural network model, the influence of each blockage position on the cooling efficiency of the liquid cooling system is evaluated to obtain the corresponding blockage influence degree.

[0025] Furthermore, the step of performing congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node to obtain the corresponding node congestion process prediction results includes:

[0026] According to the blockage prediction results of each node and the preset computational fluid dynamics technology, a blockage simulation analysis is performed on each pipeline node to be analyzed to obtain the corresponding blockage impact analysis results;

[0027] When the blocking impact analysis result is a blocking impact flow vector, a blocking development prediction is performed according to a pre-constructed blocking development trend model to obtain a corresponding node blocking process prediction result.

[0028] Further, the adjusted system operation data includes adjusted flow data, adjusted pressure data, adjusted flow rate, adjusted temperature and adjusted heat load;

[0029] The step of analyzing the adjustment effect based on the obtained adjusted system operation data to optimize the liquid cooling system performance and anti-blocking strategy includes:

[0030] When the adjusted flow rate data and the adjusted pressure data are both lower than the corresponding expected values, a comprehensive analysis is performed on the current liquid cooling system according to the adjusted flow rate, the adjusted temperature and the adjusted heat load to determine whether there is a potential blocking problem;

[0031] When it is determined that there is a potential blocking problem, the blocking relief process is performed according to the preset adjustment rules and the preset adjustment times, and when the preset adjustment times are reached and the liquid cooling system parameters still do not reach the predetermined targets, the vibration and noise monitoring data are collected to analyze whether there is a mechanical performance failure;

[0032] If it is determined that there is no mechanical performance failure and the risk of leakage is eliminated, the anti-blocking strategy is optimized using a preset self-optimization strategy.

[0033] Furthermore, the method further comprises:

[0034] According to the pre-built system fault prediction model, the real-time performance data and operating environment data of the liquid cooling system are predicted and analyzed to obtain the prediction results for the preset future time period, and based on the prediction results, the corresponding liquid cooling system maintenance optimization plan is generated and executed; the prediction results include fault prediction results and system performance trends.

[0035] Furthermore, the method further comprises:

[0036] The system operation data after the execution of the liquid cooling system maintenance optimization plan is obtained, and the state of each system component is predicted based on the system operation data, a preset thermal model and a preset fluid dynamics model, and the system state optimization control is performed based on the corresponding component state prediction results.

[0037] In a second aspect, an embodiment of the present invention provides a liquid cooling system self-optimizing control system, the system comprising:

[0038] The system status analysis module is used to analyze the collected coolant flow status data in the liquid cooling system to obtain the current working status of the liquid cooling system;

[0039] The abnormal pipeline identification module is used to determine whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node when the current working condition of the liquid cooling system is abnormal, and to use the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed;

[0040] The blockage position prediction module is used to perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain the corresponding node blockage prediction result;

[0041] The congestion development prediction module is used to perform congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node, obtain the corresponding node congestion process prediction results, and generate a congestion warning signal according to the node congestion process prediction results;

[0042] A blocking response processing module, configured to, in response to receiving the blocking warning signal, acquire and execute a blocking response strategy according to a preset adjustment control logic, and perform simulation analysis on the execution of the blocking response strategy to optimize and adjust the blocking response strategy;

[0043] The strategy optimization and adjustment module is used to analyze the adjustment effect based on the obtained adjusted system operation data, so as to optimize the liquid cooling system performance and anti-blocking strategy.

[0044] In a third aspect, an embodiment of the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0045] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of the above method when executed by a processor.

[0046] The above-mentioned application provides a liquid cooling system self-optimization control method, system, computer equipment and medium. The method realizes the analysis of the collected coolant flow state data in the liquid cooling system to obtain the current working condition of the liquid cooling system. When the current working condition of the liquid cooling system is abnormal, it is judged whether there is abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and the key pipeline nodes with abnormal flow phenomenon are used as pipeline nodes to be analyzed. Then, the real-time monitoring data of each pipeline node to be analyzed are subjected to flow vector analysis to obtain the corresponding node blocking prediction result, and blocking simulation analysis and blocking process prediction are performed according to each node blocking prediction result to obtain the corresponding node blocking process prediction result, and a blocking warning signal is generated according to the node blocking process prediction result. In response to the reception of the blocking warning signal, a blocking response strategy is obtained and executed according to the preset adjustment control logic, and a simulation analysis is performed on the execution of the blocking response strategy to optimize and adjust the blocking response strategy, and an adjustment effect analysis is performed according to the obtained adjusted system operation data to optimize the liquid cooling system performance and anti-blocking strategy. Technical solution. Compared with the prior art, the self-optimizing control method for a liquid cooling system not only effectively prevents and solves system blockage problems by real-time monitoring and predicting the flow state in the liquid cooling system to ensure smooth flow of the coolant, but also reduces the risk of system overheating and avoids damage to sensitive electronic components by accurately locating potential blockage points and taking preventive measures in advance, thereby ensuring stable operation of system equipment; that is, the self-optimizing control method for a liquid cooling system provided by the present invention can effectively prevent and solve the blockage problem of the liquid cooling system based on real-time and reliable monitoring of the flow state of the liquid cooling system, thereby achieving true self-optimizing control of the liquid cooling system, significantly improving the cooling efficiency and overall performance of the system, and thereby ensuring reliable and stable operation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a flow chart of a self-optimizing control method for a liquid cooling system in an embodiment of the present invention;

[0048] Figure 2 2 is another schematic diagram of the process architecture of the self-optimization control method for a liquid cooling system according to an embodiment of the present invention;

[0049] Figure 3 is another process architecture schematic diagram of the liquid cooling system self-optimization control method in an embodiment of the present invention;

[0050] Figure 4 is a schematic diagram of the structure of a self-optimizing control system for a liquid cooling system according to an embodiment of the present invention;

[0051] Figure 5 is another structural schematic diagram of a self-optimizing control system for a liquid cooling system according to an embodiment of the present invention;

[0052] Figure 6is another structural schematic diagram of the self-optimizing control system of the liquid cooling system in an embodiment of the present invention;

[0053] Figure 7 It is a diagram of the internal structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to make the purpose, technical scheme and beneficial effects of the present application clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. Obviously, the embodiments described below are part of the embodiments of the present invention and are only used to illustrate the present invention, but are not used to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] The self-optimization control method for a liquid cooling system provided by the present invention can be understood as a method for independently optimizing and controlling a liquid cooling system by comprehensively applying a variety of cross-disciplinary knowledge such as fluid mechanics, sensing technology, data processing, and intelligent algorithms to dynamically analyze flow vector data under different working conditions, predict and adapt to flow state changes in real time, and design and optimize the global temperature control and self-optimization state prediction of the liquid cooling system. The method is based on real-time and reliable monitoring of the flow state of the liquid cooling system, effectively preventing and solving the problem of blockage of the liquid cooling system, realizing true self-optimization control of the liquid cooling system, significantly improving the cooling efficiency and overall performance of the system, and thereby ensuring reliable and stable operation of the system. The following embodiments will provide a detailed description of the self-optimization control method for a liquid cooling system of the present invention.

[0056] In one embodiment, Figure 1 As shown, a liquid cooling system self-optimization control method is provided, comprising the following steps:

[0057] S11. Analyze the collected coolant flow state data in the liquid cooling system to obtain the current working condition of the liquid cooling system; wherein, the coolant flow state data includes but is not limited to the coolant flow in the liquid cooling system measured by a high-sensitivity flow sensor, the pressure measured by a pressure sensor, the temperature measured by a temperature sensor, the flow direction monitored by a vortex flowmeter, the flow velocity monitored by a flow velocity sensor, and the turbulence intensity measured by a turbulence measurement device, etc. The current working condition of the liquid cooling system obtained by the corresponding analysis mainly includes a normal state and a suspected abnormal state, and the suspected abnormal state includes suspected blockage, suspected leakage, suspected erosion, incorrect valve or pipeline position, and insufficient adaptability to the working environment, etc.

[0058] Specifically, the process of analyzing the coolant flow status data collected in real time to obtain the current working condition of the liquid cooling system is as follows: if an unexpected drop in flow is detected, the pressure is judged to be normal, and when the pressure is confirmed to be abnormal and the pressure difference increases, it is determined that there is a blockage in the liquid cooling system, that is, the current working condition of the liquid cooling system is suspected blockage; if the flow direction obtained by real-time monitoring of the vortex flowmeter changes, it is determined that there may be problems such as valve position error or pipeline configuration, that is, the current working condition of the liquid cooling system is valve or pipeline position error; if the monitored coolant temperature rises abnormally, it is inferred that the flow rate is too slow and the heat exchange is insufficient. At the same time, if the local flow rate is too fast, it is judged that there is an erosion risk in the area, that is, the current working condition of the liquid cooling system is suspected erosion; if the monitored turbulence intensity exceeds the predetermined range, the system parameters are adjusted to reduce the impact of turbulence on the system, and at the same time, the transient changes of the flow are analyzed based on the flow change rate data to determine whether the system is leaking or whether the performance of the pump and valve is abnormal, and then determine whether there is a suspected leakage problem; at the same time, the work environment adaptability analysis is performed based on the obtained working environment data. If the ambient temperature, pressure or chemical properties affect the sensor performance, the sensor or system settings are adjusted to adapt to environmental changes.

[0059] In actual applications, if the flow sensor detects that the coolant flow rate has dropped from 5 liters per minute to 3 liters per minute, it means that the flow rate has decreased by 40%; the pressure sensor shows that the pressure of the cooling system is 2 bar under normal circumstances, but the current measured pressure is 3 bar, which indicates that the pressure difference has increased by 1 bar, which may be due to some kind of blockage in the system; through the real-time monitoring of the flow direction by the vortex flowmeter, it is found that the coolant should flow from pipe A to pipe B, but in fact it is flowing from pipe A to pipe C, which may be caused by the wrong valve position; if the normal operating temperature range of the coolant is 30°C to 35°C, but the temperature sensor shows that the current temperature is 40°C, this may indicate that the heat exchange is insufficient due to the slow flow rate; The flow rate measured by the flow sensor in a specific area suddenly increased from 1 liter per minute to 2 liters per minute, which may mean that there is a risk of erosion in the area due to excessive flow rate; it is assumed that the upper limit of turbulence intensity is 1500rpm, but real-time monitoring shows that the turbulence intensity is 1800rpm, which exceeds the predetermined range and requires adjustment of system parameters to reduce turbulence intensity; flow rate change analysis shows that the flow rate suddenly dropped by 20% in a short period of time, which may be due to a leak in the system or abnormal performance of the pump and valve; the working environment temperature increased by 10°C from 25°C to 35°C, and the calibration data of the pressure sensor showed that its accuracy decreased by 5%, which requires recalibration of the sensor or adjustment of the system settings to adapt to the ambient temperature change.

[0060] S12. When the current working condition of the liquid cooling system is abnormal, determine whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and use the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed; wherein, the abnormal state can be understood as the suspected abnormal state given above, that is, when the liquid cooling system may have suspected abnormal risks such as blockage, leakage and insufficient adaptability to the working environment after the above steps are analyzed, it is necessary to configure flow sensors, pressure sensors and temperature sensors at the key pipeline nodes of the liquid cooling system to monitor the flow, pressure and temperature data of the coolant at each key pipeline node in real time, so as to monitor in real time whether the node position has abnormal flow phenomena such as a sudden drop in flow rate or an abnormal increase in pressure; it should be noted that the key pipeline nodes in this embodiment include but are not limited to the positions of pipeline inlet, pipeline outlet and pipeline branch, which can be set according to actual application requirements;

[0061] In order to ensure the efficiency and accuracy of abnormal flow phenomenon analysis, this embodiment preferably adopts flow vector analysis method to analyze and monitor system blockage phenomenon; specifically, the step of judging whether there is abnormal flow phenomenon according to the coolant state data at each preset key pipeline node includes:

[0062] The coolant status data is compared and analyzed with a pre-constructed historical status database to obtain a corresponding comparison result; wherein the historical status database can be understood as a benchmark status database that stores flow, pressure and temperature data during historical normal operation, and provides a reference basis for subsequent comparison and analysis of the coolant status data obtained by real-time monitoring with the historical normal status data.

[0063] If the comparison result is that the flow rate is abnormal and the pressure is normal, the flow rate is predicted according to the obtained liquid viscosity and liquid density, and the corresponding liquid temperature to obtain the corresponding theoretical flow rate, and when the theoretical flow rate is greater than the actual flow rate, the flow pattern is identified based on the flow fluctuation analysis and the pressure fluctuation analysis to obtain the corresponding coolant flow pattern; that is, by developing a data analysis algorithm to compare and analyze the real-time monitoring data with the historical data, if the current flow data shows a downward trend, and the pressure data has no obvious change, it is necessary to further analyze the possible reasons; by introducing the real-time monitoring data of liquid viscosity and liquid density to determine whether the flow rate drop is related to the change of coolant properties, if the increase in coolant temperature causes the viscosity to decrease, the flow rate may be affected; at the same time, the algorithm is used to predict the theoretical flow rate based on the current temperature, viscosity and density data and compare it with the current actual flow rate. If the theoretical flow rate is significantly different from the actual flow rate, the possibility of determining it as an abnormal flow phenomenon increases, and it is necessary to further analyze the flow fluctuation and pressure fluctuation to determine the flow pattern of the coolant in the pipeline and obtain the corresponding coolant flow pattern.

[0064] If the coolant flow pattern suddenly changes from laminar flow to turbulent flow, the isolation forest algorithm is used to detect anomalies in the coolant status data; that is, the sudden change of the flow pattern from laminar flow to turbulent flow may indicate blockage or bubble generation. In order to further confirm the anomaly, the isolation forest algorithm is needed to automatically identify abnormal patterns in the flow, pressure and temperature data. The system will issue corresponding abnormal alarms based on the detected abnormal patterns and judge potential leakage or blockage events.

[0065] When an abnormality is detected, a corresponding abnormal alarm is issued, and the abnormal flow phenomenon is classified in combination with the acquired pump working status data; the abnormal flow phenomenon includes system leakage, filter blockage, internal pipe blockage and valve closure; wherein, the pump working status data includes but is not limited to rotation speed and power consumption, and the corresponding abnormal flow phenomenon classification is understood to be a further refined analysis of the cause of flow abnormality based on the pump working status data: if the pump speed is normal but the flow rate decreases, it may be considered that the system is leaking or the filter is clogged; if the pressure data shows an abnormal upward trend, it is necessary to combine the liquid level monitoring data to determine whether the liquid flow is obstructed. If the pressure rises and the liquid level is normal or increases, it indicates that there is a blockage inside the pipeline or the valve is closed.

[0066] In actual applications, if under normal operating conditions, the flow sensor shows a flow rate of 5L / min, the pressure sensor shows a pressure of 3bar, and the temperature sensor shows a coolant temperature of 20°C, these data will be stored in the historical status database; when the real-time monitoring data is compared and analyzed with the historical data using the developed data analysis algorithm, if the current flow data shows that the flow rate has dropped to 0L / min and the pressure data has not changed significantly and remains at around 3bar, this may indicate that there is a problem that requires further analysis. Introducing real-time monitoring data of viscosity and density can help determine whether the drop in flow rate is related to a change in the coolant properties. If the coolant temperature increases from 20°C to 30°C, causing the viscosity to drop from 0cP to 8cP, this may affect the flow rate. At the same time, the algorithm is used to predict the theoretical flow rate based on the current temperature, viscosity, and density data and compare it with the actual flow rate. If the theoretical flow rate is 5L / min and the actual flow rate is only 0 L / min, then the difference is large, indicating that the possibility of abnormal flow phenomenon has increased, and flow pattern recognition is further implemented; for example, by analyzing the flow fluctuation, if a sudden drop in flow is found, it may indicate a change from laminar flow to turbulent flow, which may indicate blockage or bubble generation, and the isolated forest algorithm can be used to automatically identify abnormal patterns in flow, pressure and temperature data. If the system suddenly detects that the flow rate drops to 8L / min and the pressure rises to 4bar, the system will issue an alarm; in addition, combined with the pump's operating status data, if the pump speed is 3000rpm and the power consumption is 2kW, but the flow rate is reduced from 5L / min to 0L / min, then this indicates that there may be a leak in the system or the filter is clogged; if the pressure data shows an abnormal increase from 3bar to 4bar, and the liquid level monitoring data shows that the liquid level has risen by 5cm from the normal mark, this may indicate that there may be a blockage or valve closure inside the pipeline.

[0067] The above method and steps can be used to make a preliminary judgment on abnormal flow phenomena in the liquid cooling system, so as to identify key pipeline nodes that may have abnormalities, and facilitate subsequent detailed flow vector analysis to accurately determine the abnormal blockage location of the liquid cooling system.

[0068] S13. Perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain corresponding node blockage prediction results; wherein, the real-time monitoring data can be understood as node status data monitored in real time by sensors installed at key pipeline nodes of the liquid cooling system, preferably including real-time flow data and real-time pressure data; correspondingly, the node blockage prediction result can be understood as generating a real-time flow vector diagram of the coolant based on the real-time monitoring data, and performing blockage vector analysis based on the real-time flow vector diagram of the coolant to obtain the node blockage prediction result, including the blockage position and the corresponding degree of blockage impact.

[0069] Specifically, the step of performing flow vector analysis according to the real-time monitoring data to obtain a corresponding node congestion prediction result includes:

[0070] The coolant fluid characteristics and pipeline specifications of the pipeline node to be analyzed are obtained, and theoretical fluidity prediction is performed based on the coolant fluid characteristics, the pipeline specifications and a pre-constructed fluidity prediction model to obtain a corresponding unobstructed flow state; wherein the pipeline specifications include pipeline dimensions and pipeline shapes; the fluidity prediction model can be understood as a model with functions such as predicting the flow direction and flow velocity in an unobstructed condition based on relevant theoretical analysis according to data such as fluid characteristics, pipeline dimensions and shapes, and is not specifically limited here; correspondingly, the unobstructed flow state predicted based on the fluidity prediction model includes the flow direction and flow velocity.

[0071] The real-time monitoring data is subjected to fluidity analysis by means of a preset computational fluid dynamics analysis tool to generate a corresponding real-time flow vector diagram of the coolant, and abnormal flow areas are identified based on the real-time flow vector diagram of the coolant and the unobstructed flow state to obtain the corresponding abnormal flow area to be analyzed; wherein, the preset computational fluid dynamics analysis tool can be selected and determined according to actual application requirements and is not limited at this time; the process of obtaining the abnormal flow area to be analyzed can be understood as generating a corresponding real-time flow vector diagram based on the actual flow and pressure data in the real-time monitoring data by means of computational fluid dynamics software or analysis tools, and comparing and analyzing it with the expected normal condition (unobstructed flow state), and identifying the area with the state deviation as the abnormal flow area to be analyzed.

[0072] According to the obtained pipeline material data and roughness parameters in the abnormal flow area to be analyzed, the corresponding fluid resistance influence is evaluated, and when it is determined that there is an influence, the coolant flow vector distribution is obtained through vector analysis of streamline distribution and flow field distribution; wherein, the coolant flow vector distribution preferably includes flow direction distribution and flow velocity distribution, and the specific acquisition process can be understood as: obtaining material data and roughness parameters of the abnormal flow area to be analyzed, and evaluating their influence on fluid resistance in actual application, and determining whether these factors are related to abnormal flow phenomena; at the same time, external environmental data including temperature and pressure can also be recorded and analyzed to determine their potential influence on fluid characteristics and pipeline performance; after evaluating and determining that the material data and roughness parameters have an influence on fluid resistance, further through detailed vector analysis of streamline and flow field distribution, the specific direction and velocity distribution of the flow are determined, so as to provide more accurate data support for the subsequent positioning of the blockage point.

[0073] The area with abnormally increased pressure drop is identified according to the coolant flow vector distribution, and the obtained area with abnormally increased pressure drop is used as the blockage position; wherein, the process of determining the blockage position can be understood as: according to the coolant flow vector distribution, the distribution of pressure data along the pipeline is obtained, and the pressure difference calculation is applied to identify the area with abnormally increased pressure drop, and use it as a potential blockage point, that is, to obtain each blockage position.

[0074] According to the pre-constructed recurrent neural network model, the influence of each blockage position on the cooling efficiency of the liquid cooling system is evaluated to obtain the corresponding blockage influence degree; wherein, the recurrent neural network model can be understood as a model that is pre-trained based on historically collected data such as flow changes, pressure drops, flow vectors, cooling efficiency and external environmental impact information, and can predict and evaluate the influence of blockage points on the cooling efficiency of the liquid cooling system. The specific model structure can be selected according to actual application requirements and is not specifically limited here.

[0075] In actual applications, in a normally operating liquid cooling system, the flow sensor installed at the key pipeline node may show a flow rate of 5L / min, while the pressure sensor shows a pressure of 8MPa. If at a certain moment, the flow sensor reading suddenly drops to 8L / min, and the pressure sensor reading rises to 0MPa, this may indicate a blockage. The moving average algorithm can also be used to analyze the flow data in the last 10 minutes. If the current flow reading is lower than twice the standard deviation of the moving average, an abnormality is reported. If the moving average of the flow in the past 10 minutes is 4L / min and the standard deviation is 1L / min, any reading lower than 2L / min is considered abnormal. In the flow prediction model, if the pipe diameter is 50mm, the length is 10m, and there is no blockage in the pipe, the Bernoulli equation is used to predict that the ideal flow rate under a pressure of 8MPa should be 0L / min, while the actual monitored flow rate is 8L / min, which is much lower than the prediction, further verifying the assumption that there is a blockage. Computational fluid dynamics software is then used to simulate the fluid flow vector diagram under normal conditions, showing a uniform streamline distribution. In actual situations, if the streamline is observed to be compact or broken in a certain pipe section, this may indicate that there is a blockage there, that is, the abnormal flow area to be analyzed has been found. Further analysis of the pipeline material stainless steel and the surface roughness of 0.045mm, the calculated friction coefficient of 0.028 can be used to determine the flow resistance. If the roughness of the abnormal flow area to be analyzed is found to be higher than the normal value, this may be the cause of the blockage. At the same time, the external ambient temperature recorded by the temperature sensor may rise from 20°C to 30°C, which will also affect the viscosity of the fluid and may affect the flow characteristics. If the viscosity of the coolant decreases at high temperatures, the flow rate should theoretically increase, but if the flow rate actually decreases, this will further prove that there is a blockage in the system; when determining the presence of resistance, vector analysis of streamlines and flow fields is performed. It can be found that the flow velocity decreases near the blocking point, and the flow velocity decreases from 5m / s to 2m / s, indicating that this is the location where the blockage occurs; if the pressure at 5 meters away from the blocking point is 9MPa, and the pressure at 1 meter upstream of the blocking point is 1MPa, it shows that the pressure drop in this area is abnormally increased, revealing the existence of blockage; in addition, by collecting flow, pressure, temperature and other data, and using the recurrent neural network algorithm to train the model, the impact of the blocking point on the cooling efficiency of the liquid cooling system is evaluated, and the model predicts that if the blocking problem is not solved, the cooling efficiency of the system may drop by 15%, and the impact on the liquid cooling system is large.

[0076] S14, performing a blocking simulation analysis and a blocking process prediction according to the blocking prediction results of each node to obtain a corresponding node blocking process prediction result, and generating a blocking warning signal according to the node blocking process prediction result; wherein, the blocking simulation analysis can be understood as using computational fluid dynamics to simulate the flow vector of the coolant under blocking conditions with different blocking influence degrees; the blocking process prediction can be understood as performing a predictive analysis on the time series state data of each pipeline node to be analyzed based on a neural network model to infer the coolant flow state change and the blocking development trend; specifically, the step of performing a blocking simulation analysis and a blocking process prediction according to the blocking prediction results of each node to obtain the corresponding node blocking process prediction result includes:

[0077] According to the blockage prediction results of each node and the preset computational fluid dynamics technology, a blockage simulation analysis is performed on each pipeline node to be analyzed to obtain the corresponding blockage impact analysis results; wherein, the blockage impact analysis results can be understood as the results of simulating and analyzing the flow vector data of the coolant under different blockage impact levels according to the blockage position and blockage impact level of the key pipeline nodes of the liquid cooling system. The specific acquisition process includes: taking into account the flow velocity (real-time flow velocity of the coolant in the pipeline) and pressure difference data (pressure difference changes indicate a possible increase in flow resistance) of the key nodes collected through flow velocity and pressure sensors, and based on the pipeline size and physical property information, using computational fluid dynamics CFD simulation technology to simulate the flow of coolant under blockage conditions of different severity, analyze the impact of blockage on the flow vector, and obtain an analytical conclusion on whether the blockage will affect the flow vector of the key pipeline node.

[0078] When the blocking impact analysis result is a blocking impact flow vector, a blocking development prediction is performed based on a pre-constructed blocking development trend model to obtain a corresponding node blocking process prediction result; wherein, the blocking development trend model can be understood as a network model that is based on a long short-term memory network algorithm, and is trained based on historical flow vector data collected in actual applications and real-time monitoring data collected by sensors such as flow rate, pressure, and temperature, and is able to predict changes in the flow state of the coolant and the development trend of blocking. The model can be used to analyze the development rate and degree of the blocking and predict the impact of the blocking situation on the system; if an impending blockage is predicted, the system automatically generates an early warning signal and performs dynamic analysis through preset data analysis rules to ensure the accuracy of the early warning signal; based on the dynamic analysis results, the sampling frequency is dynamically adjusted to capture more accurate flow vector data, and the data is processed, including cleaning, feature extraction, and noise reduction, and the specific characteristics of the blockage, including type, location, and degree, are judged by further analyzing the processed data. The system automatically initiates preventive measures based on this, and adjusts the flow rate or increases the cleaning frequency to prevent blockage from forming.

[0079] In practical applications, if there is a pipeline node to be analyzed in a liquid cooling system, the pipeline diameter is 50 mm, and a flow sensor and a pressure sensor are installed on the node, where the flow sensor shows that the coolant flow rate at this point is 5 m / s, and the pressure sensor shows a pressure difference of 200 Pascals; using computational fluid dynamics CFD simulation technology, the flow of this point under different blockage conditions can be simulated. When 25% of the pipeline diameter is blocked in the simulation, the CFD simulation may show that the flow rate drops to 2 m / s and the pressure difference increases to 250 Pascals. At the same time, based on the flow rate and pressure difference data collected from the sensor every minute by the real-time monitoring system, the long short-term memory network model is trained and processed. Based on the analysis of historical data by the model, it can be predicted that under the current situation, if no measures are taken, 100% blockage may occur within the next 48 hours; when the model detects a potential blockage warning, the system will automatically generate a warning signal. For example, an early warning signal may be triggered when the blockage reaches 50%, and the accuracy of this trend is confirmed by preset data analysis. The preset data analysis process may decide to increase the sampling frequency from once a minute to once every 30 seconds in order to more accurately capture the changes in the flow vector; when processing this data, if the data cleaning link removes 10% of the noise data points, the feature extraction link identifies a trend of an increase in the pressure difference of 10 Pascals per hour; in addition, further analysis determines that the blockage is caused by sediment accumulation in the pipe, which is located approximately 10 cm downstream from the flow sensor, and the blockage is 50%. Based on this, the system automatically initiates preventive measures, adjusts the flow rate to 8 meters per second, and increases the cleaning frequency to once every 24 hours to prevent blockage from forming and developing.

[0080] Based on the complex interaction between the nonlinear behavior of the fluid in the liquid cooling system (including turbulence, sudden changes in flow velocity) and pipeline blockage, the above prediction of the liquid cooling system blockage development rate and the issuance of early warning signals can effectively prevent or slow down the development of blockage. However, considering how to define the blockage development rate to accurately describe the relationship between the nonlinear flow characteristics and the blockage development rate, the accuracy of the node blockage process prediction results can be effectively improved. Preferably, in this embodiment, sensors are arranged to collect flow rate change data, suspended particle diameter and other related parameters, and a blockage development rate prediction formula is defined through data analysis:

[0081]

[0082] Where k is an adjustment factor used to match the response of the actual system; V max 、V min and V avgare the maximum, minimum and average flow rates during the monitoring period; α and β are exponents determined based on the experience of system characteristics to adjust the influence of flow rate changes and particle diameter on the blockage development rate, D p is the average diameter of suspended particles in the fluid, which is a key factor in the occurrence of blockage; based on experimental data and historical records, the values ​​of k, α and β are adjusted to ensure that the FND formula can accurately reflect the actual operating status of the system.

[0083] After verifying the accuracy of the FND formula under actual operating conditions and correcting the model through experiments or simulation tests, the adjusted FND formula is integrated into the real-time monitoring system, and the blockage development rate is calculated using real-time data to issue early warning signals in a timely manner; at the same time, the system operating parameters, including flow rate and temperature, are dynamically adjusted according to the calculation results of the FND formula to prevent or slow down the development of blockage. For example, the flow rate data of a pipe in the liquid cooling system is: Maximum flow rate V max =2.5m / s, minimum flow velocity V min =0.5m / s, average flow velocity V avg =1.5m / s; In addition, the average diameter of suspended particles in the fluid has been measured to be D p =0.02m; According to the defined formula for predicting the blockage development rate FND, when the parameters k=0.6, α=1.8, β=0.5 have been determined, the specific values ​​are substituted into the formula to calculate the quantitative value FND of the nonlinear flow characteristics, FND=0.6*(1.0) 1.8 *0.1414≈0.126, that is, the calculated FND value is about 0.126, which indicates the influence of nonlinear flow characteristics on the blockage development rate under the current flow conditions. By continuously monitoring and calculating FND, the development trend of pipeline blockage can be predicted in time, and corresponding measures can be taken to prevent or slow down the formation of blockage, thereby ensuring the normal operation of the data center liquid cooling system.

[0084] S15. In response to receiving the blockage warning signal, a blockage response strategy is acquired and executed according to a preset adjustment control logic, and a simulation analysis is performed on the execution of the blockage response strategy to optimize and adjust the blockage response strategy; wherein, the preset adjustment control logic may be a system control logic pre-set according to actual application requirements, which is not specifically limited here; the corresponding blockage response strategy preferably includes increasing the operating frequency of the pump, adjusting the pump speed, or adjusting the degree of opening of the valve.

[0085] Considering that the blocking response strategy preset in the system is not necessarily applicable to all coolant flow state scenarios, in order to further improve the application effect of the blocking response strategy in practical applications, and thus improve the efficiency of preventing and solving blocking problems, this embodiment preferably uses fluid dynamics simulation analysis of real-time data to optimize the blocking warning response and maintenance strategy of the liquid cooling system; specifically, the process of simulating and analyzing the execution of the blocking response strategy to optimize and adjust the blocking response strategy includes:

[0086] The system receives a blocking warning signal. If it is determined that there is a risk of blocking, it will obtain and execute the blocking response strategy based on the abnormal characteristics and preset adjustment control logic. For example, the operating frequency or pump speed of the pump is adjusted through the liquid cooling system. If the flow rate is lower than the set threshold, the pump speed is increased to increase the flow rate in order to flush out potential blockages; adjust the valve opening degree. If the temperature rise in a specific area is detected, the related valve is opened to increase the coolant flow rate and heat dissipation capacity in the area. At the same time, in order to ensure the actual application effect of the blocking response strategy, fluid dynamics simulation software is used to simulate the adjusted flow conditions. If the simulation results show that the flow is improved, continue to monitor, otherwise further fine-tune the pump frequency and valve opening degree; in addition, the automatic response of the liquid cooling system is optimized by adjusting the parameters in the control logic. If the sensor data shows that the pressure is stable, the current setting is maintained, otherwise, the adjustment is continued; monitor the coolant characteristics. If the viscosity change is detected to affect the flow performance, the liquid cooling system is instructed to automatically add an appropriate amount of additives or replace the coolant, and a real-time data analysis tool is used to monitor and analyze the data fed back by a specific sensor. If the feedback data of a specific sensor is found to be inconsistent with expectations, corrections are made or maintenance inspections are arranged. Feedback the adjusted liquid cooling system status through the user interface. If the liquid cooling system shows that all parameters have returned to normal, the monitoring is maintained. If a new warning is displayed, the above adjustment process is repeated. Preferably, the present embodiment analyzes long-term data records to monitor the recurrence of abnormal patterns, and when found, adjusts the maintenance strategy and further plans systematic preventive maintenance to reduce future blockage risks.

[0087] In actual applications, if during the monitoring of the liquid cooling system, the sensor data shows that the coolant flow rate suddenly drops from the normal value of 5 L / min to 8 L / min, and the pressure increases from the normal 100 kPa to 135 kPa, after data analysis and determination of potential blockage risks, in order to flush potential blockages, the liquid cooling system will increase the pump speed from 2000 RPM to 2500 RPM to increase the flow rate; if the temperature in a specific area is monitored to rise from the normal 30°C to 35°C, the system will increase the opening degree of the connected valve from 50% to 70% to improve cooling efficiency. At the same time, the adjusted flow conditions are simulated using fluid dynamics simulation software (such as ANSYS Fluent, etc.). If the simulation results show that the flow rate returns to 4 L / min and the pressure drops to 110 kPa, which is close to the normal level, when the pump speed is 2500 RPM and the valve is opened 70%, the system will continue with such monitoring settings; in terms of maintaining the coolant characteristics, if the data measured by the viscometer shows that the viscosity of the coolant has increased from the general 0 mPa·s to 5 mPa·s, the system will automatically add 100 mL of additives to the coolant to adjust the viscosity; if the real-time data analysis tool indicates that the reading of a temperature sensor deviates from the average value of other sensors by more than 5°C, the system will consider this to be a sensor failure or calibration problem and mark it as requiring maintenance inspection. If the user interface shows that all monitoring parameters such as flow and pressure have returned to the normal range after adjustment, the system will continue to operate under these settings. If a new warning appears after the adjustment, such as the flow rate drops to 7 L / min again, the system will repeat the above adjustment process. Additionally, when analyzing long-term data records, if it is found that the system experiences a slow pressure increase in the last week of each month, with an average pressure increase of 10 kPa, then the maintenance strategy of the liquid cooling system may be adjusted, such as performing preventive cleaning in the third week of each month to reduce the risk of blockage.

[0088] S16. Performing adjustment effect analysis based on the obtained adjusted system operation data to optimize the performance and anti-blocking strategy of the liquid cooling system; wherein the adjusted system operation data preferably includes adjusted flow data, adjusted pressure data, adjusted flow rate, adjusted temperature and adjusted heat load; correspondingly, the adjustment effect analysis can be understood as analyzing the deviation of the flow and pressure data in real time after obtaining the adjusted flow and pressure data to determine whether the adjustment is effective; specifically, the step of performing adjustment effect analysis based on the obtained adjusted system operation data to optimize the performance and anti-blocking strategy of the liquid cooling system includes:

[0089] When the adjusted flow data and the adjusted pressure data are both lower than the corresponding expected values, a comprehensive analysis is performed on the current liquid cooling system based on the adjusted flow rate, the adjusted temperature and the adjusted heat load to determine whether there is a potential blockage problem; wherein, the process of comprehensive analysis to determine whether there is a potential blockage problem can be understood as if the flow and pressure values ​​are lower than expected, a preset software analysis tool is used to perform a comprehensive analysis on relevant data such as flow rate, temperature, heat load, etc. to determine whether there is a potential blockage problem, and if the software analysis shows that the outlet temperature is abnormally increased and the flow rate is reduced, the resistance of the liquid cooling system is further checked, especially whether the radiator, filter, valve and other components show signs of blockage.

[0090] When it is determined that there is a potential blocking problem, blocking relief processing is performed according to the preset adjustment rules and the preset adjustment times, and when the preset adjustment times are reached and the liquid cooling system parameters still do not reach the predetermined target, vibration and noise monitoring data are collected to analyze whether there is a mechanical performance failure; wherein the preset adjustment rules can be set according to actual application requirements in principle. In order to improve the adjustment optimization effect, this embodiment preferably performs the adjustment operation according to the following rules:

[0091] Adjust the pump speed according to the pump performance data, observe the changes in flow and pressure data after adjusting the pump speed, and determine whether the pump speed adjustment is effective in alleviating potential blockages; if the flow increases but the pressure does not change much after adjusting the pump speed, check the pipe diameter and length data to determine whether fluid dynamics problems may be caused by inappropriate pipe size; if the resistance of the liquid cooling system is reduced, but the flow and pressure data still do not reach the target value, adjust the pump and pipe parameters, and use the data recording function to continuously monitor the changes in temperature and pressure; through continuous adjustment and monitoring, if the pressure and flow data of the liquid cooling system reach the predetermined target and there is no abnormal temperature change, it is judged that the precursor to blockage has been successfully eliminated. Conversely, if the liquid cooling system parameters still do not reach the predetermined target after the preset number of adjustments, vibration and noise monitoring data can be used to further analyze whether there are other mechanical failures that cause performance degradation.

[0092] If it is determined that there is no mechanical performance failure and the risk of leakage is eliminated, the preset self-optimization strategy is used to optimize the anti-blocking strategy; that is, if no abnormality is found through vibration and noise monitoring, the probability of leakage is eliminated through leakage detection, and advanced self-optimization strategies and algorithms (such as genetic algorithms, etc.) are used to try to find new optimization solutions.

[0093] In practical applications, the ideal flow rate of a liquid cooling system is 200L / min and the pressure is 5bar. If the current flow rate detected by the sensor is only 180L / min and the pressure is 5bar, both of which are lower than the expected values, it indicates that there may be potential problems. By accessing the data through the preset software analysis tool, it is calculated that the outlet temperature of the radiator is 5°C higher than the design temperature, and the flow rate is 10% lower than expected, which may indicate that there is a blockage inside the radiator. Further inspection of the radiator found that the pressure drop of the filter increased by 2bar, which is significantly greater than the normal value of 5bar, which strengthens the judgment of blockage. Subsequently, the pump speed is adjusted from 1800rpm to 2000rpm, and the data is re-monitored. If the adjusted flow rate increases to 190L / min, but the pressure only rises to 7bar, it shows that the pump speed adjustment has a certain effect on alleviating the blockage, but the problem has not been completely solved. At the same time, checking the pipeline data, it is found that the pipeline diameter is 50mm and the length is 10m, which matches the design, and the possibility of improper pipeline size is also ruled out. After continuously monitoring the changes in temperature and pressure, the data shows that after 30 minutes of continuous operation, the flow rate has steadily increased to 195L / min and the pressure has reached 8bar, indicating that the system has begun to gradually return to normal, but has not yet reached the target value. In addition, after multiple adjustments, if the liquid cooling system parameters are still not satisfied, for example, vibration monitoring shows that the vibration frequency of the pump has changed from the normal 50Hz to 55Hz, which may indicate wear or imbalance of mechanical parts; if vibration and noise monitoring show that everything is normal, then perform leakage detection. If there is still no sign of leakage, use advanced self-optimization strategies (such as implementing genetic algorithms) to optimize pump and pipeline parameters. Assuming that after the advanced self-optimization strategy algorithm runs, it is recommended to adjust the pump speed to 2100rpm and re-insulate the local pipeline. After these adjustments are implemented, the data finally shows that the flow rate reaches 205L / min and the pressure is 2bar, which exceeds the design target. The system runs stably for more than 1 hour without abnormal temperature rise, indicating that the potential blockage problem has been successfully solved.

[0094] The present embodiment provides an analysis of the collected coolant flow state data in the liquid cooling system to obtain the current working state of the liquid cooling system. When the current working state of the liquid cooling system is abnormal, it is determined whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and the key pipeline node with abnormal flow phenomenon is used as the pipeline node to be analyzed. The real-time monitoring data of each pipeline node to be analyzed is subjected to flow vector analysis to obtain the corresponding node blocking prediction result, and blocking simulation analysis and blocking process prediction are performed according to the blocking prediction result of each node to obtain the corresponding node blocking process prediction result. After a blocking warning signal is generated according to the node blocking process prediction result, in response to the blocking prediction result, The technical solution not only monitors and predicts the flow state in the liquid cooling system in real time to effectively prevent and solve system blockage problems and ensure smooth flow of coolant, but also accurately locates potential blockage points and takes preventive measures in advance to reduce the risk of system overheating, avoid damage to sensitive electronic components, ensure stable operation of equipment, and thus provide reliable guarantee for the normal and stable operation of the liquid cooling system.

[0095] In addition, in order to further improve the ability of the liquid cooling system to cope with actual application failures and enhance the system's self-optimization control capabilities, this embodiment preferably uses a system failure prediction model to combine system performance data and operating environment data obtained through real-time monitoring to perform failure prediction analysis on the liquid cooling system, so as to timely perceive and maintain system failures and ensure the stable operation of the liquid cooling system as much as possible. Figure 2 As shown, the method also includes:

[0096] S17. According to the pre-constructed system fault prediction model, the real-time performance data and operating environment data of the liquid cooling system are predicted and analyzed to obtain the prediction results for the preset future time period, and based on the prediction results, a corresponding liquid cooling system maintenance optimization plan is generated and executed; the prediction results include fault prediction results and system performance trends; wherein, the system fault prediction model can be understood as a fault prediction model established based on the historical fault data of the liquid cooling system, which is used to predict the faults and system performance changes of the liquid cooling system within a preset time period in the future. Based on the preset results of the fault prediction model, a response strategy can be formulated first, and then based on the response strategy, a corresponding liquid cooling system maintenance optimization plan including a blockage prevention strategy is generated.

[0097] The process of obtaining the specific liquid cooling system maintenance optimization plan is as follows:

[0098] Through sensors or related data acquisition systems, real-time performance data of the liquid cooling system is obtained, including key parameters such as flow, pressure, temperature, heat load and energy consumption, as well as coolant composition, to provide basic data for the training and construction of the system fault prediction model; at the same time, according to maintenance records and fault history data, the fault type, frequency, and location are classified, and data preprocessing technology is used to clean and normalize the data to facilitate model analysis and identification of fault modes. The processed data is then fused with environmental factor data including ambient temperature and humidity, as well as data such as the usage time and workload of the liquid cooling system to enhance the model's adaptability to changes in external conditions and prediction accuracy. Equipment information including equipment specifications, lifespan, and manufacturer information is used to build equipment feature profiles, and the characteristics and performance limitations of different equipment are added to the model; machine learning algorithms, including decision trees or random forests, are applied to train the system fault prediction model. If the model performs well on the test set, it is determined that the model can accurately predict future failures and performance changes; at the same time, through sensitivity analysis, the parameters that have the greatest impact on the model prediction results are identified, feature selection is performed, and the model is optimized to pay more attention to these key parameters, improve the accuracy and reliability of the prediction, and the model is evaluated using a cross-validation method to obtain prediction accuracy and computational efficiency indicators, and the model is adjusted and optimized based on these indicators to achieve better performance; in addition, the system fault prediction model used in this embodiment is also regularly trained and updated through a preset data update strategy, so that the model can capture the latest trends in system performance and environmental conditions, and ensure the timeliness of the prediction model; after the real-time performance data and operating environment data of the liquid cooling system are predicted and analyzed by the constructed system fault prediction model to obtain the corresponding fault prediction results and system performance trends, a maintenance and optimization plan for the liquid cooling system can be formulated based on the actual prediction results, the flow rate can be adjusted, and a blocking prediction strategy can be formulated, thereby guiding the actual operation and maintenance work to ensure stable and efficient operation of the system.

[0099] In practical applications, a sensor in a liquid cooling system may measure flow once a minute and show a flow of 5 liters / minute. If the flow drops to 3 liters / minute, this may indicate a blockage or leak in the system. In terms of pressure monitoring, the data recorded by the sensor may show that the pressure of the system is usually 2 bar during normal operation. If the monitored pressure suddenly rises to 4 bar, this may mean that there is an abnormality somewhere in the system or that the flow of coolant is restricted. For temperature monitoring, the sensor may show that the normal inlet temperature of the coolant is 25°C and the outlet temperature is 30°C. If the outlet temperature rises to 35°C, this may indicate that the system is thermally cross-linked. The efficiency of the exchange system decreases or the heat load increases; in terms of monitoring energy consumption, the data may show that the system consumes 1 kilowatt-hour (kWh) of electricity per hour under standard operating conditions. If the energy consumption increases to 5kWh / hour, this may mean that the system efficiency is reduced or maintenance is required; the analysis of the coolant composition may show that the coolant contains 2% impurities under normal conditions. If this value rises to 5%, the coolant may need to be replaced or processed; in the statistical analysis of fault data, the maintenance records show that the pump of the liquid cooling system has the most failures in the past year, with a total of 10 failures, followed by the radiator, which has 5 failures. Corresponding marks can be made based on this fault data set. At the same time, in the sensitivity analysis process of model training, it may be found that the pressure monitoring data of the system has the greatest impact on the output of the fault prediction model. In this way, pressure data will be given priority in feature selection to improve the accuracy of the model; in the model evaluation stage, the accuracy of the model is 90% using the cross-validation method, and it takes 5 seconds to calculate the prediction of a fault; based on these indicators, the model parameters can be further adjusted, such as increasing the training data or adjusting the complexity of the algorithm. Finally, the system failure prediction model obtained through training may predict that there is a 20% probability of failure of the cooling system's radiator in the next month. Based on this information, the operation and maintenance team can check the radiator in advance in the subsequent maintenance optimization plan to ensure the continuous and efficient operation of the system.

[0100] At the same time, in order to further ensure the reliability of the liquid cooling system maintenance optimization plan formulated based on the system fault prediction model, this embodiment preferably monitors the effect of the liquid cooling system after the maintenance optimization plan is executed, and predicts the components and fluid states of the liquid cooling system through thermal models and fluid dynamics models based on the actual application effect and real-time operation data of the liquid cooling system, and adjusts the pump speed control parameters and valve opening based on the Q-Learning algorithm, optimizes the heat dissipation efficiency in combination with the radiator fan control, implements system state feedback and fault recovery mechanisms, adjusts control parameters to cope with environmental changes and load fluctuations, and realizes reliable system state optimization and regulation; specifically, Figure 3 As shown, the method also includes:

[0101] S18. Obtain the system operation data after the execution of the liquid cooling system maintenance optimization plan, and predict the state of each system component according to the system operation data, the preset thermal model and the preset fluid dynamics model, and perform system state optimization control according to the corresponding component state prediction results; wherein the system operation data can be understood as the data of the temperature sensor, flow rate sensor, pressure sensor and power consumption sensor obtained by the relevant real-time data acquisition system, based on which the temperature of each component of the liquid cooling system and the current state information of the fluid can be obtained: according to the collected temperature data, the temperature to be reached by each component in the future is predicted by the thermal model, and the system state optimization control is performed. It determines whether the cooling flow needs to be adjusted. If the thermal model shows that the temperature of a component is about to exceed the safe range, the pump speed control parameters are adjusted to increase the flow of the corresponding path to strengthen the cooling of the component; the fluid dynamics model is used to analyze the current flow rate and pressure data to determine the optimal valve opening to ensure that the coolant is evenly distributed in the multi-channel liquid cooling system; through the radiator fan control, the fan speed is adjusted to optimize the heat dissipation effect according to the power consumption data and the output of the prediction model to reduce the overall energy consumption of the liquid cooling system; the Q-Learning algorithm is used to optimize the control parameters of the liquid cooling system to achieve more efficient cooling performance and lower energy consumption. In addition, the system status feedback mechanism can also be used to obtain the results of error detection and performance monitoring. If an abnormality is detected, the control parameters are adjusted immediately or the fault recovery process is executed to ensure the continuous operation of the liquid cooling system. The machine learning algorithm is used to analyze historical and real-time data, continuously adjust and optimize each prediction model, improve the accuracy of the prediction of future thermal load changes, and obtain ambient temperature and workload data to adjust each prediction model and optimization algorithm to cope with environmental changes and load fluctuations to ensure that the multi-channel liquid cooling system is always in the best working state.

[0102] In actual applications, the real-time data acquisition system shows that the temperature sensor reads the temperature of a processor as 70°C, and the preset thermal model predicts that based on the current cooling flow and processor usage, the processor temperature is expected to rise to 85°C within the next 30 minutes, exceeding its safe operating temperature threshold of 80°C; in order to avoid overheating, the control system increases the cooling flow rate by adjusting the pump speed, such as increasing the pump speed from 2000 rpm to 2500 rpm, increasing the flow rate of the coolant, thereby improving the cooling efficiency on this path; at the same time, the fluid dynamics model analyzes the data of the current flow rate of 5 m / s and the pressure of 2 bar, and calculates that the optimal valve opening should be adjusted to 45% opening to ensure that the coolant can be evenly distributed to the various components of the multi-channel liquid cooling system; the radiator fan control system, based on the data displayed by the power consumption sensor, obtains that the current power consumption of the processor is 150 watts, and the predicted The test model recommends increasing the fan speed from 1500 rpm to 1800 rpm, which can improve the heat dissipation efficiency while keeping the energy consumption within a reasonable range. In addition, the control parameters are optimized through genetic algorithms or particle swarm optimization algorithms. For example, the Q-Learning algorithm is used for the optimal combination strategy, and the optimal combination of pump speed, valve opening and fan speed is found, which reduces the total energy consumption of the entire liquid cooling system by 8%. If the system state feedback mechanism detects an abnormal flow sensor reading, such as a sudden drop in flow rate to 5 m / s, the fault recovery process is immediately started, the valve on the affected cooling circuit is closed, and the flow rate of the backup cooling path is increased to maintain the cooling efficiency of the system. The machine learning algorithm is used to analyze historical and real-time data. The support vector machine algorithm is used to train the model obtained through the data of the past week, which reduces the error in predicting future processor temperatures by 5%. The ambient temperature sensor showed that the room temperature increased from 22°C to 28°C and the workload increased from 50% to 75%. The control system adjusted the parameters of the prediction model and optimization algorithm accordingly, for example by adjusting the weight coefficient of the valve opening to ensure that the multi-channel liquid cooling system can remain in the optimal state despite the changes in ambient temperature and load.

[0103] The embodiment of the present application obtains the current working condition of the liquid cooling system by analyzing the collected coolant flow state data in the liquid cooling system. When the current working condition of the liquid cooling system is abnormal, it is determined whether there is abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and the key pipeline nodes with abnormal flow phenomenon are used as pipeline nodes to be analyzed. Then, the real-time monitoring data of each pipeline node to be analyzed are subjected to flow vector analysis to obtain the corresponding node blocking prediction result, and blocking simulation analysis and blocking process prediction are performed according to each node blocking prediction result to obtain the corresponding node blocking process prediction result. After a blocking warning signal is generated according to the node blocking process prediction result, in response to the reception of the blocking warning signal, a blocking response strategy is obtained and executed according to the preset adjustment control logic, and a simulation analysis is performed on the execution of the blocking response strategy to optimize and adjust the blocking response strategy, and an adjustment effect analysis is performed according to the obtained adjusted system operation data to optimize the performance and anti-blocking strategy of the liquid cooling system. At the same time, the real-time prediction of the liquid cooling system is also optimized based on the pre-constructed system fault prediction model. The prediction results of the preset future time period are obtained by predictive analysis of performance data and operating environment data, and the corresponding liquid cooling system maintenance optimization plan is generated and executed. After the liquid cooling system maintenance optimization plan is executed, the state of each system component is predicted according to the monitored system operation data, the preset thermal model and the preset fluid dynamics model, and the technical solution of system state optimization control is performed according to the corresponding component state prediction results. It not only effectively prevents and solves the system blockage problem and ensures the smooth flow of coolant by real-time monitoring and predicting the flow state in the liquid cooling system, but also reduces the risk of system overheating and avoids damage to sensitive electronic components by accurately locating potential blockage points and taking preventive measures in advance, ensuring the stable operation of the equipment, and also reduces the system maintenance cost and extends the equipment service life by automatically adjusting the flow and pressure. Through continuous data collection and analysis, the system self-optimization capability is enhanced, which can flexibly respond to changes in internal conditions, maintain an efficient operating state, and significantly improve the cooling efficiency, reliability and operating performance of the system, thereby providing reliable guarantee for the normal and stable operation of the liquid cooling system.

[0104] It should be noted that although the steps in the above flowchart are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders.

[0105] In one embodiment, Figure 4 As shown, a liquid cooling system self-optimizing control system is provided, the system comprising:

[0106] System status analysis module 1, used to analyze the collected coolant flow status data in the liquid cooling system to obtain the current working status of the liquid cooling system;

[0107] The abnormal pipeline identification module 2 is used to determine whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node when the current working condition of the liquid cooling system is abnormal, and to use the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed;

[0108] The blocking position prediction module 3 is used to perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain the corresponding node blocking prediction result;

[0109] The congestion development prediction module 4 is used to perform congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node, obtain the corresponding node congestion process prediction results, and generate a congestion warning signal and start congestion prevention measures according to the node congestion process prediction results;

[0110] A blocking response processing module 5 is used to obtain and execute a blocking response strategy according to a preset adjustment control logic in response to the reception of the blocking warning signal, and to simulate and analyze the execution of the blocking response strategy to optimize and adjust the blocking response strategy;

[0111] The strategy optimization and adjustment module 6 is used to analyze the adjustment effect according to the obtained adjusted system operation data, so as to optimize the performance and anti-blocking strategy of the liquid cooling system.

[0112] In one embodiment, Figure 5 As shown, a liquid cooling system self-optimizing control system is provided, and the system also includes:

[0113] The system fault prediction module 7 is used to perform predictive analysis on the real-time performance data and operating environment data of the liquid cooling system according to a pre-built system fault prediction model, obtain prediction results for a preset future time period, and generate and execute a corresponding liquid cooling system maintenance optimization plan based on the prediction results; the prediction results include fault prediction results and system performance trends.

[0114] In one embodiment, Figure 6 As shown, a liquid cooling system self-optimizing control system is provided, and the system also includes:

[0115] The system state optimization module 8 is used to obtain the system operation data after the execution of the liquid cooling system maintenance optimization plan, and to predict the state of each system component based on the system operation data, a preset thermal model and a preset fluid dynamics model, and to perform system state optimization control based on the corresponding component state prediction results.

[0116] For the specific limitations of the self-optimizing control system of the liquid cooling system, please refer to the limitations of the self-optimizing control method of the liquid cooling system mentioned above, and the corresponding technical effects can also be obtained equivalently, which will not be repeated here. Each module in the above-mentioned self-optimizing control system of the liquid cooling system can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0117] Figure 7 FIG. 1 shows an internal structure diagram of a computer device in an embodiment, and the computer device may specifically be a terminal or a server. Figure 7 As shown, the computer device includes a processor, a memory, a network interface, a display, a camera and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a self-optimization control method for a liquid cooling system is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball or a touchpad set on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0118] It can be understood by those skilled in the art that Figure 7 The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the computer device to which the present application scheme is applied. The specific computing device may include more or fewer components than shown in the figure, or combine certain components, or have the same component arrangement.

[0119] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the computer program.

[0120] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0121] In summary, the embodiments of the present invention provide a liquid cooling system self-optimization control method, system, computer equipment and medium. The liquid cooling system self-optimization control method realizes the analysis of the collected coolant flow state data in the liquid cooling system to obtain the current working state of the liquid cooling system. When the current working state of the liquid cooling system is abnormal, it is determined whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and the key pipeline nodes with abnormal flow phenomena are used as pipeline nodes to be analyzed. Then, the real-time monitoring data of each pipeline node to be analyzed are subjected to flow vector analysis to obtain the corresponding node blocking prediction result, and blocking simulation analysis and blocking process prediction are performed according to each node blocking prediction result to obtain the corresponding node blocking process prediction result, and a blocking warning signal is generated according to the node blocking process prediction result. In response to the reception of the blocking warning signal, a blocking response strategy is obtained and executed according to the preset adjustment control logic, and the execution of the blocking response strategy is simulated and analyzed to optimize and adjust the blocking response strategy, and the adjustment effect analysis is performed according to the obtained adjusted system operation data to optimize the performance of the liquid cooling system and the anti-blocking strategy. At the same time, based on The pre-built system fault prediction model predicts and analyzes the real-time performance data and operating environment data of the liquid cooling system to obtain the prediction results of the preset future time period, generates and executes the corresponding liquid cooling system maintenance optimization plan, and after the liquid cooling system maintenance optimization plan is executed, the state of each system component is predicted according to the monitored system operation data, the preset thermal model and the preset fluid dynamics model, and the technical solution of system state optimization control is performed according to the corresponding component state prediction results. This method not only effectively prevents and solves the system blockage problem by real-time monitoring and predicting the flow state in the liquid cooling system to ensure the smooth flow of coolant, but also reduces the risk of system overheating by accurately locating potential blockage points and taking preventive measures in advance, avoids damage to sensitive electronic components, and ensures the stable operation of the equipment. It also automatically adjusts the flow and pressure to reduce the system maintenance cost and extend the equipment life. Through continuous data collection and analysis, the system self-optimization capability is enhanced, which can flexibly respond to changes in internal conditions, maintain an efficient operating state, and significantly improve the cooling efficiency, reliability and operating performance of the system, thereby providing reliable guarantee for the normal and stable operation of the liquid cooling system.

[0122] Each embodiment in this specification is described in a progressive manner, and the same or similar parts of each embodiment can be directly referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. It should be noted that the technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, all possible combinations of the technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0123] The above-mentioned embodiments only express several preferred implementation modes of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in the technical field, several improvements and substitutions can be made without departing from the technical principles of the present invention, and these improvements and substitutions should also be regarded as the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be based on the protection scope of the claims.

Claims

1. A liquid cooling system self-optimization control method, characterized in that: The method comprises the following steps: Analyze the collected data on the flow status of the coolant in the liquid cooling system to obtain the current working status of the liquid cooling system; When the current working condition of the liquid cooling system is abnormal, judging whether there is abnormal flow phenomenon according to the coolant state data at each preset key pipeline node, and taking the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed; Perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain a corresponding node blockage prediction result; the real-time monitoring data includes real-time flow data and real-time pressure data; the node blockage prediction result includes a blockage location and a corresponding blockage impact degree; Performing congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node to obtain the corresponding node congestion process prediction results, and generating a congestion warning signal according to the node congestion process prediction results; In response to receiving the blocking warning signal, acquiring and executing a blocking response strategy according to a preset adjustment control logic, and performing simulation analysis on the execution of the blocking response strategy to optimize and adjust the blocking response strategy; Analyze the adjustment effect based on the obtained adjusted system operation data to optimize the liquid cooling system performance and anti-blocking strategy; The step of performing flow vector analysis according to the real-time monitoring data to obtain corresponding node congestion prediction results includes: The coolant fluid characteristics and pipeline specifications of the pipeline node to be analyzed are obtained, and a theoretical fluidity prediction is performed according to the coolant fluid characteristics, the pipeline specifications and a pre-built fluidity prediction model to obtain a corresponding unobstructed flow state; the pipeline specifications include pipeline size and pipeline shape; the unobstructed flow state includes flow direction and flow velocity; Performing fluidity analysis on the real-time monitoring data by using a preset computational fluid dynamics analysis tool to generate a corresponding real-time flow vector diagram of the coolant, and identifying abnormal flow areas based on the real-time flow vector diagram of the coolant and the unobstructed flow state to obtain corresponding abnormal flow areas to be analyzed; According to the obtained pipeline material data and roughness parameters in the abnormal flow area to be analyzed, the corresponding fluid resistance influence is evaluated, and when it is determined that there is an influence, the coolant flow vector distribution is obtained through vector analysis of streamline distribution and flow field distribution; the coolant flow vector distribution includes flow direction distribution and flow velocity distribution; Identify the area with abnormally increased pressure drop according to the coolant flow vector distribution, and use the obtained area with abnormally increased pressure drop as the blocking position; According to the pre-built recurrent neural network model, the influence of each blocking position on the cooling efficiency of the liquid cooling system is evaluated to obtain the corresponding blocking influence degree; The step of performing blocking simulation analysis and blocking process prediction according to the blocking prediction results of each node to obtain the corresponding node blocking process prediction results comprises: According to the blockage prediction results of each node and the preset computational fluid dynamics technology, a blockage simulation analysis is performed on each pipeline node to be analyzed to obtain the corresponding blockage impact analysis results; When the blocking impact analysis result is a blocking impact flow vector, a blocking development prediction is performed according to a pre-built blocking development trend model to obtain a corresponding node blocking process prediction result; the node blocking process prediction result includes a blocking development rate; the prediction expression of the blocking development rate is: Where FND is the blocking development rate; k is the adjustment coefficient; V max 、V min and V avg They represent the maximum, minimum and average flow rates during the monitoring period respectively; α and β represent the index values ​​of the influence of the flow rate change and particle diameter on the blockage development rate respectively; D p Indicates the average diameter of particles suspended in the coolant.

2. The liquid cooling system self-optimization control method according to claim 1, characterized in that: The step of judging whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node includes: Compare and analyze the coolant status data with a pre-built historical status database to obtain corresponding comparison results; If the comparison result is that the flow rate is abnormal and the pressure is normal, the flow rate is predicted according to the obtained liquid viscosity and liquid density, and the corresponding liquid temperature to obtain the corresponding theoretical flow rate, and when the theoretical flow rate is greater than the actual flow rate, the flow pattern is identified based on the flow fluctuation analysis and the pressure fluctuation analysis to obtain the corresponding coolant flow pattern; If the coolant flow pattern suddenly changes from laminar flow to turbulent flow, an isolation forest algorithm is used to perform anomaly detection on the coolant state data; When an abnormality is detected, a corresponding abnormal alarm is issued, and abnormal flow phenomena are classified in combination with the acquired pump working status data; the abnormal flow phenomena include system leakage, filter blockage, internal pipe blockage and valve closure.

3. The liquid cooling system self-optimization control method according to claim 1, characterized in that: The adjusted system operation data includes adjusted flow data, adjusted pressure data, adjusted flow rate, adjusted temperature and adjusted heat load; The step of analyzing the adjustment effect based on the obtained adjusted system operation data to optimize the liquid cooling system performance and anti-blocking strategy includes: When the adjusted flow rate data and the adjusted pressure data are both lower than the corresponding expected values, a comprehensive analysis is performed on the current liquid cooling system according to the adjusted flow rate, the adjusted temperature and the adjusted heat load to determine whether there is a potential blocking problem; When it is determined that there is a potential blocking problem, the blocking relief process is performed according to the preset adjustment rules and the preset adjustment times, and when the preset adjustment times are reached and the liquid cooling system parameters still do not reach the predetermined targets, the vibration and noise monitoring data are collected to analyze whether there is a mechanical performance failure; If it is determined that there is no mechanical performance failure and the risk of leakage is eliminated, the anti-blocking strategy is optimized using a preset self-optimization strategy.

4. The liquid cooling system self-optimization control method according to claim 1, characterized in that: The method further comprises: According to the pre-built system fault prediction model, the real-time performance data and operating environment data of the liquid cooling system are predicted and analyzed to obtain the prediction results for the preset future time period, and based on the prediction results, the corresponding liquid cooling system maintenance optimization plan is generated and executed; the prediction results include fault prediction results and system performance trends.

5. The liquid cooling system self-optimization control method according to claim 4, characterized in that: The method further comprises: The system operation data after the execution of the liquid cooling system maintenance optimization plan is obtained, and the state of each system component is predicted based on the system operation data, a preset thermal model and a preset fluid dynamics model, and the system state optimization control is performed based on the corresponding component state prediction results.

6. A liquid cooling system self-optimizing control system, characterized in that: The system comprises: The system status analysis module is used to analyze the collected coolant flow status data in the liquid cooling system to obtain the current working status of the liquid cooling system; The abnormal pipeline identification module is used to determine whether there is an abnormal flow phenomenon according to the coolant state data at each preset key pipeline node when the current working condition of the liquid cooling system is abnormal, and to use the key pipeline nodes with abnormal flow phenomenon as pipeline nodes to be analyzed; The blockage position prediction module is used to perform flow vector analysis on the real-time monitoring data of each pipeline node to be analyzed to obtain the corresponding node blockage prediction result; the real-time monitoring data includes real-time flow data and real-time pressure data; the node blockage prediction result includes the blockage position and the corresponding blockage impact degree; The congestion development prediction module is used to perform congestion simulation analysis and congestion process prediction according to the congestion prediction results of each node, obtain the corresponding node congestion process prediction results, and generate a congestion warning signal according to the node congestion process prediction results; A blocking response processing module, configured to, in response to receiving the blocking warning signal, acquire and execute a blocking response strategy according to a preset adjustment control logic, and perform simulation analysis on the execution of the blocking response strategy to optimize and adjust the blocking response strategy; The strategy optimization and adjustment module is used to analyze the adjustment effect based on the obtained adjusted system operation data, so as to optimize the performance and anti-blocking strategy of the liquid cooling system; The performing of flow vector analysis according to the real-time monitoring data to obtain corresponding node blocking prediction results includes: The coolant fluid characteristics and pipeline specifications of the pipeline node to be analyzed are obtained, and a theoretical fluidity prediction is performed according to the coolant fluid characteristics, the pipeline specifications and a pre-built fluidity prediction model to obtain a corresponding unobstructed flow state; the pipeline specifications include pipeline size and pipeline shape; the unobstructed flow state includes flow direction and flow velocity; Performing fluidity analysis on the real-time monitoring data by using a preset computational fluid dynamics analysis tool to generate a corresponding real-time flow vector diagram of the coolant, and identifying abnormal flow areas based on the real-time flow vector diagram of the coolant and the unobstructed flow state to obtain corresponding abnormal flow areas to be analyzed; According to the obtained pipeline material data and roughness parameters in the abnormal flow area to be analyzed, the corresponding fluid resistance influence is evaluated, and when it is determined that there is an influence, the coolant flow vector distribution is obtained through vector analysis of streamline distribution and flow field distribution; the coolant flow vector distribution includes flow direction distribution and flow velocity distribution; Identify the area with abnormally increased pressure drop according to the coolant flow vector distribution, and use the obtained area with abnormally increased pressure drop as the blocking position; According to the pre-built recurrent neural network model, the influence of each blocking position on the cooling efficiency of the liquid cooling system is evaluated to obtain the corresponding blocking influence degree; The blocking simulation analysis and blocking process prediction are performed according to the blocking prediction results of each node to obtain the corresponding node blocking process prediction results, including: According to the blockage prediction results of each node and the preset computational fluid dynamics technology, a blockage simulation analysis is performed on each pipeline node to be analyzed to obtain the corresponding blockage impact analysis results; When the blocking impact analysis result is a blocking impact flow vector, a blocking development prediction is performed according to a pre-built blocking development trend model to obtain a corresponding node blocking process prediction result; the node blocking process prediction result includes a blocking development rate; the prediction expression of the blocking development rate is: Where FND is the blocking development rate; k is the adjustment coefficient; V max 、V min and V avg They represent the maximum, minimum and average flow rates during the monitoring period respectively; α and β represent the index values ​​of the influence of the flow rate change and particle diameter on the blockage development rate respectively; D p Indicates the average diameter of particles suspended in the coolant.

7. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images