A smart control system for the operation of computer room infrastructure

Abstract

This invention discloses an intelligent control system for the operation of data center infrastructure, belonging to the field of data center infrastructure monitoring technology. It includes a data center monitoring module, a heat flow analysis module, and a data center control module. The data center monitoring module monitors the data center and obtains monitoring data. The heat flow analysis module performs heat flow analysis, identifies heat flow monitoring data within the monitoring data, generates a heat flow dynamic graph based on the monitoring data, analyzes the data center temperature regulation based on the heat flow dynamic graph, obtains the temperature adjustment method, and sends the temperature regulation method to the data center control module. The system also acquires real-time control records from the data center control module, performs infrastructure evaluation based on the real-time adjustment records and the heat flow dynamic graph, obtains the facility evaluation result, and sends the facility evaluation result to the data center control module. The data center control module controls the infrastructure based on the temperature adjustment method and the facility evaluation result.

Description

Technical Field

[0001] This invention belongs to the field of data center infrastructure monitoring technology, specifically an intelligent control system for the operation of data center infrastructure. Background Technology

[0002] With the rapid development of information technology, the construction of AI big data models and intelligent computing centers is accelerating. Data center scale is continuously expanding, equipment density is constantly increasing, single rack power is significantly rising, equipment heat generation is surging, and heat flow distribution is becoming increasingly complex. Heat flow data has become core data reflecting the operational status of data centers, ensuring equipment safety, and optimizing operational efficiency. Data center infrastructure encompasses multiple subsystems such as power supply and distribution, and cooling. Its operational stability directly affects the normal operation of IT equipment and maintenance costs. However, currently, the utilization of heat flow data in data center monitoring and management is rudimentary or even nonexistent.

[0003] Specifically, this manifests in the following ways: First, comprehensive and accurate heat flow data collection is not conducted. Only the overall temperature of the computer room is monitored, ignoring key data such as local heat flow distribution and flow patterns, and failing to capture core information such as peak heat values ​​and heat flow blind spots around the equipment. Second, in-depth analysis of the collected heat flow data is not performed, failing to uncover the intrinsic correlation between heat flow data and equipment operating status, cooling system energy efficiency, and power supply parameters. Heat flow data is only used as an auxiliary reference and cannot provide support for control decisions. Third, coordinated linkage between various subsystems is not achieved based on heat flow data. The control of subsystems such as cooling and power supply is deviated from the heat flow change patterns, still adopting fixed parameters or passive response modes, resulting in a lack of targeted control.

[0004] In order to solve the above problems, the present invention provides an intelligent control system for the operation of data center infrastructure. Summary of the Invention

[0005] To address the problems of the above solutions, this invention provides an intelligent control system for the operation of data center infrastructure.

[0006] The objective of this invention can be achieved through the following technical solutions: An intelligent control system for the operation of data center infrastructure includes a data center monitoring module, a heat flow analysis module, and a data center control module; The data center monitoring module is used to monitor the data center and obtain data center monitoring data. The data center monitoring data includes basic monitoring data and heat flow monitoring data. The basic monitoring data is the monitoring data of non-heat flow monitoring data in the data center monitoring.

[0007] Furthermore, the acquisition of heat flow monitoring data includes: A multi-point sensor array is installed, and monitoring is performed based on the installed multi-point sensor array to obtain heat flow monitoring data.

[0008] The heat flow analysis module is used to perform heat flow analysis, identify heat flow monitoring data in the data center monitoring data, generate a heat flow dynamic graph based on the heat flow monitoring data, perform data center temperature control analysis based on the heat flow dynamic graph, obtain the temperature adjustment method, and send the temperature control method to the data center control module; obtain the real-time control records of the data center control module, perform infrastructure assessment based on the real-time adjustment records and the heat flow dynamic graph, obtain the facility assessment results, and send the facility assessment results to the data center control module.

[0009] Furthermore, based on the heat flow dynamics diagram, analysis of data center temperature control is performed, including: A digital twin model of the data center is established based on digital twin technology. Simulation is performed based on the digital twin model to obtain the optimal heat flow map of the data center. The optimal heat flow map is marked as the heat flow target map. The digital twin model is adjusted accordingly based on the heat flow dynamic diagram. The temperature adjustment method is obtained by performing adjustment simulation based on the heat flow target diagram and the digital twin model.

[0010] Furthermore, the simulation is adjusted based on the heat flow target map and digital twin model, including: Set the heat flux parameters in the heat flux target diagram to the heat flux control target values ​​of the digital twin model; The system pre-defines heat flux adjustment items and adjustment criteria for these items; it performs real-time adjustment and evaluation of each heat flux adjustment item, and marks those items that meet the corresponding adjustment criteria as simulation adjustment items. In the digital twin model, the control parameters of the simulation adjustment items are adjusted one by one to form the alternative adjustment methods and temperature control simulation data; Priority analysis is performed based on temperature control simulation data, and the candidate adjustment method corresponding to the highest priority temperature control simulation data is marked as the temperature control adjustment method.

[0011] Furthermore, priority analysis is performed based on temperature control simulation data, including: Set temperature control standard data, and mark the heat flux change effect, energy consumption data, and adjustment response speed in the temperature control standard data as standard heat flux change effect, standard energy consumption data, and standard adjustment response speed, respectively. Based on temperature control simulation data, identify the corresponding heat flux change effect, energy consumption data, and adjustment response speed. Based on similarity algorithm, calculate the effect difference, energy consumption difference, and response speed difference between the heat flux change effect, energy consumption data, and adjustment response speed and the standard heat flux change effect, standard energy consumption data, and standard adjustment response speed. The overall difference value is calculated based on the difference in performance, energy consumption, and response speed. The candidate adjustment methods corresponding to temperature control simulation data with a comprehensive difference value greater than the threshold X1 are eliminated. The remaining alternative adjustment methods are prioritized in ascending order of their comprehensive difference values.

[0012] Furthermore, the temperature control standard data selected includes the best heat flow change effect, the best energy consumption data, and the best adjustment response speed.

[0013] Furthermore, the comprehensive difference value is the sum of the effect difference value, energy consumption difference value, and response speed difference value.

[0014] Furthermore, based on the heat flow dynamics diagram, analysis of data center temperature control is performed, including: Step SA1: Establish a digital twin model of the computer room based on digital twin technology, perform simulation based on the digital twin model, obtain the optimal heat flow map and simulation effect information database of the computer room, and mark the optimal heat flow map as the heat flow target map; the simulation effect information database stores the corresponding heat flow effect difference between the corresponding heat flow dynamic map and the heat flow target map; Step SA2: Determine the difference in heat flow effect between the dynamic heat flow diagram and the target heat flow diagram based on the simulation effect information database; When the difference in heat flow effect is less than the threshold X2, the temperature adjustment method is none; When the difference in heat flux effect is not less than the threshold X2, the target heat flux adjustment area is determined according to the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is the heat flux difference area that needs to be adjusted first in the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is simulated based on the digital twin model to obtain the temperature adjustment method of the target heat flux adjustment area. The temperature control adjustment method is output. Step SA3: Obtain the new heat flow dynamics and return to step SA2.

[0015] Furthermore, infrastructure assessment is conducted based on real-time adjustment records and heat flow dynamics, including: Based on the real-time identification of the basic monitoring data corresponding to the heat flow dynamic diagram according to the data of the computer room monitoring, the heat flow dynamic diagram is evaluated based on the basic monitoring data to obtain the heat flow evaluation result; When the heat flow assessment result indicates an abnormal heat flow, the cause of the abnormal heat flow shall be determined based on the aforementioned heat flow dynamic diagram. When the heat flow assessment result is that the heat flow is normal, no corresponding action is taken; Identify the control information in the real-time adjustment record, obtain the heat flow dynamic diagram corresponding to the control information, evaluate the heat flow change based on the heat flow dynamic diagram and the control information, and obtain the heat flow change evaluation result. When the heat flow change is assessed as normal, no corresponding action is taken; When assessing abnormal heat flow changes, the cause of the abnormal heat flow changes is determined based on the heat flow dynamic diagram and control information. The results of heat flow assessment, the causes of heat flow anomalies, the results of heat flow change assessment, and the causes of heat flow change anomalies are integrated into the facility assessment results.

[0016] Furthermore, heat flux changes are assessed based on the heat flux dynamics diagram and control information, including: The heat flow simulation data is obtained by performing simulation based on the control information using a preset digital twin model; the heat flow dynamic diagram is calibrated based on the heat flow simulation data to obtain the heat flow change evaluation result.

[0017] The computer room control module is used to control the computer room infrastructure, receive temperature adjustment methods and facility evaluation results, and control the infrastructure based on the temperature adjustment methods and facility evaluation results.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention combines three methods—fiber optic grating monitoring, sensor array monitoring, and infrared thermal imaging monitoring—to comprehensively collect heat flow data (temperature gradient, heat flow density, flow direction, flow velocity, etc.) across the entire computer room. This overcomes the limitations of previous technologies that simply monitored the overall temperature of the computer room while ignoring local heat flow distribution and flow patterns. By comparing the dynamic heat flow map with the target heat flow map, it accurately identifies areas of heat flow difference, uncovers the intrinsic correlation between heat flow data and the optimal heat flow state of the computer room, and upgrades heat flow data from an auxiliary reference to a core basis for regulation. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a block diagram illustrating the principle of the present invention. Detailed Implementation

[0021] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figure 1 As shown, an intelligent control system for the operation of data center infrastructure includes a data center monitoring module, a heat flow analysis module, and a data center control module. The data center monitoring module is used to monitor the data center and obtain data center monitoring data. It includes monitoring units for various data monitoring, such as power monitoring units (mains power, UPS, batteries, power distribution cabinets, etc.), humidity monitoring units, air quality monitoring units, and heat flow monitoring units, which are used to monitor corresponding data. In order to distinguish the monitoring units, the monitoring units other than the heat flow monitoring units are marked as basic monitoring units. That is, they can be monitored according to the original or other existing monitoring methods to obtain basic monitoring data and heat flow monitoring data. The basic monitoring data and heat flow monitoring data are then integrated into data center monitoring data.

[0023] In one embodiment, the heat flow monitoring unit adopts a distributed fiber optic grating monitoring method, which uses distributed fiber optic sensing technology to continuously deploy special sensing optical fibers with multiple built-in fiber optic grating sensors along the server rack array, cold aisle, hot aisle, air conditioning pipes, ceiling and underfloor areas of the computer room. The optical fibers are directly attached to the surface of the rack, the outer wall of the pipeline or arranged overhead.

[0024] By transmitting laser signals to the sensing fiber through an optical fiber demodulator, and utilizing the physical characteristics of fiber optic grating wavelength changing with temperature and heat flux gradient, the temperature gradient, heat flux density, heat flux distribution, and long-distance continuous temperature field are collected in real time. Combined with multi-point synchronous measurement along the optical fiber, kilometer-level full-area continuous heat flux monitoring can be achieved, which can accurately capture pipeline micro-leakage heat anomalies, hidden interlayer heat accumulation, and long-distance airflow temperature changes.

[0025] This method is a contact-based continuous monitoring method, which has strong resistance to electromagnetic interference, high accuracy, and can achieve uninterrupted full-area monitoring. However, it can only acquire indirect data on temperature-related heat flow and cannot directly measure airflow speed and direction.

[0026] In one embodiment, the heat flow monitoring unit is an infrared thermal imaging non-contact heat flow monitoring method; infrared thermal imaging monitoring uses infrared thermal imaging cameras installed on the top of the computer room and on both sides of the passage to obtain thermal images of the temperature distribution on the surface of the computer room, cabinets, and equipment in a non-contact manner based on the principle of infrared radiation of objects.

[0027] By analyzing image pixels to determine temperature, the system identifies surface hotspots, abnormal equipment heating, heat buildup on cabinet surfaces, and boundaries between hot and cold airflows. Combined with image algorithms, it infers the heat diffusion trend and approximate flow direction. It can quickly scan the entire area and intuitively present a visual image of the thermal field. No wiring or contact with equipment is required, making deployment convenient.

[0028] This method is a non-contact visual monitoring method with fast response speed and can achieve large-scale rapid inspection; however, it cannot penetrate cabinets, walls, or mezzanines, cannot measure internal heat flow and airflow speed, and is greatly affected by dust, light, and obstruction. It can only obtain the surface temperature field, and the measurement accuracy is lower than the previous two methods. It is mostly used for coarse hotspot location and visual assistance.

[0029] In one embodiment, the heat flow monitoring unit is a multi-point sensor array heat flow monitoring method; the sensor array monitoring uses distributed temperature and humidity, wind speed, and heat flow density micro sensors, and two-dimensional or three-dimensional sensor arrays are deployed in a grid layout in the computer room space, the upper and lower layers inside the cabinet, the air inlet, the air outlet, the cold / hot aisle, and the air conditioning supply and return air area.

[0030] The sensors directly collect point temperature, air velocity, wind direction, heat flux density, temperature difference between the air entering and exiting the cabinet, and local heat exchange volume, and upload the data via wired or wireless means; based on the multi-point data of the array, the heat flow field of the computer room is reconstructed by interpolation, and hot spots, thermal dead zones, and airflow short-circuit areas are located to achieve cabinet-level fine-grained heat flow monitoring.

[0031] This method is a contact-based discrete monitoring system that can directly measure the core physical quantity of heat flow. It is flexible in location, simple to deploy, and moderately costly, making it suitable for precise monitoring inside cabinets and in local areas. However, due to the limited number of sensors, the spatial coverage is discrete, and continuous full-area monitoring cannot be achieved. The measurement accuracy is easily affected by interference in strong electromagnetic environments.

[0032] For example, the computer room is analyzed to determine the location of each sensor, and the sensors are installed according to the location. A corresponding digital twin model can be established based on digital twin technology, and simulation can be performed based on the digital twin model to install a multi-point sensor array.

[0033] In one embodiment, the heat flow monitoring unit primarily uses a multi-point sensor array for heat flow monitoring, combined with fiber optic grating distributed monitoring and infrared thermal imaging non-contact heat flow monitoring for collaborative monitoring.

[0034] The heat flow analysis module is used to perform heat flow analysis, identify heat flow monitoring data in the data center monitoring data, and generate a heat flow dynamic graph based on the heat flow monitoring data; it performs data center temperature control analysis based on the heat flow dynamic graph to obtain temperature adjustment methods, and sends the temperature control methods to the data center control module; it obtains the real-time control records of the data center control module, and the real-time adjustment records include various adjustment records that will affect heat flow changes, such as temperature adjustment and basic equipment control; it performs infrastructure assessment based on the real-time adjustment records and the heat flow dynamic graph to obtain facility assessment results, and sends the facility assessment results to the data center control module.

[0035] In one embodiment, the heat flow monitoring data is obtained by using a multi-point sensor array heat flow monitoring method as the main method, combined with a fiber optic grating distributed monitoring method and an infrared thermal imaging non-contact heat flow monitoring method for collaborative monitoring. Fiber Bragg grating data: acquire temperature gradients, long-distance heat flow distributions, and heat flow data in concealed areas of pipelines / interlayers over continuous space, serving as basic data for the continuous field; Sensor array data: provides heat flux density, airflow velocity, wind direction, and cabinet inlet and outlet air temperature difference at discrete points, serving as precise point location constraint data; Infrared thermal imaging data: Outputs surface temperature field images and hotspot region outlines as visualization boundary and texture data.

[0036] Time alignment, outlier removal, and unit normalization are performed on fiber optic grating data, sensor array data, and infrared thermal imaging data. Spatiotemporal interpolation algorithms are used to fill in the blank areas of discrete points and a unified three-dimensional thermal flow field numerical matrix is ​​constructed to solve the problems of incomplete coverage and insufficient accuracy of single monitoring methods. Based on the fundamental principles of heat conduction and convection heat transfer, heat flux vector calculation is performed on the fused data; Temperature gradient → direction of heat flow (arrow direction); Heat flux density, wind speed → arrow size, color intensity; Temperature range → Thermal chromatography (blue-green-yellow-orange-red); Continuous temporal data → dynamic frame sequence.

[0037] By using grid-based modeling, the computer room space, server racks, cold / hot aisles, ceilings, and underfloor spaces are divided into three-dimensional grid units. The magnitude and direction of heat flow are calculated for each unit to construct a vector-type heat flow field model. Dynamic heat flow field visualization is achieved using web terminals, large screen terminals, and twin platforms to obtain dynamic heat flow diagrams.

[0038] The heat flow dynamic map is a real-time dynamic visualization of thermal vector spectrum. Based on the physical space of the computer room, it intuitively presents the spatiotemporal changes of the thermal field, such as temperature gradient, heat flow density, heat flow direction, heat flow velocity, hot spot evolution, airflow short circuit, hot and cold collision, and eddy dead zone, through thermal chromatogram, dynamic vector arrows, and time-series animation. It realizes the transformation of the heat flow field from static data to dynamic visualization, and can reflect the heat flow distribution and evolution process of the entire space, including the inside of the cabinet, hot and cold aisles, ceiling, and floor mezzanine, in real time.

[0039] In one embodiment, a dynamic heat flow graph can also be generated based on heat flow monitoring data using various existing methods.

[0040] In one embodiment, data center temperature control analysis based on heat flow dynamics includes: Preliminary preparations: Establish a digital twin model of the data center based on digital twin technology, conduct simulation based on the digital twin model to obtain the optimal heat flow map of the data center, and mark it as the heat flow target map; Real-time analysis: Based on the heat flow dynamic diagram, the digital twin model is adjusted accordingly. Taking the room temperature control target of the heat flow target diagram as the target, the digital twin model is used to adjust and simulate the temperature control method.

[0041] In one embodiment, the simulation is adjusted using a digital twin model, based on the current simulation method.

[0042] In one embodiment, the adjusted simulation based on the heat flow target map and the digital twin model includes: The heat flow parameters (optimal temperature gradient, heat flow density, flow direction, etc.) in the heat flow target diagram are set as the heat flow control target values ​​of the digital twin model; The system pre-sets heat flux adjustment items and adjustment standards for these items. Common heat flux adjustment items include air conditioner operating frequency, supply and return air temperatures, outlet angle, fan speed, cold / hot aisle closure status, cabinet airflow angle, fresh air intake and temperature, and load adjustments for non-core equipment. Adjustment standards define the conditions under which an item can or cannot be adjusted. These standards can be set based on historical data or manually. The system performs real-time evaluation of each heat flux adjustment item and marks those meeting the corresponding adjustment standards as simulation adjustment items. In the digital twin model, the control parameters of the simulation adjustment items are adjusted one by one to simulate the process of the evolution of the heat flow state of the computer room to the heat flow target diagram under different combinations of control parameters of the simulation adjustment items. The heat flow change effect, energy consumption data and adjustment response speed corresponding to each set of adjustment schemes are recorded and summarized into candidate adjustment methods and temperature control simulation data. Priority analysis is performed based on temperature control simulation data. The candidate adjustment method corresponding to the highest priority temperature control simulation data is marked as the temperature control adjustment method. For example, the highest priority is the adjustment method with the highest efficiency, lowest energy consumption, fastest heat flow to the target value, and no impact on the normal operation of the equipment. Other methods are calculated based on the degree of difference between the ideal and the optimal method, and the smaller the difference, the higher the priority. Priority can also be compared based on the similarity to the ideal and optimal method.

[0043] In one embodiment, each heat flux adjustment item is evaluated in real time. Based on the adjustment criteria, corresponding characteristic data is collected, and the characteristic data is calibrated according to the adjustment criteria to determine whether it meets the adjustment criteria and can be adjusted.

[0044] In one embodiment, because not all heat flux adjustment terms may be used as simulation adjustment terms, in some cases, adjusting to the target heat flux map may result in poor heat flux change effects, high energy consumption, and slow adjustment response. Therefore, in this embodiment, the target heat flux map is no longer used as the absolute target. Priority analysis is performed based on temperature control simulation data, including: Based on the digital twin model or historical data, select a set of heat flux change effects, energy consumption data, and adjustment response speed as temperature control standard data, such as optimal heat flux change effect, optimal energy consumption data, and optimal adjustment response speed. The temperature control standard data is used to form a reference standard for subsequent priority comparison, which facilitates the determination of priorities. Any set can be selected, but the optimal temperature control data is preferred for easy calculation. The heat flux change effect, energy consumption data, and adjustment response speed in the temperature control standard data are marked as standard heat flux change effect, standard energy consumption data, and standard adjustment response speed, respectively. Based on temperature control simulation data, identify the corresponding heat flux change effect, energy consumption data, and adjustment response speed. Based on similarity algorithms (such as cosine similarity), calculate the effect difference, energy consumption difference, and response speed difference between the heat flux change effect, energy consumption data, and adjustment response speed and the standard heat flux change effect, standard energy consumption data, and standard adjustment response speed. Energy consumption and response speed are directly calculated based on numerical values. A negative value indicates better performance, and a positive value indicates worse performance. For example, (energy consumption data - standard energy consumption data) ÷ standard energy consumption data, (standard adjustment response speed - adjustment response speed) ÷ standard adjustment response speed. The comprehensive difference value is calculated based on the difference in performance, energy consumption, and response speed. The weighted coefficients corresponding to the difference in performance, energy consumption, and response speed are multiplied and then summed. If the weighted difference is not considered, the difference in performance, energy consumption, and response speed are simply summed to obtain the comprehensive difference value. The candidate adjustment methods corresponding to temperature control simulation data with a comprehensive difference value greater than the threshold X1 are eliminated. The remaining alternative adjustment methods are prioritized in ascending order of their comprehensive difference values.

[0045] In one embodiment, data center temperature control analysis based on heat flow dynamics includes: Step SA1: Establish a digital twin model of the computer room based on digital twin technology, perform simulation based on the digital twin model to obtain the optimal heat flow map of the computer room, and mark the optimal heat flow map as the heat flow target map; In the process of obtaining the optimal heat flow map of the computer room through simulation based on the digital twin model, the difference in heat flow effect between different heat flow dynamic maps and the target heat flow map is determined based on the accumulated simulation data. The difference in heat flow effect can be determined by the difference in the operating status of the infrastructure. For example, the individual impact value of each infrastructure can be calculated based on the functional impact of each infrastructure. The comprehensive difference of the facilities can be calculated based on the preset weight coefficients of each infrastructure. If factors such as cost are not considered, the comprehensive difference of the facilities can be marked as the difference in heat flow effect. If other factors are considered, a comprehensive calculation is performed based on the weight coefficients and effect differences of each factor and the comprehensive difference of the facilities. A simulation effect information database is established, and the difference in heat flow effect between the corresponding heat flow dynamic map and the target heat flow map is determined through the simulation effect information database.

[0046] Step SA2: Determine the difference in heat flow effect between the dynamic heat flow diagram and the target heat flow diagram based on the simulation effect information database; When the difference in heat flow effect is less than the threshold X2, no adjustment is made, and the temperature adjustment mode is set to none, indicating that no adjustment is required. When the difference in heat flux effect is not less than the threshold X2, the target heat flux adjustment area is determined based on the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is the heat flux difference area that is prioritized for adjustment in the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is simulated based on the digital twin model to obtain the temperature adjustment method for the target heat flux adjustment area (the temperature control adjustment method can be determined and simulated according to the above embodiment), and the temperature control adjustment method is output. Step SA3: Obtain the new heat flow dynamic diagram (the computer room control module adjusts according to the temperature control adjustment method) to achieve more efficient and flexible temperature control adjustment, and return to step SA2.

[0047] In one embodiment, the target heat flow adjustment region is determined based on the heat flow dynamic map and the heat flow target map, the heat flow difference region between the heat flow dynamic map and the heat flow target map is identified, the adjustment priority analysis is performed on each heat flow difference region, the priority is determined according to the degree of difference and importance, and the heat flow difference region with the highest priority is marked as the target heat flow adjustment region; the heat flow difference region that should be adjusted first can also be determined based on other methods.

[0048] For example, a point-by-point comparison is performed between the dynamic heat flux map and the target heat flux map across the entire region. Based on the differences in the core heat flux parameters between the two maps, all areas where the heat flux state does not meet the target standard are identified, i.e., areas of heat flux difference. The specific identification process is as follows: The core heat flow parameters (temperature gradient, heat flow density, heat flow direction, flow velocity, hot spot range, etc.) of the dynamic heat flow map and the target heat flow map are aligned in time and space to ensure consistent comparison benchmarks. A difference analysis method is used to calculate the difference in heat flow parameters at corresponding locations on the dynamic heat flow map and the target heat flow map, grid by grid and point by point, generating a heat flow difference quantification matrix to clarify the magnitude of differences in each region. A heat flow parameter difference threshold is set (which can be flexibly adjusted according to the safety standards and energy-saving targets of the computer room equipment), and regions exceeding the threshold in the difference quantification matrix are extracted and marked as heat flow difference regions. Simultaneously, each difference region is independently numbered, and its specific location, coverage area, core difference parameters, and difference magnitude are recorded to form a list of heat flow difference regions.

[0049] Evaluation indicators are set, such as the importance level of equipment in the area (core equipment / auxiliary equipment / non-core equipment), equipment density, heat flow conduction influence (the diffusion effect of abnormal heat flow in this area on the surrounding area), and whether it is a hidden area (areas that are difficult to adjust, such as ceiling / floor mezzanine, which are prone to causing hidden dangers). Based on the above evaluation indicators, each area with heat flow difference is scored (out of 100 points). The higher the score, the greater the impact of the area on the overall heat flow control and equipment safety of the computer room, and the higher its importance. The comprehensive priority score of each area with heat flow difference is calculated using the formula: difference degree score × weight A + area importance score × weight B (weights A and B can be adjusted according to the focus of computer room control, with A=0.5 and B=0.5 by default, taking into account both difference degree and importance). All areas with heat flow difference are sorted from high to low according to the comprehensive priority score.

[0050] In one embodiment, infrastructure assessment based on real-time adjustment records and heat flow dynamics includes: Based on the real-time identification of the basic monitoring data corresponding to the heat flow dynamic graph from the computer room monitoring data, the heat flow dynamic graph is evaluated based on the basic monitoring data to determine whether the heat flow dynamic graph is abnormal under the current background, and the heat flow evaluation result is obtained. When the heat flow evaluation result is that the heat flow is abnormal, the cause of the heat flow abnormality is determined according to the heat flow dynamic graph; when the heat flow evaluation result is that the heat flow is normal, no corresponding action is taken. The system identifies control information in real-time adjustment records, such as the time and operation performed, and forms control information according to the entire control process corresponding to the control target, such as temperature adjustment, specifying which devices are adjusted to achieve the temperature adjustment. It also identifies the corresponding heat flow dynamic graph, i.e., determines the heat flow dynamic graph according to the time corresponding to the control information, such as the preset duration before control + control period + heat flow response time after control, and determines whether the changes in the heat flow dynamic graph are normal under the control adjustment corresponding to the control information. Based on the heat flow dynamic graph and control information, it performs heat flow change evaluation to obtain the evaluation result. When the evaluated heat flow change is normal, no corresponding processing is performed; when the evaluated heat flow change is abnormal, the cause of the abnormal heat flow change is determined based on the heat flow dynamic graph and control information.

[0051] The results of heat flow assessment, the causes of heat flow anomalies, the results of heat flow change assessment, and the causes of heat flow change anomalies are integrated into the facility assessment results.

[0052] In one embodiment, the heat flow dynamic diagram is evaluated based on basic monitoring data. The range of the heat flow dynamic diagram under normal conditions is determined based on the basic monitoring data and corresponding historical data, and then the heat flow dynamic diagram is evaluated to determine whether it is abnormal. It can also be determined by matching based on historical heat flow anomaly data. Alternatively, an anomaly identification model can be established based on machine learning, deep learning algorithms, isolated forest algorithms, etc., and heat flow anomaly identification can be performed through the anomaly identification model. If the heat flow density in a certain area is consistently high, the difference in heat flow density between areas is too large, or the heat flow density fluctuates drastically in a short period of time, it indicates uneven heat dissipation, hot spot accumulation, and unbalanced cold energy distribution. If the flow direction is chaotic and local eddies / backflows occur, if the cold air does not flow to the equipment but flows back to the air conditioner, or if the hot air flows back into the cold aisle.

[0053] In one embodiment, evaluating the heat flow dynamics graph based on baseline monitoring data includes: Based on the digital twin model established above, simulation verification is performed to determine whether there is an abnormal heat flow, and at the same time, the cause of the abnormal heat flow can be determined.

[0054] In one embodiment, the cause of the heat flow anomaly is determined based on the heat flow dynamic diagram, the abnormal heat flow is identified, the possible causes of the heat flow anomaly are determined based on historical data, and the possibility of the cause being possible is analyzed to determine the cause of the heat flow anomaly. If only one cause cannot be found, multiple causes are considered as causes of the heat flow anomaly and then further determined.

[0055] In one embodiment, heat flux change assessment is performed based on heat flux dynamics and control information. The normal heat flux dynamics range under the corresponding background can be determined in the manner described above, and then the heat flux change assessment can be performed.

[0056] In one embodiment, assessing heat flux changes based on a heat flux dynamics graph and control information includes: The heat flow simulation data is obtained by performing simulation based on control information using a digital twin model; the heat flow dynamic diagram is calibrated based on the heat flow simulation data to determine whether the heat flow change is abnormal, whether the deviation from the heat flow simulation data exceeds a preset value, or whether the difference from the abnormal heat flow change data is within a preset range; and the heat flow change evaluation result is obtained. When assessing abnormal heat flow changes, the cause of the abnormal heat flow changes can be determined simultaneously using a digital twin model.

[0057] In one embodiment, the cause of abnormal heat flow changes can be determined based on the heat flow dynamic diagram and control information. Alternatively, an intelligent model can be established based on machine learning, deep learning algorithms, etc., to determine the cause of abnormal heat flow changes, and training can be performed based on historical data labeled with corresponding training sets.

[0058] The computer room control module is used to control the computer room infrastructure, receive temperature adjustment methods and facility evaluation results, and control the infrastructure based on the temperature adjustment methods and facility evaluation results.

[0059] Temperature control can be implemented according to temperature adjustment methods; the infrastructure can be assessed to determine if there are any abnormalities, and if so, emergency measures can be taken according to the preset procedures.

[0060] The above formulas are all numerical calculations after removing dimensions. The formulas are obtained by software simulation based on a large amount of data and are closest to the real situation. The preset parameters and preset thresholds in the formulas are set by those skilled in the art according to the actual situation or obtained by simulation based on a large amount of data.

[0061] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An intelligent control system for the operation of computer room infrastructure, characterized in that, Includes a data center monitoring module, a heat flow analysis module, and a data center control module; The computer room monitoring module is used to monitor the computer room and obtain computer room monitoring data. The computer room monitoring data includes basic monitoring data and heat flow monitoring data. The basic monitoring data is the monitoring data of non-heat flow monitoring data in the computer room monitoring. The heat flow analysis module is used to perform heat flow analysis, identify heat flow monitoring data in the data room monitoring data, generate a heat flow dynamic graph based on the heat flow monitoring data, perform data room temperature regulation analysis based on the heat flow dynamic graph, obtain the temperature adjustment method, and send the temperature regulation method to the data room control module. The system acquires real-time control records from the data center control module, performs infrastructure assessment based on real-time adjustment records and heat flow dynamics, obtains facility assessment results, and sends the facility assessment results to the data center control module. The computer room control module is used to control the computer room infrastructure, receive temperature adjustment methods and facility evaluation results, and control the infrastructure based on the temperature adjustment methods and facility evaluation results.

2. The intelligent control system for the operation of computer room infrastructure according to claim 1, characterized in that, Acquisition of heat flow monitoring data includes: A multi-point sensor array is installed, and monitoring is performed based on the installed multi-point sensor array to obtain heat flow monitoring data.

3. The intelligent control system for the operation of computer room infrastructure according to claim 1, characterized in that, Analysis of computer room temperature control based on heat flow dynamics diagrams, including: A digital twin model of the data center is established based on digital twin technology. Simulation is performed based on the digital twin model to obtain the optimal heat flow map of the data center. The optimal heat flow map is marked as the heat flow target map. The digital twin model is adjusted accordingly based on the heat flow dynamic diagram. The temperature adjustment method is obtained by performing adjustment simulation based on the heat flow target diagram and the digital twin model.

4. The intelligent control system for the operation of computer room infrastructure according to claim 3, characterized in that, The simulation was adjusted based on the heat flow target map and digital twin model, including: Set the heat flux parameters in the heat flux target diagram to the heat flux control target values ​​of the digital twin model; The system pre-defines heat flux adjustment items and adjustment criteria for these items; it performs real-time adjustment and evaluation of each heat flux adjustment item, and marks those items that meet the corresponding adjustment criteria as simulation adjustment items. In the digital twin model, the control parameters of the simulation adjustment items are adjusted one by one to form the alternative adjustment methods and temperature control simulation data; Priority analysis is performed based on temperature control simulation data, and the candidate adjustment method corresponding to the highest priority temperature control simulation data is marked as the temperature control adjustment method.

5. The intelligent control system for the operation of computer room infrastructure according to claim 4, characterized in that, Priority analysis based on temperature control simulation data includes: Set temperature control standard data, and mark the heat flux change effect, energy consumption data, and adjustment response speed in the temperature control standard data as standard heat flux change effect, standard energy consumption data, and standard adjustment response speed, respectively. Based on temperature control simulation data, identify the corresponding heat flux change effect, energy consumption data, and adjustment response speed. Based on similarity algorithm, calculate the effect difference, energy consumption difference, and response speed difference between the heat flux change effect, energy consumption data, and adjustment response speed and the standard heat flux change effect, standard energy consumption data, and standard adjustment response speed. The overall difference value is calculated based on the difference in performance, energy consumption, and response speed. The candidate adjustment methods corresponding to temperature control simulation data with a comprehensive difference value greater than the threshold X1 are eliminated. The remaining alternative adjustment methods are prioritized in ascending order of their comprehensive difference values.

6. The intelligent control system for the operation of computer room infrastructure according to claim 5, characterized in that, The temperature control standard data selected is the best heat flow change effect, the best energy consumption data, and the best adjustment response speed.

7. The intelligent control system for the operation of computer room infrastructure according to claim 5, characterized in that, The overall difference value is the sum of the difference in performance, energy consumption, and response speed.

8. The intelligent control system for the operation of computer room infrastructure according to claim 1, characterized in that, Analysis of computer room temperature control based on heat flow dynamics diagrams, including: Step SA1: Establish a digital twin model of the computer room based on digital twin technology, perform simulation based on the digital twin model, obtain the optimal heat flow map and simulation effect information database of the computer room, and mark the optimal heat flow map as the heat flow target map; the simulation effect information database stores the corresponding heat flow effect difference between the corresponding heat flow dynamic map and the heat flow target map; Step SA2: Determine the difference in heat flow effect between the dynamic heat flow diagram and the target heat flow diagram based on the simulation effect information database; When the difference in heat flow effect is less than the threshold X2, the temperature adjustment method is none; When the difference in heat flux effect is not less than the threshold X2, the target heat flux adjustment area is determined according to the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is the heat flux difference area that needs to be adjusted first in the heat flux dynamic diagram and the heat flux target diagram. The target heat flux adjustment area is simulated based on the digital twin model to obtain the temperature adjustment method of the target heat flux adjustment area. The temperature control adjustment method is output. Step SA3: Obtain the new heat flow dynamics and return to step SA2.

9. The intelligent control system for the operation of computer room infrastructure according to claim 1, characterized in that, Infrastructure assessment based on real-time adjustment records and heat flow dynamics includes: Based on the real-time identification of the basic monitoring data corresponding to the heat flow dynamic diagram according to the data of the computer room monitoring, the heat flow dynamic diagram is evaluated based on the basic monitoring data to obtain the heat flow evaluation result; When the heat flow assessment result indicates an abnormal heat flow, the cause of the abnormal heat flow shall be determined based on the aforementioned heat flow dynamic diagram. When the heat flow assessment result is that the heat flow is normal, no corresponding action is taken; Identify the control information in the real-time adjustment record, obtain the heat flow dynamic diagram corresponding to the control information, evaluate the heat flow change based on the heat flow dynamic diagram and the control information, and obtain the heat flow change evaluation result. When the heat flow change is assessed as normal, no corresponding action is taken; When assessing abnormal heat flow changes, the cause of the abnormal heat flow changes is determined based on the heat flow dynamic diagram and control information. The results of heat flow assessment, the causes of heat flow anomalies, the results of heat flow change assessment, and the causes of heat flow change anomalies are integrated into the facility assessment results.

10. The intelligent control system for the operation of computer room infrastructure according to claim 9, characterized in that, Assessment of heat flux changes based on heat flux dynamics diagrams and control information includes: The heat flow simulation data is obtained by performing simulation based on the control information using a preset digital twin model; the heat flow dynamic diagram is calibrated based on the heat flow simulation data to obtain the heat flow change evaluation result.

Images