Data center air conditioning system and energy efficiency optimization control method thereof

Abstract

The application discloses a data center air conditioning system and an energy efficiency optimization control method thereof, relates to the technical field of data center environment control, and specifically relates to a data center air conditioning system and an energy efficiency optimization control method thereof. The system comprises an environment sensing module, a cold and hot channel isolation unit, a multi-mode air conditioning unit and a central control unit. The environment sensing module collects data room temperature and humidity, airflow and equipment heat dissipation data; the cold and hot channel isolation unit uses an adjustable flow guide plate and a heat insulation curtain to guide the airflow; the multi-mode air conditioning unit comprises a fixed frequency and variable frequency compressor combined refrigeration unit, a sensible heat / total heat exchange unit and an auxiliary heat dissipation unit; and the central control unit analyzes data and sends control instructions. The method comprises basic data collection, heat load characteristic analysis, operation mode matching, dynamic parameter adjustment and closed loop verification and correction, adjusts parameters according to low, medium and high load modes of the heat load, and realizes energy efficiency optimization.

Description

Technical Field

[0001] This invention relates to the field of data center environmental control technology, specifically to data center air conditioning systems and their energy efficiency optimization control methods. Background Technology

[0002] With the rapid development of information technology and the widespread adoption of emerging applications such as cloud computing, big data, and artificial intelligence, the number and scale of data centers worldwide continue to expand. As the core infrastructure supporting various digital businesses, data centers house a large number of high-performance servers, storage devices, and network switching equipment, which continuously generate significant amounts of heat during operation. To ensure stable equipment operation, data centers must maintain a constant temperature and humidity environment, making air conditioning systems a crucial component of data center energy consumption. Statistics show that the energy consumption of traditional data center air conditioning systems can account for 30% to 40% of their total energy consumption, and even higher in some high-density deployment scenarios. Therefore, the energy efficiency level of air conditioning systems directly affects the operating costs and environmental sustainability of data centers.

[0003] Currently, most data centers use room-level or row-level air conditioning to dissipate heat through natural diffusion of hot and cold air or simple mechanical guidance. However, this traditional approach has significant drawbacks: First, hot and cold air can easily mix within the server room, creating airflow short-circuiting and resulting in some areas being overheated while others are overcooled, affecting equipment reliability and wasting energy. Second, the operating modes of air conditioning equipment are relatively fixed, often running continuously according to preset parameters, making it difficult to precisely adjust based on real-time changes in heat load. For example, running the compressor at full load during low-load periods or failing to increase cooling capacity in localized high-heat areas will reduce the overall energy efficiency of the system.

[0004] Regarding airflow organization, while existing data centers generally adopt the basic concept of hot and cold aisle isolation, most isolation measures have a simple structure, such as fixed baffles or simple curtains, which cannot dynamically adjust the guidance direction according to the airflow distribution, resulting in a mismatch between the airflow path and the actual heat dissipation requirements of the equipment. At the same time, the monitoring of environmental parameters is mostly limited to a few key points, making it difficult to comprehensively reflect the temperature gradient, humidity differences, and airflow status of different areas of the computer room. This results in the central control system lacking sufficiently detailed data support when making load analysis and control decisions.

[0005] In terms of air conditioning unit configuration, common solutions mainly rely on a single type of compressor or a fixed heat exchange mode, lacking adaptability to different load stages. For example, fixed-frequency compressors are inefficient and energy-intensive under light loads, while heat exchangers often operate in full-heat exchange mode for extended periods, exchanging humidity even when only temperature regulation is required, thus consuming additional energy. Furthermore, auxiliary cooling units (such as liquid-cooled terminals or external fans) are often independently controlled and do not operate in coordination with the main air conditioning system, making it difficult to respond quickly during peak loads.

[0006] At the control strategy level, most systems rely on static thresholds or manual experience to set operating parameters, lacking the ability to analyze the spatiotemporal distribution characteristics of heat load in real time, and failing to form a closed-loop optimization mechanism of "sensing-analysis-control-verification". When the layout of the computer room or the heating characteristics of the equipment change, the original control logic is prone to deviation, leading to decreased energy efficiency or increased risk of local overheating.

[0007] In summary, traditional data center air conditioning systems have limitations in terms of airflow organization flexibility, comprehensive environmental perception, diverse equipment operation modes, and intelligent control strategies, making it difficult to meet the energy-saving and temperature control requirements of modern high-density, high-reliability data centers. Therefore, a new type of air conditioning system and its energy efficiency optimization method are needed, capable of accurately sensing multi-dimensional environmental parameters, dynamically adjusting the hot and cold aisle isolation structure, flexibly configuring multi-mode cooling and heat exchange units, and performing closed-loop optimization control based on real-time heat load characteristics, in order to significantly reduce data center energy consumption and improve operational stability. Summary of the Invention

[0008] The purpose of this invention is to provide a data center air conditioning system and its energy efficiency optimization control method. Through environmental perception, hot and cold aisle isolation, multi-mode air conditioning unit coordination, and central control, precise energy efficiency optimization is achieved, heat dissipation efficiency is improved, energy consumption of the air conditioning system is reduced, and stable operation of the equipment is ensured.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a data center air conditioning system and its energy efficiency optimization control method, comprising: The environmental sensing module is distributed in different areas of the computer room. Each area is equipped with at least one set of sensors. Each set includes an ambient temperature sensor, an ambient humidity sensor, an airflow direction sensor, and an airflow speed sensor. These sensors are used to collect data on ambient temperature, humidity, airflow status (including airflow direction and speed), and actual heat dissipation temperature of the equipment. The hot and cold aisle isolation unit is set between the hot and cold aisles formed by two adjacent rows of cabinets. It includes a guide plate that can be adjusted around the horizontal axis and an openable heat insulation curtain arranged longitudinally along the top of the aisle. The angle adjustment range of the guide plate is 0°-90°. When the heat insulation curtain is fully extended, it can cover more than 90% of the cross-sectional area of ​​the aisle. It is used to guide the airflow along a predetermined path of air intake at the front and air exhaust at the back of the cabinet and reduce the mixing of hot and cold air. The multi-mode air conditioning unit includes a refrigeration unit with a fixed-frequency compressor and an inverter compressor connected in parallel, a heat exchange unit with a built-in sensible heat / total heat exchange core and the ability to switch between sensible heat / total heat exchange modes, and an auxiliary heat dissipation unit (which is an air-cooled cooling fan or a water-cooled cooling coil) that starts and stops according to heat load requirements. The central control unit receives real-time data transmitted by the environmental sensing module, processes the data using a preset heat load distribution characteristic analysis algorithm, identifies spatial distribution differences of heat load and abnormal airflow organization characteristics in the computer room, generates control commands including target angle of the guide vane, target opening and closing degree of the heat insulation curtain, target operating frequency of the compressor, target working mode of the heat exchange unit, and target start and stop status of the auxiliary heat dissipation unit, and sends them to the cold and hot aisle isolation unit and each air conditioning unit component.

[0010] Furthermore, the environment perception module includes: Temperature and humidity sensors are installed at the top of each rack, 100-200mm from the top surface of the rack, and at the bottom of each floor, 300-500mm directly above the air vent, to collect the ambient temperature and humidity at the corresponding locations. Airflow velocity sensors are installed in the middle of the side wall of the channel at the junction of the hot and cold channels. At least two sensors are installed in each junction area and are arranged diagonally to measure the lateral airflow velocity in the area where hot and cold air meet. The server rack exhaust temperature sensor is embedded in the back panel of the rack near the edge of the exhaust vent. One sensor is installed on each rack back panel to collect the actual temperature of the hot air exhausted from the server.

[0011] Furthermore, the guide plate of the hot and cold aisle isolation unit is angle-adjusted by a stepper motor. The stepper motor is fixed on the mounting bracket on the side of the guide plate, and the motor output shaft is connected to the guide plate shaft through a gear pair. The angle adjustment accuracy is ±1°. The heat insulation curtain is made of fireproof soft curtain and is opened and closed by an electric slide rail. The electric slide rail is installed along the guide rail at the top of the hot and cold aisle. The two sides of the heat insulation curtain are fixedly connected to the sliders in the slide rail. The opening and closing stroke covers the entire length of the aisle, and the opening and closing speed is 0.5-1.0m / s.

[0012] Furthermore, the refrigeration unit of the multi-mode air conditioning unit includes at least one fixed-frequency compressor and at least two variable-frequency compressors connected in parallel. The rated power of the fixed-frequency compressor is 50-100kW, and the rated power of a single variable-frequency compressor is 30-60kW and the speed can be adjusted within the frequency range of 20-100Hz. All compressors are connected to the condenser and evaporator through the same refrigerant pipeline to form a parallel refrigeration circuit.

[0013] Furthermore, the heat exchange unit of the multi-mode air conditioning unit has a built-in sensible heat / total heat exchange core. The sensible heat / total heat exchange core is a cross-flow metal foil heat exchanger structure. In sensible heat mode, the core only allows heat driven by temperature gradient to pass through. In total heat exchange mode, the core transfers heat and moisture driven by temperature gradient and humidity gradient at the same time. Mode switching is achieved by controlling the opening and closing of the air valves on both sides of the core. The switching response time does not exceed 30 seconds.

[0014] Furthermore, it includes the following steps: Basic data acquisition: The environmental sensing module collects temperature and humidity, airflow speed and equipment outlet temperature data of each area according to the set period, and records the current operating status of the air conditioning unit (including the start and stop status of each compressor, the working mode of the heat exchange unit, the start and stop status of the auxiliary heat dissipation unit and the fan speed). Heat load characteristic analysis: The central control unit classifies and statistically analyzes the collected data, divides the area into units according to the rack rows, identifies areas with heat load values ​​per unit area higher than the average as high heat density areas, areas where the airflow velocity direction deviates from the predetermined path by more than 30° as airflow short-circuit areas, and areas where the difference between the highest and lowest temperatures in the same area is greater than 5℃ as heat and cold imbalance areas. The heat load value per unit area (unit: W / m²) and airflow efficiency value (airflow efficiency value is the ratio of the effective airflow through the rack to the total air supply volume) of each area are calculated. Operating mode matching: Based on the heat load characteristic analysis results, the current operating conditions are divided into three modes: low load equilibrium state, medium load zone adjustment state, and high load enhancement state; Dynamic parameter adjustment: Adjust the angle of the guide plate of the hot and cold aisle isolation unit, the opening and closing status of the heat insulation curtain, and the compressor frequency, heat exchange mode, and start / stop status of the auxiliary heat dissipation unit for different modes; Closed-loop verification and correction: After adjustment, data is re-acquired, and the deviation between the target temperature range (target temperature range is 22-27℃) and the actual temperature is compared. If the deviation exceeds 2℃ for three consecutive cycles, the heat load characteristic identification error is analyzed back and the mode division logic is adjusted.

[0015] Furthermore, the basic data collection period is set to 10-60 seconds, and the mean filtering process is performed on five consecutive sets of data. The filtered data is the arithmetic mean of the five sets of data. After removing abnormal data that deviates from the mean by more than 20% in a single collection, the average is taken again.

[0016] Furthermore, in the operation mode matching, the determination condition for low load equilibrium state is: the heat load value per unit area of ​​more than 80% of the area is less than 200W / m² and the airflow efficiency value is higher than 0.8; the determination condition for medium load zoning adjustment state is: there is at least one area with a heat load value per unit area between 200-800W / m² and an airflow efficiency value less than 0.8; the determination condition for high load enhancement state is: the heat load value per unit area of ​​the entire computer room exceeds 800W / m² and the duration exceeds 5 minutes.

[0017] Furthermore, in the dynamic parameter adjustment, the operation in the low-load balanced state includes: shutting down the fixed-frequency compressors of some redundant air conditioning units (keeping at least one fixed-frequency compressor running), switching the heat exchange unit to sensible heat mode, adjusting the baffle angle to 40°-50° to balance the air supply of each cabinet, and at the same time reducing the fan speed of the air conditioning unit to 60%-70% of the rated speed.

[0018] Furthermore, in the dynamic parameter adjustment, the operation in the high-load enhanced state includes: starting the auxiliary heat dissipation unit, switching the heat exchange unit to the full heat exchange mode, reducing the angle of the guide vanes in non-critical areas to 10°-30° to reduce ineffective airflow, and simultaneously increasing all variable frequency compressors to operate at a frequency of 80-100Hz, while all fixed frequency compressors are turned on.

[0019] This invention provides a data center air conditioning system and its energy efficiency optimization control method, which has the following beneficial effects: 1. Precise Environmental Sensing Enhances Targeted Control: The environmental sensing module deploys temperature, humidity, airflow velocity, and equipment outlet temperature sensors at key locations such as the top of the server rack, underfloor air vents, hot and cold aisle junctions, and the back panel of the server rack. This enables fine-grained collection of environmental parameters in different areas of the data center. This distributed layout accurately detects anomalies such as localized heat islands and airflow short circuits, avoiding blind spots in control caused by traditional single-point monitoring. The central control unit identifies heat load distribution characteristics based on real and comprehensive on-site data, making subsequent operations such as adjusting the angle of isolation units and switching air conditioning unit modes more aligned with actual needs. This reduces ineffective control from the data source and significantly improves the accuracy of energy efficiency optimization.

[0020] 2. Hot and cold aisle isolation enhances airflow efficiency: The hot and cold aisle isolation unit uses a stepper motor-driven adjustable guide vane and an electrically controlled fireproof soft curtain, which can change the airflow guidance path and mixing degree as needed. The flexible adjustment of the guide vane angle can balance the air supply uniformity of different cabinets, and the opening and closing of the heat insulation curtain can block ineffective mixing when there is a local imbalance of hot and cold. Compared with a fixed isolation structure, this design can dynamically adapt to changes in heat load—for example, tightening the aisle in high-load areas to reduce cooling waste, and widening the aisle in low-load areas to reduce fan energy consumption, directly improving the utilization rate of cold air and reducing the extra energy consumed by the air conditioning system to offset mixing losses.

[0021] 3. Multi-mode air conditioning units achieve tiered energy consumption control: The refrigeration unit is configured with parallel fixed-frequency and variable-frequency compressors, and the heat exchange unit supports switching between sensible heat and total heat exchange core modes. Combined with an auxiliary cooling unit that starts and stops on demand, a tiered energy supply system of "basic cooling + dynamic energy replenishment" is formed. Under low load, redundant fixed-frequency compressors are shut down and sensible heat mode is switched to reduce the operating time of high-power components; under high load, auxiliary cooling is activated and total heat mode is switched to utilize waste heat from return air to improve heat exchange efficiency. This on-demand combined operation avoids the inefficiency of traditional single-mode units that are "overpowered," significantly reducing the overall energy consumption of the air conditioning system.

[0022] 4. Three-mode dynamic control to match scenario-based energy saving: Based on heat load characteristics, the operating conditions are divided into three modes: low, medium, and high load, and quantitative judgment standards are established (e.g., low load requires 80% of the area to have a heat load <200W / m² and an airflow efficiency >0.8). Differentiated adjustment strategies are implemented under different modes: low load equilibrium mode focuses on shutting down redundant equipment and balancing air supply; medium load zoned adjustment mode targets local hotspots for precise intervention; high load enhanced mode prioritizes heat dissipation in key areas. This closed-loop logic of "diagnosis-matching-control" changes the traditional "one-size-fits-all" operation mode of air conditioning systems, making energy allocation highly consistent with actual needs and avoiding ineffective energy consumption.

[0023] 5. Closed-loop verification mechanism ensures continuous optimization: After adjustment, average filtered data is collected at 10-60 second intervals to compare the deviation between the target temperature and the actual temperature. If the deviation continuously exceeds the threshold, the heat load identification error is traced back and the mode logic is adjusted. This mechanism forms a complete closed loop of "collection-analysis-adjustment-verification-correction," which can promptly detect control failures caused by interference factors such as sensor errors and sudden changes in heat load, and dynamically optimize the feature recognition algorithm and mode division rules. Compared with the problem of open-loop control being prone to failure due to environmental changes, closed-loop verification ensures that the system maintains a high-efficiency operating state in the long term, continuously improving the stability and reliability of energy efficiency optimization. Attached Figure Description

[0024] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0025] Figure 1 This is a diagram showing the overall architecture of the data center air conditioning system of the present invention. Figure 2 This is the main flowchart of the energy efficiency optimization control method of the present invention; Figure 3 This is a flowchart illustrating the basic data acquisition process of this invention. Figure 4 This is a flowchart of the operation mode matching and determination process of the present invention; Figure 5 This is a flowchart of the dynamic parameter adjustment process under low load equilibrium state according to the present invention. Detailed Implementation

[0026] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses consistent with some aspects of this disclosure as detailed in the appended claims.

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0028] How to use: I. Basic Data Collection and Environmental Sensing Deployment First, the environmental sensing module is deployed: temperature and humidity sensors are installed near the air vents on the top of the server rack and under the floor; airflow velocity sensors are deployed at the junction of hot and cold aisles; and the server rack exhaust temperature sensor is embedded in the rack back panel. After the system starts, the environmental sensing module automatically collects ambient temperature, humidity, airflow velocity, and equipment exhaust temperature data in each area at a set interval (e.g., 10-60 seconds). It performs mean filtering on multiple consecutive sets of data and simultaneously records the current operating status of the air conditioning unit, providing basic data support for subsequent analysis.

[0029] II. Heat Load Characteristics Analysis and Pattern Classification After receiving the collected data, the central control unit performs classification, statistics, and analysis, focusing on identifying areas with high heat density, airflow short-circuit areas, and areas of thermal imbalance. It calculates the heat load and airflow efficiency values ​​per unit area for each area, and based on this, classifies the current operating conditions into three modes: if more than 80% of the areas have a heat load per unit area below 200 W / m² and an airflow efficiency above 0.8, it is determined to be a low-load equilibrium state; if there are areas with a heat load per unit area between 200-800 W / m² and an airflow efficiency below 0.8, it is determined to be a medium-load zoned control state; if the overall heat load per unit area exceeds 800 W / m², it is determined to be a high-load enhanced state.

[0030] III. Dynamic Parameter Adjustment and Execution Based on the mode division results, the central control unit sends control commands to the hot and cold aisle isolation units and multi-mode air conditioning units: Low-load balanced state: The fixed-frequency compressors of some redundant air conditioning units are shut down, the heat exchange unit is switched to sensible heat mode (only exchanging temperature), and the adjustable guide vanes of the hot and cold aisle isolation unit are driven to 40°-50° by a stepper motor to balance the air supply of each cabinet; the heat insulation curtain is kept closed to reduce the mixing of hot and cold air.

[0031] Medium-load zone control: For areas with high heat density, adjust the angle of the deflector plate in the corresponding area to enhance local air supply, and appropriately reduce the angle of the deflector plate in non-critical areas; the air conditioning unit starts some variable frequency compressors according to the zone heat load, the heat exchange unit maintains sensible heat or switches to mode as needed, and the auxiliary heat dissipation unit is temporarily not used.

[0032] High-load enhanced mode: The auxiliary heat dissipation unit is activated, and the heat exchange unit switches to full heat exchange mode (exchanging temperature and humidity simultaneously); the angle of the deflector plate in non-critical areas is reduced to 10°-30° to reduce ineffective airflow, and the heat insulation curtain is fully closed by the electric sliding rail to enhance the isolation effect of the hot and cold aisles; the fixed-frequency and variable-frequency compressors of the multi-mode air conditioning unit work together to increase the cooling capacity.

[0033] IV. Closed-loop verification and continuous optimization After adjustment, the environmental sensing module re-collects data, and the central control unit compares the deviation between the target temperature range and the actual temperature. If the deviation exceeds the threshold for multiple consecutive cycles, the system automatically backtracks and analyzes the heat load characteristic identification error (such as abnormal sensor data or area division deviation), adjusts the mode division logic (such as correcting the heat load threshold or airflow efficiency judgment standard), and ensures that the control strategy continuously adapts to the actual thermal environment of the computer room to achieve the energy efficiency optimization goal. Example:

[0034] Example 1: System usage scenario under low load equilibrium state A data center server room is currently experiencing a light workload, with most server racks operating at stable loads and overall low cooling requirements, resulting in a low-load equilibrium state. The environmental sensing module's temperature and humidity sensors continuously collect data near the air vents on the top of the server racks and under the floor. Airflow velocity sensors monitor airflow at the junction of hot and cold aisles, and server rack exhaust temperature sensors provide real-time feedback on the actual cooling temperature of the equipment. All data is synchronized to the central control unit at set intervals. Analysis by the central control unit reveals that over 80% of the area has a heat load per unit area below the threshold and high airflow efficiency, confirming the current low-load equilibrium state.

[0035] Control commands were immediately issued: the adjustable guide vanes of the hot and cold aisle isolation units were driven to an angle of 40°-50° by stepper motors, ensuring even airflow across the air intake surfaces of each cabinet; the heat-insulating curtains made of fireproof soft curtains were kept closed by electric sliding rails, effectively preventing the mixing of hot and cold air. In the multi-mode air conditioning unit, some redundant fixed-frequency compressors were shut down to reduce ineffective energy consumption; the heat exchange unit switched to sensible heat mode, performing only temperature exchange to reduce the additional load from humidity handling; the variable-frequency compressor maintained its base frequency operation, and in conjunction with the strategy of temporarily not starting the auxiliary cooling unit, an energy-saving operating state was achieved. During the closed-loop verification phase, the environmental sensing module re-collected data, and the deviation between the target temperature range and the actual temperature was small. The system maintained stable operation in the current mode, achieving energy efficiency optimization under low-load scenarios.

[0036] Example 2: System usage scenario under medium load zone regulation During a certain period, the addition of new computing nodes led to increased load in some areas of the server room, resulting in localized high heat density and airflow short-circuiting. The system determined this to be a medium-load zone adjustment. The environmental sensing module's sensors continued to operate. The temperature and humidity sensors detected the temperature rise trend in the high-heat area, the airflow velocity sensor detected abnormally low-speed airflow (reduced airflow efficiency) at the junction of the hot and cold aisles, and the server rack exhaust temperature sensor reported that the equipment in that area was dissipating heat at a higher temperature. The data was aggregated and sent to the central control unit.

[0037] The central control unit identified high-heat-density areas and airflow short-circuit areas. Based on heat load characteristics analysis, it determined that zoned adjustments were necessary: ​​the hot and cold aisle isolation unit increased the angle of the adjustable deflectors in these areas using stepper motors to enhance local airflow guidance; the deflectors in non-critical areas had their angles appropriately reduced to avoid airflow waste. In the multi-mode air conditioning unit, the inverter compressors in high-heat-density areas increased their operating frequency to enhance cooling capacity; the heat exchange unit maintained sensible heat mode to simplify control, and the auxiliary heat dissipation unit was temporarily deactivated. Insulation curtains were partially closed according to regional needs to further isolate hot and cold airflows. Closed-loop verification showed that the temperature in high-heat-density areas gradually approached the target range, airflow efficiency recovered, and the system balanced local heat load and overall energy efficiency through zoned adjustments.

[0038] Example 3: System usage scenario under high load enhancement state The data center experienced a surge in business activity, causing a sharp increase in overall load. The heat load per unit area exceeded the high-load threshold, and the system entered a high-load intensified state. The temperature and humidity sensors in the environmental sensing module detected a rapid rise in the overall temperature and humidity of the server room. The airflow velocity sensor showed increased airflow turbulence in the hot and cold aisles. The temperature sensors at the server rack exhaust vents indicated that the cooling temperatures of multiple equipment areas were approaching their upper limits. The data was then processed by averaging and filtering before being sent to the central control unit.

[0039] The central control unit urgently activated enhanced control: the adjustable guide vanes of the hot and cold aisle isolation unit were lowered to an angle of 10°-30° in non-critical areas via stepper motors to reduce ineffective airflow diffusion; the fireproof soft curtains were fully closed via electric sliding rails to maximize the prevention of hot and cold air mixing. In the multi-mode air conditioning unit, the fixed-frequency compressor and the variable-frequency compressor operated in parallel at full load to increase the total cooling capacity; the heat exchange unit switched to full heat exchange mode, simultaneously processing temperature and humidity to optimize heat exchange efficiency; the auxiliary heat dissipation unit was activated simultaneously to supplement additional heat dissipation capacity. During the closed-loop verification phase, despite the high heat load, the deviation between the actual temperature and the target range did not exceed the threshold. The system achieved energy efficiency balance under high load while ensuring equipment safety through the collaboration of multiple components.

[0040] Example 4: Application Scenarios for Airflow Short-Circuit Region Identification and Correction A server room experienced a localized airflow short circuit after rack layout adjustments, causing cool air to be drawn away by return air before effectively flowing through the equipment. The system resolved this issue through an environmental sensing and control process. The airflow velocity sensor in the environmental sensing module detected abnormal airflow direction (uneven velocity distribution) at the junction of the hot and cold aisles. The temperature and humidity sensors showed that the temperature in the short-circuit area was low, but the temperature at the equipment outlet was high. Data from the server rack outlet temperature sensor confirmed poor equipment heat dissipation, and the data was uploaded to the central control unit.

[0041] The central control unit analyzed and identified an airflow short-circuit area, classifying it as a medium-load zone (with areas exhibiting reduced airflow efficiency). Control commands instructed: the adjustable guide vanes of the hot and cold aisle isolation unit in this area should be angled via stepper motors to guide airflow along the front of the cabinet, preventing direct short circuits; the thermal insulation curtains should be partially closed to enhance isolation. The multi-mode air conditioning units should increase the number of inverter compressors operating in this area to increase airflow intensity; the heat exchange unit should maintain sensible heat mode, and the auxiliary cooling unit should remain off. After adjustments, data collected again showed that airflow efficiency had recovered, and the equipment outlet temperature was approaching the target. The system, by accurately identifying airflow anomalies and adjusting isolation and cooling parameters, resolved the energy efficiency loss caused by the short circuit.

[0042] Example 5: Application Scenarios for Cold and Heat Imbalance Area Control and Model Backtracking A temporary sensor malfunction in a computer room caused a deviation in heat load characteristic identification. Initially, it was judged to be in a low-load equilibrium state, but in reality, there was a localized temperature imbalance. The system triggered a backtracking correction through closed-loop verification. The environmental sensing module collected data periodically (including some outliers during the malfunction). The central control unit initially analyzed the data and misjudged it as a low-load equilibrium state, executing operations such as shutting down redundant fixed-frequency compressors and adjusting the deflector to 40°-50°. However, closed-loop verification found that the actual temperature deviated from the target range by more than a threshold for several consecutive cycles. The system automatically backtracked and analyzed the data, identifying the heat load characteristic identification error (due to inaccurate area division caused by abnormal sensor data).

[0043] The central control unit adjusted its mode division logic and reassessed the heat load and airflow efficiency. The revised analysis revealed areas of thermal imbalance (reduced airflow efficiency and locally high heat load), classifying these areas as medium-load zone control. New control commands were issued: the hot and cold aisle isolation unit specifically adjusted the angle of the deflector plates in the imbalance area, and optimized the closing status of the thermal insulation curtains; multi-mode air conditioning units started some inverter compressors and switched heat exchange modes, while auxiliary cooling units prepared as needed. After further verification, the temperature deviation returned to the normal range. The system corrected the identification error through a closed-loop mechanism, ensuring that the control strategy was consistent with the actual thermal environment, demonstrating the system's adaptive optimization capabilities.

[0044] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A data center air conditioning system and its energy efficiency optimization control method, characterized in that, include: The environmental sensing module is distributed in different areas of the computer room and is used to collect data on ambient temperature, humidity, airflow status and actual heat dissipation temperature of the equipment. The hot and cold aisle isolation unit, installed between rack rows, includes an adjustable-angle baffle and an openable heat insulation curtain, used to guide airflow along a predetermined path and reduce the mixing of hot and cold air; The multi-mode air conditioning unit includes a refrigeration unit combining a fixed-frequency compressor and a variable-frequency compressor, a heat exchange unit that can switch between sensible heat and total heat exchange modes, and an auxiliary heat dissipation unit that can be started and stopped as needed. The central control unit receives data from the environmental sensing module, analyzes the heat load distribution characteristics, generates and sends control commands to the hot and cold aisle isolation unit and each air conditioning unit component.

2. The data center air conditioning system and its energy efficiency optimization control method according to claim 1, characterized in that, The environment sensing module includes: Temperature and humidity sensors are placed at the top of the cabinet and near the air vents under the floor. An airflow velocity sensor is installed at the junction of the hot and cold aisles. The server rack exhaust temperature sensor is embedded in the rack back panel.

3. The data center air conditioning system and its energy efficiency optimization control method according to claim 1, characterized in that: The guide plate of the hot and cold aisle isolation unit is angle-adjusted by a stepper motor, and the heat insulation curtain is made of fireproof soft curtain and is opened and closed by an electric slide rail.

4. The data center air conditioning system and its energy efficiency optimization control method according to claim 1, characterized in that: The refrigeration unit of the multi-mode air conditioning unit includes at least one fixed-frequency compressor and at least two variable-frequency compressors connected in parallel.

5. The data center air conditioning system and its energy efficiency optimization control method according to claim 1, characterized in that: The heat exchange unit of the multi-mode air conditioning unit has a built-in sensible heat / total heat exchange core, which can switch between exchanging only temperature or exchanging both temperature and humidity at the same time.

6. The data center air conditioning system and its energy efficiency optimization control method according to any one of claims 1 to 5, characterized in that, Includes the following steps: Basic data acquisition: The environmental sensing module collects temperature and humidity, airflow speed and equipment outlet temperature data of each area according to a set period, and records the current operating status of the air conditioning unit in a synchronous manner; Heat load characteristic analysis: The central control unit classifies and statistically analyzes the collected data, identifies areas with high heat density, airflow short-circuit areas, and areas of heat-cold imbalance, and calculates the heat load value per unit area and airflow efficiency value for each area; Operating mode matching: Based on the heat load characteristic analysis results, the current operating conditions are divided into three modes: low load equilibrium state, medium load zone adjustment state, and high load enhancement state; Dynamic parameter adjustment: Adjust the angle of the guide plate of the hot and cold aisle isolation unit, the opening and closing status of the heat insulation curtain, and the compressor frequency, heat exchange mode, and start / stop status of the auxiliary heat dissipation unit for different modes; Closed-loop verification and correction: After adjustment, data is re-acquired, and the deviation between the target temperature range and the actual temperature is compared. If the deviation exceeds the threshold for multiple consecutive cycles, the heat load characteristic identification error is analyzed back and the mode division logic is adjusted.

7. The data center air conditioning system and its energy efficiency optimization control method according to claim 6, characterized in that: The basic data collection period is set to 10-60 seconds, and mean filtering is performed on multiple consecutive sets of data.

8. The data center air conditioning system and its energy efficiency optimization control method according to claim 6, characterized in that: In the operation mode matching, the criteria for determining the low load equilibrium state are: the heat load per unit area of ​​more than 80% of the area is less than 200W / m² and the airflow efficiency is greater than 0.8; the criteria for determining the medium load zone adjustment state are: the heat load per unit area of ​​some areas is between 200-800W / m² and the airflow efficiency is less than 0.8; the criteria for determining the high load enhancement state are: the overall heat load per unit area exceeds 800W / m².

9. The data center air conditioning system and its energy efficiency optimization control method according to claim 6, characterized in that: In the dynamic parameter adjustment, the operation in the low-load balanced state includes: shutting down the fixed-frequency compressors of some redundant air conditioning units, switching the heat exchange unit to sensible heat mode, and adjusting the angle of the baffle plate to 40°-50° to balance the air supply of each cabinet.

10. The data center air conditioning system and its energy efficiency optimization control method according to claim 6, characterized in that: In the dynamic parameter adjustment, the operation in the high-load enhanced state includes: starting the auxiliary heat dissipation unit, switching the heat exchange unit to the full heat exchange mode, and reducing the angle of the guide vane in non-critical areas to 10°-30° to reduce ineffective airflow.

Images