On-demand refrigeration method for energy-saving inter-row and inter-column high-precision cabinet air conditioner

Abstract

The invention relates to the technical field of data center air conditioner refrigeration, in particular to an on-demand refrigeration method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner. Comprising the following steps that dynamic thermal load parameters of a refrigeration area, operation state parameters of key equipment in a cabinet and operation parameters of an air conditioning system are collected in real time; on the basis of the parameters, a real-time overall refrigeration demand is obtained through calculation of a preset refrigeration demand model; decomposing the total refrigeration demand into an inter-row air conditioner refrigeration demand and an inter-column air conditioner refrigeration demand based on a preset distribution strategy; correspondingly generating an inter-row air conditioner control instruction and an inter-column air conditioner control instruction according to the inter-row air conditioner refrigeration requirement and the inter-column air conditioner refrigeration requirement so as to dynamically adjust the running states of the inter-row air conditioner and the inter-column air conditioner and achieve refrigeration according to requirements; the invention provides an on-demand refrigeration method based on multi-parameter dynamic sensing, real-time demand calculation and energy efficiency feedback so as to improve the system adaptability and the energy efficiency level.

Description

Technical Field

[0001] This invention relates to the field of data center air conditioning and refrigeration technology, and in particular to an on-demand refrigeration method for energy-saving inter-row and inter-column high-precision cabinet air conditioners. Background Technology

[0002] With the continuous expansion of data centers and communication equipment rooms, the density of internal IT equipment is constantly increasing, leading to a significant increase in the heat load on server racks. To ensure stable equipment operation, rack air conditioning systems need to have efficient and precise cooling capabilities. Traditional server room air conditioning systems often use fixed temperature settings or simple zoning control, which are difficult to adapt to dynamically changing heat load distribution, easily resulting in localized overheating or overcooling. In recent years, the coordinated application of in-row and inter-row air conditioning has gradually become an important direction for improving cooling efficiency. By placing air conditioning equipment close together between rack rows or columns, more precise airflow organization and temperature control can be achieved. However, existing control methods still have room for further improvement in terms of in-row and inter-row air conditioning coordination strategies, dynamic load response, and system energy efficiency optimization.

[0003] Currently, in practical applications, the control of inter-row and inter-column air conditioning systems largely relies on preset fixed logic or local temperature feedback, lacking a comprehensive assessment of the operating status of equipment within the cabinet, dynamic heat load distribution, and the performance of the air conditioning system itself. This control method struggles to match complex and changing thermal environments in real time, easily leading to a mismatch between cooling capacity and actual demand, resulting in energy waste. Furthermore, existing methods exhibit lag in response to local hotspots or heat island effects, and the allocation of cooling tasks between inter-row and inter-column air conditioning systems lacks global optimization, impacting the overall energy efficiency and cooling accuracy of the system. Summary of the Invention

[0004] This invention provides an on-demand cooling method based on multi-parameter dynamic sensing, real-time demand calculation, and energy efficiency feedback to improve system adaptability and energy efficiency.

[0005] The technical solution adopted in this invention is: an on-demand cooling method for energy-saving inter-row and inter-column high-precision cabinet air conditioners, comprising the following steps:

[0006] S1: Real-time acquisition of dynamic heat load parameters of the cooling area, operating status parameters of key equipment in the cabinet, and operating parameters of the air conditioning system itself;

[0007] S2: Based on the dynamic heat load parameters, operating status parameters, and its own operating parameters, the real-time overall cooling demand is calculated using a preset cooling demand model;

[0008] S3: Based on a preset allocation strategy, the total cooling demand is decomposed into inter-row air conditioning cooling demand and inter-column air conditioning cooling demand;

[0009] S4: Based on the inter-row air conditioning cooling demand and the inter-row air conditioning cooling demand, generate corresponding inter-row air conditioning control commands and inter-row air conditioning control commands respectively, so as to dynamically adjust the operating status of the inter-row air conditioning and inter-row air conditioning and realize on-demand cooling;

[0010] S5: Real-time monitoring of energy efficiency indicators after on-demand cooling, including the ratio of the total power consumption of the cooling system to the total power consumption of the IT equipment;

[0011] S6: Based on the energy efficiency index, optimize the parameters and allocation strategy in the cooling demand model.

[0012] As a further improvement of the present invention, in step S1, the dynamic heat load parameters include at least the rack inlet temperature, rack outlet temperature, rack hot spot temperature, and return air temperature; the operating status parameters include at least the server fan speed and CPU load rate; and the self-operating parameters include at least the air conditioner supply air temperature, compressor operating frequency, and fan speed.

[0013] As a further improvement of the present invention, in step S2, the method for constructing the cooling demand model is implemented by one of the following two methods:

[0014] Method 1: Based on historical and experimental data, establish a machine learning regression model with dynamic heat load parameters and operating status parameters as inputs and cooling capacity as output;

[0015] Method 2: Based on the principle of thermal balance, a physical model is established that includes the heat output of the equipment, air flow rate and temperature difference, and the heat output of the equipment is dynamically corrected through the operating state parameters.

[0016] As a further improvement of the present invention, in step S3, the preset allocation strategy is as follows: the in-row air conditioner shall take priority in handling the precise cooling demand for individual and local cabinets based on the temperature of the cabinet inlet and outlet; the portion of the total cooling demand that exceeds the cooling capacity of the in-row air conditioner, as well as the background cooling demand required to eliminate the heat island effect in the entire cooling area, shall be handled by the in-row air conditioner.

[0017] As a further improvement of the present invention, in step S3, the allocation strategy further includes: when the temperature of any of the cabinet hotspots exceeds the first safety threshold, increasing the cooling output of the inter-row air conditioner in the corresponding area; when the return air temperature exceeds the second safety threshold, increasing the cooling output of the inter-row air conditioner.

[0018] As a further improvement of the present invention, in step S4, the dynamic adjustment of the operating status of the inter-row air conditioner and the inter-row air conditioner specifically includes: for the inter-row air conditioner, controlling its air supply temperature and air volume by adjusting the compressor frequency, the opening degree of the electronic expansion valve and the fan speed; for the inter-row air conditioner, controlling its air outlet speed and air outlet temperature by adjusting the fan speed and the refrigerant flow rate.

[0019] As a further improvement of the present invention, in step S5, the energy efficiency index also includes the performance coefficient of the refrigeration system, and the overall energy efficiency level is evaluated by calculating the weighted comprehensive energy efficiency index of the power utilization efficiency and the performance coefficient.

[0020] As a further improvement of the present invention, in step S6, the feedback optimization adopts a recursive least squares algorithm to update the parameters in the cooling demand model online, and adopts fuzzy logic control to adjust the allocation strategy.

[0021] The beneficial effects of this invention are as follows: By constructing a closed-loop control system that integrates real-time sensing, dynamic modeling, intelligent allocation, and energy efficiency feedback, this invention achieves coordinated on-demand cooling between rows and columns of air conditioners. Its beneficial effects are: it can accurately match the dynamically changing heat load of the data center, effectively eliminate local hot spots and heat island effects, and significantly improve the overall energy efficiency of the cooling system while ensuring the safe operation of IT equipment, thus achieving a balance between high precision and energy saving. Attached Figure Description

[0022] Figure 1 This is a flowchart of an on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to the present invention. Detailed Implementation

[0023] To make the technical problems, technical solutions, and beneficial effects to be solved by this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the embodiments described herein are merely illustrative and are not intended to limit the scope of this application.

[0024] This invention provides an on-demand cooling method for energy-saving inter-row and inter-column high-precision cabinet air conditioners, comprising the following steps:

[0025] S1: Real-time acquisition of dynamic heat load parameters of the cooling area, operating status parameters of key equipment in the cabinet, and operating parameters of the air conditioning system itself;

[0026] S2: Based on the dynamic heat load parameters, operating status parameters, and its own operating parameters, the real-time overall cooling demand is calculated using a preset cooling demand model;

[0027] S3: Based on a preset allocation strategy, the total cooling demand is decomposed into inter-row air conditioning cooling demand and inter-column air conditioning cooling demand;

[0028] S4: Based on the inter-row air conditioning cooling demand and the inter-row air conditioning cooling demand, generate corresponding inter-row air conditioning control commands and inter-row air conditioning control commands respectively, so as to dynamically adjust the operating status of the inter-row air conditioning and inter-row air conditioning and realize on-demand cooling;

[0029] S5: Real-time monitoring of energy efficiency indicators after on-demand cooling, including the ratio of the total power consumption of the cooling system to the total power consumption of the IT equipment;

[0030] S6: Based on the energy efficiency index, optimize the parameters and allocation strategy in the cooling demand model.

[0031] In step S1 of this invention, the dynamic heat load parameters include at least the rack inlet temperature, rack outlet temperature, rack hot spot temperature, and return air temperature; the operating status parameters include at least the server fan speed and CPU load rate; and the self-operating parameters include at least the air conditioner supply air temperature, compressor operating frequency, and fan speed.

[0032] In step S2 of this invention, the method for constructing the cooling demand model is implemented in one of the following two ways:

[0033] Method 1: Based on historical and experimental data, establish a machine learning regression model with dynamic heat load parameters and operating status parameters as inputs and cooling capacity as output;

[0034] Method 2: Based on the principle of thermal balance, a physical model is established that includes the heat output of the equipment, air flow rate and temperature difference, and the heat output of the equipment is dynamically corrected through the operating state parameters.

[0035] In step S3 of this invention, the preset allocation strategy is as follows: the in-row air conditioner is given priority to handle the precise cooling demand for individual and local cabinets based on the temperature of the cabinet inlet and outlet. The portion of the overall cooling demand that exceeds the cooling capacity of the in-row air conditioner, as well as the background cooling demand required to eliminate the heat island effect in the entire cooling area, is handled by the in-row air conditioner.

[0036] In step S3 of the present invention, the allocation strategy further includes: when the temperature of any of the cabinet hotspots exceeds the first safety threshold, increasing the cooling output of the inter-row air conditioner in the corresponding area; and when the return air temperature exceeds the second safety threshold, increasing the cooling output of the inter-row air conditioner.

[0037] In step S4 of this invention, the dynamic adjustment of the operating status of the inter-row air conditioner and the inter-row air conditioner specifically includes: for the inter-row air conditioner, controlling its air supply temperature and air volume by adjusting the compressor frequency, the opening degree of the electronic expansion valve and the fan speed; for the inter-row air conditioner, controlling its air outlet speed and air outlet temperature by adjusting the fan speed and refrigerant flow rate.

[0038] In step S5 of this invention, the energy efficiency index also includes the performance coefficient of the refrigeration system, and the overall energy efficiency level is evaluated by calculating the weighted comprehensive energy efficiency index of the power utilization efficiency and the performance coefficient.

[0039] In step S6 of this invention, the feedback optimization uses a recursive least squares algorithm to update the parameters in the cooling demand model online, and uses fuzzy logic control to adjust the allocation strategy.

[0040] Example:

[0041] The following is in conjunction with the appendix Figure 1 Taking a specific data center cooling scenario as an example, the implementation of the method described in this invention will be explained in detail.

[0042] This embodiment is implemented in a standard data center server room, which is arranged with multiple server racks forming a rack row. In-row air conditioners are deployed between the rack rows, and in-row air conditioners are deployed adjacent to the racks.

[0043] Step 1: Real-time acquisition of multiple parameters

[0044] The system collects the following three types of parameters in real time through a sensor network deployed in the computer room:

[0045] ① Dynamic thermal load parameters: Temperature sensors are installed at the air inlet and outlet of each rack to collect the rack air inlet temperature (25℃) and rack air outlet temperature (32℃); sensors are installed in several high-power server chassis to collect the rack hot spot temperature (38℃); sensors are installed in the common area of ​​the computer room to collect the return air temperature (28℃).

[0046] ② Operating status parameters: Through the data center environmental monitoring system and server management interface, the fan speed (up to 80% of the highest setting) and CPU load rate (suddenly increase to 70%) of key servers are obtained in real time.

[0047] ③Self-operating parameters: Read the air supply temperature (18℃), compressor operating frequency (45Hz) and fan speed (1500rpm) in real time from the controller of the inter-row and inter-column air conditioners.

[0048] Step 2: Calculate the total cooling demand

[0049] The collected multi-source data is input into a preset cooling demand model. This embodiment uses the physical model described in Method 2 for calculation:

[0050] The model first dynamically estimates the real-time heat generation of the IT equipment based on the device's rated power and the current CPU load rate (70%).

[0051] The measured airflow (calculated from the server fan speed and air conditioner fan speed) and the temperature difference between the rack inlet and outlet (7℃) were used to verify the calculations using the heat balance formula Q=Cp*ρ*V*ΔT (where Q is the cooling capacity, Cp is the specific heat capacity of air, ρ is the air density, V is the air volume flow rate, and ΔT is the temperature rise).

[0052] The model ultimately calculated that the current real-time total cooling demand required to maintain the target temperature of the computer room is 85kW.

[0053] Step 3: Allocate inter-row / inter-column cooling requirements

[0054] The system decomposes the total cooling demand of 85kW according to a preset allocation strategy:

[0055] Precise cooling demand allocation: The system identifies that the localized cooling demand of most racks (totaling 60kW) can be met by in-row air conditioners. Therefore, this 60kW cooling demand is prioritized and allocated to the in-row air conditioners for precise and localized cooling.

[0056] Background and excess demand allocation: The portion of the total demand of 85kW that exceeds the current rated capacity of the in-row air conditioners (25kW), as well as the background cooling required to eliminate the heat island effect that may be caused by uneven airflow organization, totaling 25kW of cooling demand, is allocated to the in-row air conditioners.

[0057] Safety threshold intervention: During the allocation process, the system detected that the temperature of a hot spot in a cabinet (38°C) exceeded the set first safety threshold (35°C). In response, the allocation strategy was immediately triggered, increasing the cooling output of the inter-row air conditioners in that area to quickly suppress the local hot spot.

[0058] Step 4: Execute control commands

[0059] Based on the decomposed cooling requirements, the system generates specific control commands and issues them to the corresponding air conditioning equipment: ① For inter-row air conditioners, the system generates control commands including increasing the compressor frequency to 50Hz, fine-tuning the electronic expansion valve opening, and increasing the fan speed, aiming to increase the air volume and reduce the air supply temperature to achieve a cooling task of 25kW. ② For inter-row air conditioners, the system generates control commands including increasing the fan speed and adjusting the refrigerant flow, aiming to increase the outlet air velocity and optimize heat exchange efficiency to ensure the completion of a precise cooling task of 60kW, and to focus on cooling hot areas.

[0060] Through the above instructions, the operating status of the inter-row and inter-column air conditioners is dynamically adjusted, jointly achieving on-demand cooling that is precisely matched with the heat load.

[0061] Step 5: Monitor energy efficiency indicators

[0062] After the system has been running stably for a period of time, monitor energy efficiency indicators in real time:

[0063] The system calculates that the total power consumption of the current cooling system is 28kW and the total power consumption of the IT equipment is 150kW. Therefore, the core energy efficiency index PUE (Power Usage Effectiveness) = (150+28) / 150≈1.19.

[0064] At the same time, the system also calculates the coefficient of performance (COP) of the refrigeration system itself, and further derives a weighted comprehensive energy efficiency index, which is used to comprehensively evaluate the overall energy efficiency level after this control action.

[0065] Step 6: Feedback Optimization Model and Strategy

[0066] The system initiates a feedback optimization mechanism: ① It employs a recursive least squares algorithm to compare the actual results of the current run (such as the final cooling rate, temperature stability, and actual power consumption) with the model's predicted values, updating key parameters in the cooling demand model (such as the heat exchange efficiency coefficient) online to make the next demand calculation more accurate. ② It uses fuzzy logic control to fine-tune the thresholds and response strengths in the allocation strategy based on the current energy efficiency indicators (such as whether the PUE value has reached its optimal level) and the safety response effect (such as the hotspot elimination speed). For example, it optimizes the temperature threshold for triggering inter-row air conditioning capacity expansion or adjusts the load allocation ratio between row and inter-row air conditioning units.

[0067] As can be seen from the above embodiments, the present invention achieves continuous iterative upgrading of the cooling demand model and allocation strategy through the above feedback optimization mechanism.

[0068] Under the influence of the recursive least squares algorithm, the cooling demand model can dynamically correct its internal parameters based on actual operating data. For example, when a systematic deviation between the actual cooling capacity and the model's predicted value is detected, the algorithm will automatically adjust the air specific heat capacity or density correction coefficient in the heat balance formula, or optimize the weights of each input parameter in the machine learning model, thereby improving the model's adaptability to complex thermal environment changes and its prediction accuracy.

[0069] For optimizing the allocation strategy, fuzzy logic control demonstrates excellent handling capabilities for uncertainties. The system compares energy efficiency indicators (such as a PUE value of 1.19) with preset energy efficiency targets (such as PUE ≤ 1.20). If the PUE value is close to or slightly higher than the target value, the load allocation ratio between inter-row and inter-row air conditioners is adjusted through fuzzy rules. For example, the proportion of inter-row air conditioners in the overall cooling demand is appropriately increased to leverage their precise cooling advantage and further reduce overall energy consumption. If a hotspot elimination speed is detected to be slow, the hotspot temperature weighting factor in the fuzzy control is enhanced, enabling inter-row air conditioners to respond more quickly and increase cooling output when hotspots appear.

[0070] Through this dynamic feedback optimization process, the prediction accuracy of the cooling demand model and the adaptability of the allocation strategy are continuously improved, thereby ensuring that the entire cooling system can operate stably in a highly efficient, energy-saving and safe state for a long time, and achieving continuous optimization of data center cooling efficiency.

[0071] In summary, the on-demand cooling method for energy-saving inter-row and inter-column high-precision cabinet air conditioners of this invention, by constructing a closed-loop control system integrating multi-parameter real-time sensing, intelligent modeling and calculation, collaborative allocation decision-making, and dynamic energy efficiency optimization, not only effectively solves the industry pain points of insufficient cooling precision, lag response, and low energy efficiency in existing technologies, but also achieves a fundamental shift from "uniform cooling" to "precise cooling delivery." Its core value lies in transforming cooling from an energy burden on infrastructure into an intelligently schedulable IT support resource, achieving millisecond-level dynamic matching of cooling capacity and heat load through algorithm-driven mechanisms. The technical concept of this invention can be extended to high-heat-density scenarios such as edge computer rooms and 5G base stations, and has significant practical implications and broad commercial value for promoting energy conservation and emission reduction in the entire information and communication industry and achieving "dual-carbon" goals, demonstrating promising application prospects.

[0072] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for on-demand cooling of an energy-saving inter-row and inter-column high-precision cabinet air conditioner, characterized in that, Includes the following steps: S1: Real-time acquisition of dynamic heat load parameters of the cooling area, operating status parameters of key equipment in the cabinet, and operating parameters of the air conditioning system itself; S2: Based on the dynamic heat load parameters, operating status parameters, and its own operating parameters, the real-time overall cooling demand is calculated using a preset cooling demand model; S3: Based on a preset allocation strategy, the total cooling demand is decomposed into inter-row air conditioning cooling demand and inter-column air conditioning cooling demand; S4: Based on the inter-row air conditioning cooling demand and the inter-row air conditioning cooling demand, generate corresponding inter-row air conditioning control commands and inter-row air conditioning control commands respectively, so as to dynamically adjust the operating status of the inter-row air conditioning and inter-row air conditioning and realize on-demand cooling; S5: Real-time monitoring of energy efficiency indicators after on-demand cooling, including the ratio of the total power consumption of the cooling system to the total power consumption of the IT equipment; S6: Based on the energy efficiency index, optimize the parameters and allocation strategy in the cooling demand model.

2. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S1, the dynamic heat load parameters include at least the rack inlet temperature, rack outlet temperature, rack hot spot temperature, and return air temperature; the operating status parameters include at least the server fan speed and CPU load rate; and the self-operating parameters include at least the air conditioner supply air temperature, compressor operating frequency, and fan speed.

3. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S2, the method for constructing the cooling demand model is implemented in one of the following two ways: Method 1: Based on historical and experimental data, establish a machine learning regression model with dynamic heat load parameters and operating status parameters as inputs and cooling capacity as output; Method 2: Based on the principle of thermal balance, a physical model is established that includes the heat output of the equipment, air flow rate and temperature difference, and the heat output of the equipment is dynamically corrected through the operating state parameters.

4. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S3, the preset allocation strategy is as follows: the in-row air conditioner is given priority to handle the precise cooling demand for individual and local cabinets based on the temperature of the cabinet inlet and outlet. The portion of the total cooling demand that exceeds the cooling capacity of the in-row air conditioner, as well as the background cooling demand required to eliminate the heat island effect in the entire cooling area, is handled by the row air conditioner.

5. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 4, characterized in that, In step S3, the allocation strategy further includes: when the temperature of any of the cabinet hotspots exceeds the first safety threshold, increasing the cooling output of the inter-row air conditioner in the corresponding area; when the return air temperature exceeds the second safety threshold, increasing the cooling output of the inter-row air conditioner.

6. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S4, the dynamic adjustment of the operating status of the inter-row air conditioner and the inter-row air conditioner specifically includes: for the inter-row air conditioner, controlling its air supply temperature and air volume by adjusting the compressor frequency, the opening degree of the electronic expansion valve and the fan speed; for the inter-row air conditioner, controlling its air outlet speed and air outlet temperature by adjusting the fan speed and refrigerant flow rate.

7. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S5, the energy efficiency index also includes the coefficient of performance of the refrigeration system, and the overall energy efficiency level is evaluated by calculating the weighted comprehensive energy efficiency index of the power utilization efficiency and the coefficient of performance.

8. The on-demand cooling method for an energy-saving inter-row and inter-column high-precision cabinet air conditioner according to claim 1, characterized in that, In step S6, the feedback optimization uses a recursive least squares algorithm to update the parameters in the cooling demand model online, and uses fuzzy logic control to adjust the allocation strategy.

Images