Heat dissipation optimization method for server based on dynamic fan control technology of partition management

Abstract

The invention discloses a heat dissipation optimization method for a server by a dynamic fan control technology based on partition management, which comprises the following steps of: acquiring temperature data of each partition in a system through a BMC (Baseboard Management Controller), inputting the data into an algorithm module, and comparing the deviation between the real-time temperature and the set target temperature by the algorithm module to obtain the heat dissipation of the server. The appropriate values for adjusting the fan speed and the air inlet amount are calculated, and when the temperature of the core component exceeds the set threshold value, the algorithm indicates the BMC to increase the air inlet amount of the fan in front of the core component area so as to improve air flow, and therefore the cooling effect is enhanced. Through the dynamic adjustment, the system can automatically optimize the working state of the fan according to the temperature change of the core component, and ensures that the heat dissipation capability of each partition is effectively improved, so that the overall temperature of the server is kept within a constant range.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of server heat dissipation technology, and specifically to a heat dissipation optimization method for servers using dynamic fan control technology based on partition management. BACKGROUND

[0002] In modern server design, heat dissipation is a significant technical challenge. As computing power increases, server power consumption also increases, resulting in a significant increase in heat generation. Traditional heat dissipation systems usually rely on global fan control and static heat dissipation solutions, which often face efficiency problems when dealing with high-density, high-power components. Especially in the case of different heat dissipation needs within the server, traditional methods often cannot be flexibly adjusted, thereby affecting the overall heat dissipation effect and system stability.

[0003] Current heat dissipation techniques mainly include the following: Global fan control: All fans run at the same speed, unable to adjust to the specific heat generation of each component. This method is simple and easy to use, but it is not effective for uneven heat generation.

[0004] Temperature sensor monitoring: Using temperature sensors to monitor overall temperature and adjusting fan speed based on these data. Although this method is an improvement over global fan control, it still struggles to accurately manage the heat dissipation needs of different areas.

[0005] Static heat dissipation design: Based on server design specifications, such as heat pipe heat dissipation, heat sinks, etc. Although these designs can improve heat dissipation efficiency to some extent, they still struggle to adapt to changing heat loads in real-world environments.

[0006] The disadvantages of the global server heat dissipation system include: Uneven heat dissipation: The global system may not effectively address the different heat dissipation needs of individual components, leading to overheating in some areas.

[0007] Low energy efficiency: Global heat dissipation systems often require high-volume fans to run continuously, which can lead to energy waste and increased noise.

[0008] Maintenance difficulties: The maintenance and troubleshooting of global systems can be complex, as the heat dissipation of all components depends on a unified system.

[0009] Due to the limitations of these existing technologies, heat dissipation management remains a pressing issue in high-power servers and data centers. How to achieve more accurate and dynamic fan control, optimize heat dissipation, and reduce energy consumption and noise has become an important direction for current research.

[0010] Therefore, the application provides a dynamic fan control technology based on partition management for server heat dissipation optimization method. SUMMARY

[0011] The application aims to provide a dynamic fan control technology based on partition management for server heat dissipation optimization method, which specifically includes the following steps: S1, divide the heat dissipation area inside the server, and the division is based on the heat dissipation of components, function type and position S2, configure a corresponding fan system in each heat dissipation area; S3, the server is designed with an air flow channel to suck in cold air from the front, pass through the internal components, and discharge hot air from the back. The internal part adopts a partitioned air duct, and each partition corresponds to a fan system; S4, equip each heat dissipation area with a high-precision temperature sensor to monitor the temperature of the components, and manage them uniformly through BMC; S5, according to the temperature data of each area, BMC adjusts the speed of the fan through a specific heat dissipation control algorithm, and performs partition PID adjustment and synchronous control instruction.

[0012] Further, the division of the heat dissipation area in S1 includes the following: S11, core component area: including but not limited to high heat components such as central processing unit and graphics processing unit; S12, auxiliary component area: including but not limited to low heat components such as memory bar and hard disk drive; S13, power supply area: the power supply module also generates heat and needs a special heat dissipation area.

[0013] Further, the fan system configuration in S2 includes the following: S21, inlet fan: install a high-flow low-noise fan at the inlet of the core component area to introduce cold air and reduce the temperature of the core components while maintaining a low noise level; S22, outlet fan: install a high-pressure fan at the outlet of the core component area to discharge hot air, improve exhaust efficiency and prevent heat accumulation in the core component area.

[0014] Further, the temperature sensor configuration in S4 is based on the following heat dissipation requirements: S41, core component area: arrange temperature sensors near the main heat generating components to monitor temperature changes and timely discover and handle temperature abnormal conditions to prevent overheating from damaging the core components; S42, auxiliary component area: temperature sensors are arranged near the key auxiliary components of the memory and hard disk to ensure comprehensive monitoring of temperature changes in the area, maintain the auxiliary components within a safe temperature range, and optimize the overall performance and stability of the system; S43, power supply area: temperature sensors are arranged near the power supply module to monitor the temperature dynamics of the power supply area in real time, so that the power supply module operates within a normal temperature range, thereby improving the reliability and lifespan of the power supply.

[0015] Further, the control algorithm in S5 adopts PID control calculation, and the specific steps are as follows: S51, temperature data acquisition: real-time acquisition of temperature data from each area's temperature sensor; S52, the PID control calculation formula is: Output = P_out + I_out + D_out Where, Output is the control signal applied to the fan, Pout is the proportional response of the current error, Iout is the integral response of the historical error accumulation, and Dout is the differential response of the error change rate; S521, the calculation of P_out is: P_out = Kp Error Where, Kp is the proportional coefficient, which is a constant that needs to be debugged, and the larger Kp is, the faster the system responds; S522, the calculation of I_out is: I_out = Ki × ErrorΔt Where, ErrorΔt is the integral of the error change over time, and Ki is the integral coefficient, which is a constant that needs to be debugged. The larger Ki is, the stronger the ability to eliminate static error; S53, the calculation of D_out is: D_out = Kd × d(Error) / Δt Where, Kd is the differential coefficient, which is a constant that needs to be debugged. The larger Kd is, the stronger the inhibition of the change trend, making the system more stable. d(Error) / Δt is the rate of change of the error over time.

[0016] Further, the control algorithm in S5 adopts a fuzzy logic algorithm, and the specific steps are as follows: S51, temperature data acquisition: real-time acquisition of temperature data from each area's temperature sensor; S52, convert the input precision value to the membership degree of the fuzzy set, temperature error and error change rate; S521, define linguistic variables, i.e. descriptive language values, which can be defined as: negative big (NB), negative small (NS), zero (ZO), positive small (PS), and positive big (PB).

[0017] Further, the partition PID adjustment in S5 includes applying independent PID controllers to each region to meet the heat dissipation needs of different regions, and the algorithm will calculate the speed adjustment amount of each partition fan respectively to control the temperature of all regions within a predetermined range.

[0018] Further, the synchronization control instruction in S5 includes real-time transmission of the calculation results to the heat dissipation module, so that the fan can quickly respond to the control signal and adjust the speed to match the required heat dissipation effect.

[0019] Further, in S4, the BMC obtains temperature data of each partition in the system through a communication interface and inputs these data into the control algorithm.

[0020] The present application has the following advantages: through such dynamic adjustment, the system can automatically optimize the working state of the fan according to the temperature change of the core components, ensuring that the heat dissipation capacity of each partition is effectively improved, so that the overall temperature of the server is maintained within a constant range. This precise fan control and temperature management mechanism can greatly improve the stability and heat dissipation efficiency of the server, reduce the risk of overheating, and ensure efficient operation of the system. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a system block diagram of the present application; Figure 2 is a flowchart of embodiment 1 of the present application; Figure 3 is a flowchart of embodiment 2 of the present application. DETAILED DESCRIPTION

[0022] The present application provides a dynamic fan control technology based on partition management for server heat dissipation optimization method.

[0023] Embodiment 1

[0024] S1, divide the server into multiple server heat dissipation regions, and the division of each region is based on the heat generation, function type and position of the components, S11, core component area: including but not limited to high heat generating components such as central processing unit and graphics processing unit; S12, auxiliary component area: including but not limited to low heat generating components such as memory bar and hard disk drive; S13, power supply area: the power supply module also generates heat and needs a special heat dissipation area; S2, configure corresponding fan systems in each heat dissipation area to adapt to the temperature requirements of different areas, especially in the core component area, use two groups of fan configurations, which can optimize airflow distribution, improve heat dissipation effect, and reduce airflow resistance, ensuring that the temperature of the core component area is maintained within a reasonable range; S21, inlet fan: install high-volume low-noise fans at the inlet of the core component area to introduce cold air and reduce the temperature of the core components while maintaining a low noise level; S22, outlet fan: install high-pressure fans at the outlet of the core component area to exhaust hot air, improve exhaust efficiency, and prevent heat accumulation in the core component area; S3, the server internal air duct design adopts a multi-air duct layout; S31, the server is designed with multiple airflow channels to draw in cold air from the front, pass through internal components, and exhaust hot air from the back. Multiple partition air ducts are used inside, with each partition corresponding to a fan system to ensure that cold air effectively flows through each critical area; S4, equip each heat dissipation area with multiple high-precision temperature sensors to monitor the temperature of the components in real time. The BMC can collect and monitor temperature information in each area in real time and provide a centralized management platform to respond and handle any temperature anomalies in a timely manner, ensuring stable operation and efficient heat dissipation of the system; S41, core component area: place temperature sensors near the main heat-generating components to monitor temperature changes and promptly detect and handle temperature anomalies to prevent overheating from damaging core components; S42, auxiliary component area: place temperature sensors near key auxiliary components such as memory and hard drives to ensure comprehensive monitoring of temperature changes in the area and maintain auxiliary components within a safe temperature range, optimizing overall system performance and stability; S43, power area: place temperature sensors near power modules to monitor the temperature dynamics of the power area in real time, ensuring that power modules operate within a normal temperature range, thereby improving the reliability and lifespan of the power supply; S5, based on temperature data from each area, the BMC adjusts fan speed through specific control algorithms and performs partition PID adjustment and synchronous control instructions using a PID control algorithm for calculation; S51, temperature data acquisition: obtain temperature data from each area's temperature sensors in real time; S52, the PID control calculation formula is: Output = P_out + I_out + D_out Wherein, Output is the control signal applied to the fan, Pout is the proportional response of the current error, Iout is the integral response of the historical error accumulation, and Dout is the differential response of the error change rate; S521, the calculation of P_out is: P_out = Kp Error Wherein, Kp is a proportional coefficient, which is a constant that needs to be debugged, and the larger Kp is, the faster the system response is; S522, the calculation of I_out is: I_out = Ki × ErrorΔt Wherein, ErrorΔt is the integral of the error change over time, and Ki is an integral coefficient, which is a constant that needs to be debugged, and the larger Ki is, the stronger the ability to eliminate static error is; S523, the calculation of D_out is: D_out = Kd × d(Error) / Δt Wherein, Kd is a differential coefficient, which is a constant that needs to be debugged, and the larger Kd is, the stronger the inhibition of the change trend is, making the system more stable, and d(Error) / Δt is the error change rate over time.

[0025] Embodiment 2

[0026] S1, the server is divided into multiple server heat dissipation areas, and the division of each area mainly includes the heat generation of components, the type and position of functions, S11, core component area: including but not limited to high heat generating components such as central processing unit and graphics processing unit; S12, auxiliary component area: including but not limited to low heat generating components such as memory bar and hard disk drive; S13, power supply area: the power supply module also generates heat and needs a special heat dissipation area; S2, a corresponding fan system is configured in each heat dissipation area to adapt to the temperature demand of different areas, especially in the core component area, two groups of fan configurations are adopted, which can optimize air distribution, improve heat dissipation effect, reduce air resistance, and ensure that the temperature of the core component area is maintained within a reasonable range; S21, inlet fan: high air volume and low noise fan is installed at the inlet of the core component area to introduce cold air and reduce the temperature of the core component, while maintaining a low noise level; S22, outlet fan: high air pressure fan is installed at the outlet of the core component area to exhaust hot air, improve exhaust efficiency, and prevent heat accumulation in the core component area; S3, the server internal air duct design adopts a multi-air duct layout; S31, the server is designed with multiple air flow channels, inhaling cold air from the front, passing through internal components, and exhausting hot air from the back, with multiple partitioned air ducts inside, each corresponding to a fan system, ensuring that cold air effectively flows through each critical area; S4, multiple high-precision temperature sensors are equipped in each heat dissipation area to monitor the temperature of the components in real time, and the BMC is used for unified management. The BMC can collect and monitor the temperature information of each area in real time, and provide a centralized management platform to respond and handle any temperature abnormality in a timely manner, ensuring stable operation and efficient heat dissipation of the system; S41, core component area: temperature sensors are arranged near the main heat-generating components to monitor temperature changes and timely detect and handle temperature abnormalities to prevent overheating from damaging core components; S42, auxiliary component area: temperature sensors are arranged near the key auxiliary components such as memory and hard drive to ensure comprehensive monitoring of temperature changes in the area and maintain auxiliary components within a safe temperature range, optimizing overall system performance and stability; S43, power supply area: temperature sensors are placed near the power supply module to monitor the temperature dynamics of the power supply area in real time, ensuring that the power supply module operates within a normal temperature range, thereby improving the reliability and lifespan of the power supply; S5, based on the temperature data of each area, the BMC adjusts the fan speed through a specific control algorithm, using a fuzzy logic algorithm for calculation; S51, temperature data collection: real-time acquisition of temperature data from each area's temperature sensor; S52, convert precise input values (e.g. error +5°C) to fuzzy language values (e.g. "positive medium" membership degree 0.7, "positive very large" membership degree 0.3), input includes "temperature error" and "error change rate"; S53, rule inference: apply a pre-defined fuzzy rule base, for example: IF error is positive medium AND error is still rising rapidly THEN let the speed increase moderately; S54, defuzzification: combine the outputs of all activated rules and convert them back to a precise speed adjustment value, output includes "fan speed change amount".

[0027] Although the specific embodiments of the present application are described in detail with reference to the accompanying drawings, it should not be understood as limiting the scope of protection of the present application. Various modifications and variations made by those skilled in the art within the scope described in the claims are still within the scope of protection of the present application.

Claims

1. A method for optimizing heat dissipation of a server based on dynamic fan control technology of partition management, characterized in that: The method specifically comprises the following steps: S1, dividing the heat dissipation area inside the server, the division basis including the heat generation of components, function type and position S2, configuring a corresponding fan system in each heat dissipation area; S3, the server is designed with an air flow channel, cold air is sucked from the front, hot air is discharged from the rear through internal components, and a partitioned air duct is adopted inside, each partition corresponds to a fan system; S4, a high-precision temperature sensor is equipped in each heat dissipation area to monitor the temperature of components, and unified management is conducted through BMC; S5, according to the temperature data of each area, the BMC adjusts the rotating speed of the fan through a specific heat dissipation control algorithm, and conducts partitioned PID adjustment and synchronous control instruction.

2. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The division mode of the heat dissipation area in S1 comprises the following: S11, core component area: including but not limited to high heat generation components such as central processing unit and graphic processing unit; S12, auxiliary component area: including but not limited to low heat generation components such as memory bar and hard disk drive; S13, power supply area: the power supply module also generates heat and needs a special heat dissipation area.

3. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The fan system configuration in S2 comprises the following: S21, air inlet fan: a high air volume and low noise fan is installed at the air inlet of the core component area to introduce cold air and reduce the temperature of the core component while maintaining a low noise level; S22, air outlet fan: a high air pressure fan is installed at the air outlet of the core component area to discharge hot air, improve exhaust efficiency and prevent heat accumulation in the core component area.

4. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The temperature sensor configuration in S4 is based on the following heat dissipation requirements: S41, core component area: temperature sensors are arranged near the main heat generation components to monitor temperature changes, discover and handle temperature abnormal conditions in time, and prevent overheating from causing damage to the core components; S42, auxiliary component area: temperature sensors are arranged near the key auxiliary components of the memory and hard disk to ensure comprehensive monitoring of temperature changes in the area, maintain the auxiliary components within a safe temperature range, and optimize the overall performance and stability of the system; S43, power supply area: temperature sensors are arranged near the power supply module to real-time master the temperature dynamics of the power supply area, so that the power supply module operates within a normal temperature range, thereby improving the reliability and life of the power supply.

5. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The control algorithm in S5 adopts PID control calculation, and the specific steps are as follows: S51, temperature data acquisition: real-time acquisition of temperature data from each area temperature sensor; S52, the PID control calculation formula is: Output = P_out + I_out + D_out Wherein, Output is the control signal applied to the fan, Pout is the proportional response of the current error, Iout is the integral response of the historical error accumulation, and Dout is the differential response of the error change rate; S521, the calculation of P_out is: P_out = Kp * Error Wherein, Kp is the proportional coefficient, which is a constant that needs to be debugged, and the larger Kp is, the faster the system responds; S522, the calculation of I_out is: I_out = Ki × ∫ErrorΔt Wherein, ∫ErrorΔt is the integral of error change over time, Ki is the integral coefficient, which is a constant that needs to be debugged. The larger Ki is, the stronger the ability to eliminate static error is; S53, the calculation of D_out is: D_out = Kd × d(Error) / Δt Wherein, Kd is the differential coefficient, which is a constant that needs to be debugged. The larger Kd is, the stronger the inhibition of the trend is, and the more stable the system is. d(Error) / Δt is the rate of error change over time.

6. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The control algorithm in S5 adopts a fuzzy logic algorithm, and the specific steps are as follows: S51, temperature data acquisition: real-time acquisition of temperature data from the temperature sensor of each region; S52, convert the input accurate value into the membership degree of the fuzzy set, temperature error and error change rate S521, define linguistic variables, i.e. descriptive language values, which can be defined as: negative big (NB), negative small (NS), zero (ZO), positive small (PS), and positive big (PB); S522, define a membership function for each language value, triangular and trapezoidal functions, which define how much the fuzzy set is for a certain accurate input value.

7. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The partition PID regulation in S5 includes applying independent PID controllers to each region to meet the heat dissipation needs of different regions. The algorithm will calculate the speed adjustment amount of each partition fan respectively, so that the temperature of all regions can be controlled within the predetermined range.

8. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: The synchronization control instruction in S5 includes real-time transmission of the calculation results to the heat dissipation module, so that the fan can quickly respond to the control signal and adjust the speed to match the required heat dissipation effect.

9. The method for optimizing heat dissipation of a server using the dynamic fan control technique based on zone management according to claim 1, wherein: In S4, the BMC obtains the temperature data of each partition in the system through the communication interface and inputs these data into the control algorithm.

Images