Fan control methods, devices, electronic equipment, and storage media for heat dissipation systems

Abstract

This application discloses a fan control method, device, electronic device, and storage medium for a heat dissipation system. The method includes: a baseboard management controller acquiring a first total power consumption of the current cycle board and a first temperature of a target component; determining the power consumption change rate of the board and the temperature change rate of the target component based on the first total power consumption and the first temperature, as well as the second total power consumption of the board in the previous cycle and the second temperature of the target component; weighting the first temperature based on the power consumption change rate and the temperature change rate to determine a first predicted temperature of the target component in the next cycle; determining a first rotational speed of the fan based on the first temperature of the target component, the first predicted temperature, and the target temperature; generating a control signal based on the first rotational speed information and sending the control signal to the fan to adjust the fan speed based on the control signal.

Description

Technical Field

[0001] This application relates to the field of heat dissipation control technology for electronic devices, and in particular to a fan control method, device, electronic device, and storage medium for a heat dissipation system. Background Technology

[0002] During the operation of data centers and high-performance computing equipment, critical components such as processors and memory chips on server motherboards continuously generate a large amount of heat under high load conditions. Existing thermal management mechanisms generally adopt temperature threshold triggering or proportional-integral-derivative feedback adjustment strategies. The core logic is to monitor real-time temperature data from sensors and only initiate the fan speed increase program when the detected value exceeds a preset safety threshold. This passive response mode has inherent technical defects: heat is generated from critical components, conducted to temperature sensors through a medium, processed by the control system, and finally drives the fan to complete the speed adjustment. The entire process requires a time delay of multiple sampling cycles. When the workload increases dramatically, the rate of heat accumulation far exceeds the thermal response speed, causing a significant overshoot phenomenon in the temperature of critical components before the fan can fully accelerate. In severe cases, this will trigger hardware frequency reduction protection mechanisms or forced system shutdown, resulting in interruption of computing tasks and reduced service availability. On the other hand, although some research has attempted to introduce predictive control theory to intervene in temperature changes in advance, it relies on accurate thermodynamic modeling or highly complex numerical calculations, requiring a large amount of processor resources and memory space, and cannot be adapted to the lightweight operating environment of embedded monitoring units such as baseboard management controllers. Furthermore, prediction models with fixed parameters struggle to adapt to dynamic changes such as hardware aging, airflow blockage, or ambient temperature fluctuations, leading to a continuous decline in prediction accuracy over time. These technical bottlenecks result in significant shortcomings in current thermal management systems regarding response timeliness, environmental adaptability, and resource utilization, making it difficult to meet the real-time and robust requirements of modern electronic equipment for thermal management. Summary of the Invention

[0003] In view of this, embodiments of this application provide a fan control method, apparatus, electronic device, and storage medium for a heat dissipation system.

[0004] According to a first aspect of this application, a fan control method for a heat dissipation system is provided. The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. The method includes: The baseboard management controller obtains the first total power consumption of the board in the current cycle and the first temperature of the target component; Based on the first total power consumption and the first temperature, as well as the second total power consumption of the board and the second temperature of the target component in the previous cycle, the power consumption change rate of the board and the temperature change rate of the target component are determined. The first temperature is weighted based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle. Based on the first temperature of the target component, the first predicted temperature, and the target temperature, determine the first rotational speed of the fan; A control signal is generated based on the first rotation speed information and sent to the fan so that the fan adjusts its rotation speed based on the control signal.

[0005] Optionally, the formula for weighting the first temperature based on the rate of change of power consumption and the rate of change of temperature includes: ; in, Predicted temperature for the target component in the next cycle; The first temperature; The rate of change of power consumption; The rate of temperature change; The time period is denoted by α, which is the temperature change rate weighting coefficient, representing the influence of temperature inertia on the predicted temperature, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, with units of ℃·s / W, used to convert the power consumption change rate into an equivalent temperature rise rate, with an initial value of 0.05℃·s / W.

[0006] Optionally, the fan control method for the cooling system also includes: Based on the first temperature and the second predicted temperature of the target component in the previous cycle, the prediction error of the previous cycle is determined. The weighting coefficients for the rate of temperature change and the rate of power consumption change are updated based on the prediction error.

[0007] Optionally, the formula for updating the weighting coefficient of the rate of temperature change based on the prediction error includes: ; in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

[0008] Optionally, the formula for updating the weighting coefficient of the power consumption change rate based on the prediction error includes: ; in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

[0009] Optionally, based on the first temperature of the target component, the first predicted temperature, and the target temperature, the first rotational speed information of the fan is determined, including: Based on the first temperature and the target temperature, the current deviation is determined, which represents the degree to which the first temperature deviates from the target temperature; Based on the first predicted temperature and the target temperature, the predicted overshoot is determined. The predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature. The overshoot temperature difference is the difference between the predicted temperature and the target temperature. Based on the current deviation and the predicted overshoot, determine the first rotational speed of the fan.

[0010] Optionally, the formula for determining the first rotational speed of the fan based on the current deviation and the predicted overshoot includes: ; in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

[0011] Optionally, a control signal is generated based on the first revolution information, including: Based on the first rotational speed information and the rotational speed smoothing strategy, the second rotational speed information of the fan is determined; The fan control signal is generated based on the second rotation speed information.

[0012] Optionally, the fan control method for the cooling system also includes: Obtain fan speed feedback information; If the feedback information determines that the fan speed is not within the fan speed threshold range, a fan malfunction is identified, and an alarm is triggered.

[0013] According to a second aspect of this application, a fan control device for a heat dissipation system is provided, applied to a baseboard management controller. The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. The device includes: The acquisition module is used to acquire the first total power consumption of the board in the current cycle and the first temperature of the target component; The first determining module is used to determine the power consumption change rate of the board and the temperature change rate of the target component based on the first total power consumption and the first temperature, as well as the second total power consumption of the board and the second temperature of the target component in the previous cycle. The processing module is used to perform weighted processing on the first temperature based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle. The second determining module is used to determine the first rotational speed information of the fan based on the first temperature of the target component, the first predicted temperature, and the target temperature. The transmitting module is used to generate a control signal based on the first rotational speed information and send the control signal to the fan so that the fan adjusts its rotational speed based on the control signal.

[0014] Optionally, the formula for weighting the first temperature based on the rate of change of power consumption and the rate of change of temperature includes: ; in, Predicted temperature for the target component in the next cycle; The first temperature; The rate of change of power consumption; The rate of temperature change; The time period is denoted by α, which is the temperature change rate weighting coefficient, representing the influence of temperature inertia on the predicted temperature, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, with units of ℃·s / W, used to convert the power consumption change rate into an equivalent temperature rise rate, with an initial value of 0.05℃·s / W.

[0015] Optionally, the fan control device for the cooling system also includes: The update module is used to determine the prediction error of the previous cycle based on the first temperature and the second predicted temperature of the target component in the previous cycle. The weighting coefficients for the rate of temperature change and the rate of power consumption change are updated based on the prediction error.

[0016] Optionally, the formula for updating the weighting coefficient of the rate of temperature change based on the prediction error includes: ; in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

[0017] Optionally, the formula for updating the weighting coefficient of the power consumption change rate based on the prediction error includes: ; in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

[0018] Optionally, the second determining module is used to determine the current deviation based on the first temperature and the target temperature, wherein the current deviation represents the degree to which the first temperature deviates from the target temperature; Based on the first predicted temperature and the target temperature, the predicted overshoot is determined. The predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature. The overshoot temperature difference is the difference between the predicted temperature and the target temperature. Based on the current deviation and the predicted overshoot, determine the first rotational speed of the fan.

[0019] Optionally, the formula for determining the first rotational speed of the fan based on the current deviation and the predicted overshoot includes: ; in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

[0020] Optionally, the sending module is used to determine the second rotational speed information of the fan based on the first rotational speed information and the rotational speed smoothing strategy; The fan control signal is generated based on the second rotation speed information.

[0021] Optionally, the fan control device for the cooling system also includes: The third determining module is used to obtain fan speed feedback information; If the feedback information determines that the fan speed is not within the fan speed threshold range, a fan malfunction is identified, and an alarm is triggered.

[0022] According to a third aspect of this application, this application provides an electronic device, comprising: Baseboard management controller, heat dissipation system; The cooling system includes the circuit board, the target components mounted on the circuit board, and the fan; The baseboard management controller is used to execute the fan control method of the above-described heat dissipation system.

[0023] According to a fourth aspect of this application, this application provides a computer-readable storage medium storing computer instructions for causing a computer to execute the fan control method of the above-described heat dissipation system.

[0024] According to a fifth aspect of this application, this application provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the fan control method of the above-described heat dissipation system.

[0025] This application provides a fan control method, device, electronic device, storage medium, and program product for a heat dissipation system. The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. A baseboard management controller obtains the first total power consumption of the circuit board and the first temperature of the target component in the current cycle. Based on the first total power consumption and the first temperature, and the second total power consumption of the circuit board and the second temperature of the target component in the previous cycle, the power consumption change rate of the circuit board and the temperature change rate of the target component are determined. The first temperature is weighted based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle. Based on the first temperature of the target component, the first predicted temperature, and the target temperature, the first rotational speed information of the fan is determined. A control signal is generated based on the first rotational speed information and sent to the fan to adjust the fan speed according to the control signal. Thus, by fusing power consumption and the dynamic temperature change rate, real-time prediction of temperature trends is achieved. This allows for anticipating temperature increases before a sudden increase in the load on the target component and intervening in the fan speed in advance, enabling proactive sensing of the thermal state of the target component. This allows for advance adjustment of fan speed, effectively suppressing temperature overshoot that may occur when electronic devices operate under high loads. It also prevents frequency throttling protection triggered by overheating, thereby improving system stability and energy efficiency. Furthermore, the algorithm's simple structure makes it suitable for resource-constrained baseboard management controllers, enhancing overall system reliability.

[0026] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a fan control method for a heat dissipation system according to an embodiment of this application. Figure 2 This is a schematic diagram of the structure of a fan control device for a heat dissipation system according to an embodiment of this application; Figure 3 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0029] This application proposes a fan control method for a heat dissipation system. The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. The fan control method for the heat dissipation system is as follows: Figure 1 As shown, it includes: S101, The baseboard management controller obtains the first total power consumption of the board in the current cycle and the first temperature of the target component; S102, based on the first total power consumption and the first temperature, and the second total power consumption of the board and the second temperature of the target component in the previous cycle, determine the power consumption change rate of the board and the temperature change rate of the target component. S103, the first temperature is weighted based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle; S104, determine the first rotational speed information of the fan based on the first temperature of the target component, the first predicted temperature and the target temperature; S105, a control signal is generated based on the first rotation speed information, and the control signal is sent to the fan so that the fan adjusts its rotation speed based on the control signal.

[0030] For ease of understanding, the following explains some key terms in this embodiment: A circuit board can refer to a server motherboard, graphics card, or other circuit board that carries electronic components. The target component refers to the critical electronic component on the circuit board that generates heat, such as a central processing unit (CPU), graphics processing unit (GPU), storage controller, or hard drive. A fan is used to remove the heat generated by the target component by forcing airflow, thus maintaining it within a safe operating temperature range.

[0031] A baseboard management controller is typically a dedicated microcontroller used in embedded systems to monitor and manage hardware status. This controller can operate independently of the main processor, is responsible for collecting various sensor data, such as temperature, power consumption, and voltage, and controlling the system according to preset strategies or algorithms, such as adjusting fan speed, power status, or triggering alarms.

[0032] The first total power consumption represents the overall power consumption rate of the board as obtained by the board management controller during the current sampling period. This power consumption data reflects the real-time load of all components on the board and is an important indicator for evaluating the system's thermal load.

[0033] The first temperature represents the real-time temperature of the target component acquired by the substrate management controller during the current sampling period. This temperature is typically measured using a temperature sensor integrated near or inside the target component and is a key parameter directly reflecting the thermal state of the target component.

[0034] The second total power consumption represents the overall power consumption rate of the board as obtained by the board management controller in the previous sampling period immediately preceding the current sampling period.

[0035] The second temperature represents the real-time temperature of the target component acquired by the substrate management controller in the previous sampling period immediately preceding the current sampling period.

[0036] The power consumption change rate represents the rate at which the total power consumption of the board changes over two consecutive sampling periods. This rate of change reflects the dynamic trend of system load changes, such as sudden increases or decreases in load, providing advanced information for predicting future temperatures.

[0037] The rate of temperature change represents the speed at which the temperature of a target component changes over two consecutive sampling periods. This rate of change reflects the current trend of heat accumulation or dissipation in the target component.

[0038] The first predicted temperature represents the temperature of the target component in the next sampling period, estimated using a specific prediction model based on the power consumption and temperature data of the current and previous cycles. This predicted temperature aims to provide a forward-looking temperature value so that the system can take heat dissipation measures in advance.

[0039] The target temperature represents the upper limit of the ideal operating temperature set to ensure the stable and reliable operation of the target component. The goal of the fan control strategy is to make the actual temperature of the target component as close as possible to or below this target temperature.

[0040] The first RPM information indicates the fan speed or speed-related control parameters that should be achieved, calculated based on the current temperature, predicted temperature, and target temperature. This information guides the actual operation of the fan.

[0041] The control signal is an electrical signal generated by the baseboard management controller based on the first rotation speed information and sent to the fan. After receiving the control signal, the fan will adjust its rotation speed according to the signal instruction, thereby changing its heat dissipation capacity.

[0042] During the execution of the method, the baseboard management controller can obtain the real-time total power consumption of the board by reading the output data of the power management chip, or calculate the total power consumption by measuring the current and voltage of the board's power supply circuit. Simultaneously, the initial temperature of the target component can be read in real time by a temperature sensor connected to the baseboard management controller; this sensor can be integrated inside the target component or adjacent to its surface.

[0043] Furthermore, the board management controller determines the power consumption rate of the board and the temperature rate of change of the target component based on the acquired first total power consumption, first temperature, second total power consumption of the board in the previous cycle, and second temperature of the target component. For example, the power consumption rate of change can be calculated by dividing the difference between the first total power consumption of the current cycle and the second total power consumption of the previous cycle by the sampling period. Similarly, the temperature rate of change can be calculated by dividing the difference between the first temperature of the current cycle and the second temperature of the previous cycle by the sampling period. These rates of change reflect the dynamic trends of system heat load and temperature.

[0044] Based on this, the substrate management controller weights the first temperature according to the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle. For example, a linear model can be used, taking the current temperature as a baseline, and then multiplying the power consumption change rate and the temperature change rate by their respective weighting coefficients, and then adding these weighting terms to the current temperature to obtain a forward-looking predicted temperature value. This predicted temperature aims to anticipate the future thermal state of the target component in advance.

[0045] Therefore, the board management controller determines the first fan speed information based on the first temperature of the target component, the first predicted temperature, and the target temperature. For example, the fan speed can be initially set based on the difference between the current temperature and the target temperature; when the current temperature is higher than the target temperature, the speed is increased; when the current temperature is lower than the target temperature, the speed is decreased. At the same time, the initially set speed can be corrected by combining the difference between the predicted temperature and the target temperature; for example, if the predicted temperature is much higher than the target temperature, the speed can be further increased to prepare in advance.

[0046] Finally, the board management controller generates a control signal based on the first RPM information and sends the control signal to the fan, causing the fan to adjust its RPM according to the control signal. For example, the first RPM information can be a percentage value, which the board management controller converts into the duty cycle of a pulse width modulation (PWM) signal and outputs to the fan through the PWM pin. After receiving the PWM signal, the fan's internal drive circuit adjusts the motor speed according to the duty cycle, thereby changing the cooling airflow.

[0047] The fan control method of the heat dissipation system in this embodiment acquires the power consumption of the circuit board and the temperature of the target component in real time, and predicts future temperatures by combining their rate of change. This allows for anticipating temperature increases before a sudden increase in the load on the target component, enabling proactive sensing of the target component's thermal state. Consequently, the fan speed can be adjusted in advance, effectively suppressing temperature overshoot that may occur when electronic devices operate under high loads, avoiding frequency throttling protection triggered by overheating, and thus improving the system's operational stability and energy efficiency. Furthermore, the algorithm has a simple structure, is suitable for resource-constrained circuit board management controllers, and enhances the overall reliability of the system.

[0048] In an optional embodiment, this application further proposes a formula for weighting the first temperature based on the rate of change of power consumption and the rate of change of temperature. ; in, Predicted temperature for the target component in the next cycle; The first temperature; The power consumption change rate is a parameter that directly reflects the dynamic change in the heat generated by the target component. For example, when the power consumption of the board increases rapidly, A positive and large value indicates that the target component will heat up rapidly; The temperature change rate is a parameter that characterizes the current heating or cooling trend of the target component, reflecting its inherent thermal inertia. For example, even if the power consumption is temporarily stable, if... If the value is still positive, it indicates that the target component is still heating up. The time period is denoted as α; α is the temperature change rate weighting coefficient, which characterizes the influence of temperature inertia on the predicted temperature. This coefficient is a dimensionless value used to adjust the weight of the current temperature change trend in the prediction, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, which quantifies the degree of influence of power consumption change on the temperature change of the target component. The unit is ℃·s / W, and it is used to convert the power consumption change rate into an equivalent temperature rise rate. The initial value is 0.05℃·s / W.

[0049] By employing the aforementioned technical solution and using a clear linear prediction formula to weight the first temperature, the current temperature, temperature change rate, and power consumption change rate can be accurately incorporated into the calculation of the first predicted temperature of the target component in the next cycle. This formula, by introducing a weighting coefficient α for the temperature change rate and a weighting coefficient β for the power consumption change rate, allows the prediction model to flexibly adapt to changes in the thermal characteristics and workload of different target components, thereby significantly improving the accuracy of the predicted temperature. Based on this accurate prediction, the cooling system can predict the temperature trend of the target component earlier and more accurately, achieving more proactive fan control, effectively avoiding temperature overshoot and hysteresis response. This ensures that the temperature of the target component remains stable within a safe range while optimizing the fan operation strategy, reducing unnecessary fan speed and noise, and improving overall heat dissipation efficiency and system reliability.

[0050] In an optional embodiment, this application further proposes that after determining the first predicted temperature of the target component in the next cycle, the method further includes determining the prediction error of the previous cycle based on the first temperature and the second predicted temperature of the target component in the previous cycle; and updating the weighting coefficients of the temperature change rate and the power consumption change rate based on the prediction error.

[0051] Specifically, in each control cycle, the substrate management controller acquires the actual first temperature of the target component for the current cycle. Simultaneously, the system will save the predicted temperature of the target component from the previous cycle, i.e., the second predicted temperature of the target component from the previous cycle. By comparing the actual first temperature of the current cycle with the second predicted temperature of the previous cycle, the prediction error of the previous cycle can be calculated. ,like The prediction error directly reflects the accuracy of the model's prediction in the previous cycle. Its magnitude and sign indicate the degree and direction of deviation between the predicted and actual values. For example, if the actual first temperature is higher than the second predicted temperature, the prediction error is positive, indicating that the model predicted too low; conversely, if the actual first temperature is lower than the second predicted temperature, the prediction error is negative, indicating that the model predicted too high.

[0052] After obtaining the prediction error, the system uses this error to adjust the weighting coefficients for the rate of temperature change and the rate of power consumption change. This adjustment mechanism allows the model to learn and optimize itself based on actual prediction performance. When the prediction error is large, the adjustment magnitude of the weighting coefficients will increase accordingly to correct model bias more quickly; when the prediction error is small, the adjustment magnitude will decrease to maintain model stability. In this way, the weighting coefficients are no longer fixed but can dynamically adapt to the thermal response characteristics of the cooling system under different operating conditions, thereby improving the accuracy and robustness of temperature prediction.

[0053] Through the above technical solution, this application enables adaptive adjustment of the temperature prediction model. When the prediction model deviates, the system can promptly detect and utilize the prediction error to correct the weight parameters in the model, allowing these weight coefficients to dynamically reflect the true thermal characteristics of the cooling system under different operating conditions. This significantly improves the accuracy of the first predicted temperature of the target component in the next cycle, thereby enabling the fan's first rotation speed information to more accurately respond to the cooling needs of the target component. Ultimately, fan control becomes more refined and efficient, avoiding energy waste caused by excessive heat dissipation due to inaccurate predictions, and preventing the risk of component overheating due to insufficient heat dissipation, ensuring that the cooling system maintains optimal heat dissipation performance under various operating conditions.

[0054] In an optional embodiment, this application further proposes a formula for updating the weighting coefficient of the temperature change rate based on the prediction error, the formula comprising: ;in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

[0055] Specifically, the weighting coefficient for the rate of temperature change This characterizes the influence of temperature inertia on the predicted temperature, and its initial value is typically set to 1.0. Updated temperature change rate weighting coefficient. This is a new value that has been learned and adjusted by the system, and is used in the subsequent temperature prediction process. Learning rate It is a preset fixed decimal, such as 0.01, which is used to adjust the magnitude of the adjustment of the weight coefficient by the prediction error e during each update, so as to balance the stability of the update and the convergence speed.

[0056] Through the above technical solution, the system can dynamically adjust the weighting coefficient of the temperature change rate using the prediction error and the previous cycle's temperature change rate, with the learning rate as the step size. This adaptive update mechanism allows the temperature prediction model to continuously optimize its assessment of temperature change inertia based on actual operating conditions. For example, when the prediction model continuously underestimates the rate of temperature rise (i.e., the prediction error is positive and the previous cycle's temperature change rate is positive), the system can improve its performance. The value will increase accordingly, causing the model to place greater emphasis on the upward trend in temperature in subsequent predictions, thereby improving the accuracy of the predictions. Conversely, if the model continues to overestimate temperature changes, The value will decrease. This adjustment of weighting coefficients based on actual feedback ensures that the prediction model can better adapt to changes in the thermal characteristics of the board under complex operating conditions such as different loads, ambient temperature fluctuations, or component aging, significantly improving the accuracy of target component temperature prediction. More accurate temperature prediction is the foundation for achieving refined fan control. It can avoid overheating or underheating caused by inaccurate prediction, thereby optimizing the overall efficiency of the cooling system, reducing unnecessary energy consumption, and helping to extend the service life of the target components.

[0057] In an optional embodiment, this application further proposes a formula for updating the power consumption change rate weighting coefficient based on the prediction error, including: ;in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

[0058] Through the above technical solution, this application, based on predicting temperature and updating the weighting coefficients according to the prediction error, further introduces a mechanism for adaptively updating the power consumption change rate weighting coefficient β based on the prediction error. This update method allows the power consumption change rate weighting coefficient to be dynamically adjusted according to the actual prediction error and power consumption changes. When the prediction error is large, the system will increase or decrease β accordingly based on the trend of power consumption changes to enhance or weaken the impact of power consumption changes on the predicted temperature. For example, if the predicted temperature is consistently lower than the actual temperature, and the power consumption in the previous cycle shows an upward trend, the system may appropriately increase β so that the model can more fully consider the temperature rise effect caused by the increase in power consumption in the next prediction. Conversely, the same applies. This adaptive adjustment capability significantly improves the accuracy and robustness of the temperature prediction model, especially in scenarios where board power consumption fluctuates frequently or the workload changes dynamically. By more accurately predicting the first predicted temperature of the target component in the next cycle, the fan control system can adjust the first rotation speed information of the fan more timely and accurately, thereby achieving more refined and efficient heat dissipation control, effectively avoiding excessively high or low temperatures of the target component, extending equipment life, and optimizing system energy consumption.

[0059] In an optional embodiment, this application further proposes determining the first rotational speed information of the fan based on the first temperature of the target component, the first predicted temperature, and the target temperature, including: determining a current deviation based on the first temperature and the target temperature, wherein the current deviation represents the degree to which the first temperature deviates from the target temperature; determining a predicted overshoot based on the first predicted temperature and the target temperature, wherein the predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature; the overshoot temperature difference is the difference between the predicted temperature and the target temperature; and determining the first rotational speed information of the fan based on the current deviation and the predicted overshoot.

[0060] Specifically, the current deviation The first temperature used to quantify the target component Deviation from target temperature The degree of [temperature / temperature] can be obtained by calculating the difference between the target temperature and the first temperature, for example... This deviation directly reflects the current heat dissipation requirements of the system, that is, the difference between the actual temperature of the target component and the desired temperature. For example, when the first temperature is higher than the target temperature, the current deviation is negative, indicating that heat dissipation needs to be increased; when the first temperature is lower than the target temperature, the current deviation is positive, indicating that heat dissipation can be appropriately reduced.

[0061] Predicted overshoot The first predicted temperature used to characterize the target component in the next cycle. Exceeding the target temperature The overshoot temperature difference introduced at that time, for example Specifically, when the first predicted temperature is higher than the target temperature, the predicted overshoot is determined to be the difference between the first predicted temperature and the target temperature; if the first predicted temperature is not higher than the target temperature, the predicted overshoot can be set to zero. This predicted overshoot allows the system to anticipate potential overheating of the target component in the next cycle, thereby taking preventative heat dissipation measures before the actual temperature rises.

[0062] After obtaining the current deviation and predicted overshoot, the initial fan speed can be determined based on these two values. This method comprehensively considers both current cooling requirements and potential future overheating risks. For example, the current deviation and predicted overshoot can be weighted and summed, or these two values ​​can be mapped to specific fan speed information through methods such as lookup tables or fuzzy control. In this way, the fan speed not only responds to the current temperature state but also makes proactive adjustments for upcoming temperature changes.

[0063] Through the above technical solution, when determining the first fan speed, the system no longer relies solely on a simple combination of temperature values. Instead, it first explicitly quantifies the degree to which the current temperature deviates from the target temperature (i.e., the current deviation) and the predicted risk of the temperature exceeding the target temperature (i.e., the predicted overshoot). This decomposition and quantification allows the fan control strategy to respond more precisely to current cooling needs while also being forward-looking, effectively avoiding temperature overshoot. Specifically, the current deviation ensures a rapid response to immediate cooling demands, while the predicted overshoot allows the system to intervene by increasing the fan speed before the temperature actually reaches an excessively high level. This control method, which combines the current state with future trends, can significantly improve the response speed and control accuracy of the cooling system, reduce temperature fluctuations, and maintain the temperature of the target component more stably near the target value, thereby improving the overall cooling performance and reliability of the system.

[0064] In an optional embodiment, this application further proposes a formula for determining the first rotational speed of the fan based on the current deviation and the predicted overshoot, including: ; in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

[0065] Specifically, a PWM signal is a commonly used digital signal that allows for precise control of fan motor speed by adjusting the ratio of its high-level duration to its period (i.e., duty cycle). A higher duty cycle typically results in a faster fan speed, thus providing stronger heat dissipation. The board management controller can output this PWM signal to the fan drive circuitry to adjust the fan speed.

[0066] The above technical solution quantifies the current deviation and predicted overshoot using a clear linear combination formula, thereby accurately determining the fan's first rotational speed. This formula will reflect the immediate response to the current temperature deviation (from...). and (Decision) and prediction of future temperature exceeding the standard (by and This approach combines proportional control and predictive control. This allows the fan control system to not only passively adjust based on the current temperature but also actively intervene based on predictions of future temperature trends. For example, if the system predicts that the temperature of the target component will exceed the target temperature in the next cycle, even if the current temperature has not yet reached the target temperature, the system can increase the fan speed in advance by predicting the overshoot (T_over), thereby effectively suppressing a rapid temperature rise and avoiding or significantly reducing temperature overshoot. This strategy, combining proportional control and predictive control, significantly improves the response speed and control accuracy of the cooling system, enabling the temperature of the target component to be maintained more stably near the target temperature. This reduces performance degradation or potential failure risks caused by excessive temperature fluctuations, thereby optimizing overall heat dissipation performance.

[0067] In an optional embodiment, this application further proposes a step of generating a control signal based on the first rotational speed information, including: determining the second rotational speed information of the fan based on the first rotational speed information and a rotational speed smoothing strategy; and generating a control signal for the fan based on the second rotational speed information.

[0068] Specifically, the first RPM information is a theoretical RPM calculated based on the current temperature, predicted temperature, and target temperature, which may vary significantly due to instantaneous data fluctuations. To avoid such instantaneous fluctuations directly affecting the fan, this application introduces a RPM smoothing strategy. This RPM smoothing strategy aims to process the first RPM information to eliminate or mitigate its volatility, thereby making the fan RPM change more smooth and gradual. For example, the RPM smoothing strategy may include a rate-of-change limit (smooth start / stop, limiting the speed per second to prevent noise or inrush current caused by sudden changes in fan speed). The range of change does not exceed ) and dead zone limitation (settings) Threshold, Threshold, highest Threshold. If the first revolution information... If the threshold is reached, the output will be 20%. A threshold is set to ensure the fan can reliably rotate despite static friction; if the first rotation information... If the threshold is reached, the output will be 100%. Threshold).

[0069] Specific implementation method: Let the actual output revolution count information of the previous cycle be... The first revolution information is The rate of change limiting strategy outputs the revolutions information. for: .

[0070] The revolution count information output by the rate of change limiting strategy is then passed through the dead zone limiting strategy to output a second revolution count information.

[0071] After determining the fan's second rotational speed (RPS) information, the board management controller (BDC) generates a control signal for the fan based on this information. The second RPS information is a smoothed version of the desired fan RPS; this step aims to convert this desired RPS into a physical signal that the fan controller can recognize and execute. Typically, the control signal can be in the form of a pulse-width modulation (PWM) signal. In this case, the second RPS information is mapped to the duty cycle of the PWM signal; for example, 0% duty cycle corresponds to the fan stopping, 100% duty cycle corresponds to the fan running at full speed, and intermediate RPS correspond to the corresponding duty cycle. The BDC typically includes a PWM generator module that calculates the required duty cycle based on the second RPS information and outputs the corresponding PWM waveform. For fans that support analog control, the second RPS information can also be converted into an analog voltage signal (such as 0-5V or 0-10V) via a digital-to-analog converter (DAC). For more intelligent fans, the control signal may be sent via digital communication protocols such as I2C, SPI, or UART, containing the desired RPS or speed level.

[0072] By introducing a speed smoothing strategy through the above technical solution, drastic fluctuations in fan speed caused by instantaneous calculation results are avoided, resulting in smoother fan operation. This significantly reduces noise generated by the fan during frequent speed changes, improving the user experience. Simultaneously, smooth speed changes reduce mechanical shock and wear on the fan motor and bearings, thereby extending fan lifespan and reducing maintenance costs. Furthermore, stable fan speed helps maintain the overall stability of the cooling system, preventing localized overheating or overcooling caused by sudden speed changes, ensuring more precise and reliable temperature control of target components.

[0073] In an optional embodiment, this application further proposes that the fan control method of the above-mentioned heat dissipation system further includes obtaining fan speed feedback information; if it is determined based on the feedback information that the fan speed is not within the fan speed threshold range, a fan fault is determined and an alarm is triggered.

[0074] Specifically, during fan control, the board management controller acquires fan speed feedback information in real time. This feedback information is typically generated by an integrated speed sensor (such as a Hall sensor) inside the fan, which detects the actual rotational frequency of the fan blades and converts it into an electrical signal. The board management controller periodically reads these electrical signals through a specific communication interface (such as a PWM feedback line, I2C bus, or SPI bus) to obtain the actual fan speed. This is designed to monitor the fan's operating status in real time, providing a basis for subsequent fault diagnosis.

[0075] After receiving fan speed feedback, the board management controller determines whether the actual fan speed is within a preset threshold range. This threshold range is typically set based on fan specifications, system cooling requirements, and control strategies, and includes a minimum and a maximum speed threshold. For example, when the fan receives a control signal, its expected speed should fall within an allowable deviation range. If the actual feedback speed is lower than the minimum threshold (e.g., the fan is not running or its speed is too low), or if the actual speed deviates significantly from the expected speed corresponding to the control signal, the fan speed can be determined to be outside the normal operating range.

[0076] When the baseboard management controller determines that the fan speed is outside the fan speed threshold range, the system further determines that the fan has malfunctioned. To avoid misjudgments caused by transient interference or measurement errors, the system can employ a certain fault-tolerance mechanism. For example, a fan malfunction is only confirmed when abnormal fan speeds are detected in multiple consecutive sampling periods, or when the fan speed fails to reach the speed required by the control signal for an extended period. This confirmation mechanism improves the accuracy of fault diagnosis.

[0077] Once a fan failure is confirmed, the board management controller will immediately trigger an alarm. The alarm can take various forms, such as illuminating or flashing an indicator light, emitting a buzzer sound, displaying a warning message on the system management interface, logging the failure event to the system log, or sending a notification to a remote monitoring system via a network protocol (such as SNMP). The purpose of triggering the alarm is to promptly notify the user or system administrator so that they can take swift action, such as replacing the faulty fan or adjusting the system load, thereby preventing damage to target components on the board due to insufficient cooling.

[0078] Through the above technical solution, this application can monitor the actual operating status of the fan in real time, avoiding the blind spot of relying solely on control signal output without being able to perceive the actual working status of the fan. When the actual fan speed deviates from the preset normal operating threshold range, the system can promptly identify potential fan malfunctions. Once a fan malfunction is detected, the system immediately triggers an alarm, promptly reminding maintenance personnel to intervene and prevent overheating of target components due to fan failure. This effectively protects the stable operation of target components on the board and extends the service life of the equipment, significantly improving the reliability and safety of the heat dissipation system.

[0079] This application also provides a fan control device for a heat dissipation system, applied to a baseboard management controller. The heat dissipation system includes a circuit board, target components mounted on the circuit board, and a fan, such as... Figure 2 As shown, the device includes: The acquisition module 21 is used to acquire the first total power consumption of the board in the current cycle and the first temperature of the target component; The first determining module 22 is used to determine the power consumption change rate of the board and the temperature change rate of the target component based on the first total power consumption and the first temperature, as well as the second total power consumption of the board and the second temperature of the target component in the previous cycle. Processing module 23 is used to perform weighted processing on the first temperature based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle; The second determining module 24 is used to determine the first rotational speed information of the fan based on the first temperature of the target component, the first predicted temperature and the target temperature; The sending module 25 is used to generate a control signal based on the first rotational speed information and send the control signal to the fan so that the fan adjusts its rotational speed based on the control signal.

[0080] In some embodiments, the formula for weighting the first temperature based on the rate of change of power consumption and the rate of change of temperature includes: ; in, Predicted temperature for the target component in the next cycle; The first temperature; The rate of change of power consumption; The rate of temperature change; The time period is denoted by α, which is the temperature change rate weighting coefficient, representing the influence of temperature inertia on the predicted temperature, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, with units of ℃·s / W, used to convert the power consumption change rate into an equivalent temperature rise rate, with an initial value of 0.05℃·s / W.

[0081] In some embodiments, the fan control device of the cooling system further includes: The update module is used to determine the prediction error of the previous cycle based on the first temperature and the second predicted temperature of the target component in the previous cycle; and to update the weighting coefficients of the temperature change rate and the power consumption change rate based on the prediction error.

[0082] In some embodiments, the formula for updating the weighting coefficient of the rate of temperature change based on the prediction error includes: ; in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

[0083] In some embodiments, the formula for updating the power consumption change rate weighting coefficient based on the prediction error includes: ; in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

[0084] In some embodiments, the second determining module is used to determine a current deviation based on a first temperature and a target temperature, wherein the current deviation represents the degree to which the first temperature deviates from the target temperature; determine a predicted overshoot based on a first predicted temperature and a target temperature, wherein the predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature; the overshoot temperature difference is the difference between the predicted temperature and the target temperature; and determine the first rotational speed information of the fan based on the current deviation and the predicted overshoot.

[0085] In some embodiments, the formula for determining the first rotational speed of the fan based on the current deviation and the predicted overshoot includes: ; in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

[0086] In some embodiments, the sending module is used to determine the second rotational speed information of the fan based on the first rotational speed information and the rotational speed smoothing strategy; and to generate a control signal for the fan based on the second rotational speed information.

[0087] In some embodiments, the fan control device of the cooling system further includes: The third determination module is used to obtain fan speed feedback information; if it is determined based on the feedback information that the fan speed is not within the fan speed threshold range, a fan fault is determined and an alarm is triggered.

[0088] This embodiment provides an electronic device. For example... Figure 3 As shown, the electronic device includes a baseboard management controller 31 and a heat dissipation system; the heat dissipation system includes a board 321, a target component 322 disposed on the board, and a fan 323; the baseboard management controller 31 is used to execute the fan control method of the above-mentioned heat dissipation system.

[0089] In this electronic device, the board management controller, as the core control unit, is configured to operate independently of the main processor, continuously collecting data on the total power consumption of the board and the temperature of the target components. The boards in the cooling system serve as the physical carriers of electronic components, including key heat-generating components such as the central processing unit (CPU) and graphics processing unit (GPU). Fans achieve heat exchange through forced airflow. By executing the aforementioned fan control method, the board management controller can calculate the power consumption change rate and temperature change rate in real time based on current and historical power consumption and temperature data, and generate a forward-looking estimate of the future temperature using a lightweight linear prediction model. This prediction model employs a dual-input mechanism, using the power consumption change rate as a leading signal, effectively overcoming the inherent lag problem of traditional temperature threshold triggering methods. Based on this, the board management controller integrates the current temperature deviation and the predicted overshoot to generate fan speed commands, and ensures smooth fan speed adjustment through rate-of-change limits and dead-zone limits.

[0090] In this embodiment, by combining the baseboard management controller with a heat dissipation system including the board, target component, and fan, and having the baseboard management controller execute the aforementioned dual-input lightweight predictive control method based on the rate of change of power consumption and the rate of change of temperature, active suppression of temperature peaks is achieved without increasing hardware costs, thereby improving system stability and energy efficiency. Specifically, this electronic device can run adaptive predictive algorithms in real time in an embedded environment without relying on complex thermodynamic modeling or additional computing resources, significantly reducing temperature overshoot and avoiding performance jitter caused by temperature fluctuations. Because the baseboard management controller continuously corrects the predictive model coefficients online, this electronic device can automatically adapt to different hardware configurations, fan aging, and environmental changes, ensuring that the heat dissipation strategy remains optimal during long-term operation. Through the above technical solution, this electronic device effectively solves the temperature overshoot problem caused by control lag when servers and storage systems operate under high load, maintaining the safe operating temperature of the target component while extending fan life and reducing system noise.

[0091] This application provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the fan control method for the cooling system described above in this application.

[0092] This application provides a computer-readable storage medium storing executable instructions, wherein the executable instructions are stored and when executed by a processor, they will cause the processor to execute the fan control method of the heat dissipation system provided in this application.

[0093] In some embodiments, a computer-readable storage medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may be a machine-readable signal medium or a machine-readable storage medium. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0094] In some embodiments, executable instructions may take the form of a program, software, software module, script, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

[0095] As an example, executable instructions may, but do not necessarily, correspond to files in a file system. They may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a Hyper Text Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple collaborating files (e.g., a file that stores one or more modules, subroutines, or code sections).

[0096] As an example, executable instructions can be deployed to execute on a single computing device, or on multiple computing devices located in one location, or on multiple computing devices distributed across multiple locations and interconnected via a communication network.

[0097] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0098] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0099] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0100] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this application can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this application can be achieved, and this is not limited herein.

[0101] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0102] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A fan control method for a heat dissipation system, characterized in that, The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan; the method includes: The baseboard management controller obtains the first total power consumption of the board in the current cycle and the first temperature of the target component; Based on the first total power consumption and the first temperature, as well as the second total power consumption of the board and the second temperature of the target component in the previous cycle, the power consumption change rate of the board and the temperature change rate of the target component are determined. The first temperature is weighted based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle. Based on the first temperature of the target component, the first predicted temperature, and the target temperature, the first rotational speed information of the fan is determined; A control signal is generated based on the first rotation speed information, and the control signal is sent to the fan so that the fan adjusts its rotation speed based on the control signal.

2. The fan control method for the heat dissipation system according to claim 1, characterized in that, The formula for weighting the first temperature based on the power consumption change rate and the temperature change rate includes: ； in, Predicted temperature for the target component in the next cycle; The first temperature; The rate of change of power consumption; The rate of temperature change; The time period is denoted by α, which is the temperature change rate weighting coefficient, representing the influence of temperature inertia on the predicted temperature, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, with units of ℃·s / W, used to convert the power consumption change rate into an equivalent temperature rise rate, with an initial value of 0.05℃·s / W.

3. The fan control method for the heat dissipation system according to claim 1 or 2, characterized in that, Also includes: Based on the first temperature and the second predicted temperature of the target component in the previous cycle, the prediction error of the previous cycle is determined. The weighting coefficients for the rate of temperature change and the rate of power consumption change are updated based on the prediction error.

4. The fan control method for the heat dissipation system according to claim 3, characterized in that, The formula for updating the weighting coefficient of the rate of temperature change based on the prediction error includes: ； in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.

01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

5. The fan control method for the heat dissipation system according to claim 3, characterized in that, The formula for updating the power consumption change rate weighting coefficient based on the prediction error includes: ； in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.

01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

6. The fan control method for the heat dissipation system according to claim 1, characterized in that, Based on the first temperature of the target component, the first predicted temperature, and the target temperature, the first rotational speed information of the fan is determined, including: Based on the first temperature and the target temperature, a current deviation is determined, wherein the current deviation represents the degree to which the first temperature deviates from the target temperature; Based on the first predicted temperature and the target temperature, a predicted overshoot is determined, wherein the predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature; the overshoot temperature difference is the difference between the predicted temperature and the target temperature. Based on the current deviation and the predicted overshoot, the first rotational speed of the fan is determined.

7. The fan control method for the heat dissipation system according to claim 6, characterized in that, Based on the current deviation and the predicted overshoot, the formula for determining the first rotational speed of the fan includes: ； in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

8. The fan control method for the heat dissipation system according to claim 1, characterized in that, Based on the first revolution count information, a control signal is generated, including: Based on the first rotational speed information and the rotational speed smoothing strategy, the second rotational speed information of the fan is determined; The fan control signal is generated based on the second rotation speed information.

9. The fan control method for the heat dissipation system according to claim 1, characterized in that, Also includes: Obtain fan speed feedback information; If the feedback information determines that the fan speed is not within the fan speed threshold range, a fan malfunction is identified, and an alarm is triggered.

10. A fan control device for a heat dissipation system, applied to a baseboard management controller, characterized in that, The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. The device includes: The acquisition module is used to acquire the first total power consumption of the board in the current cycle and the first temperature of the target component; The first determining module is used to determine the power consumption change rate of the board and the temperature change rate of the target component based on the first total power consumption and the first temperature, as well as the second total power consumption of the board in the previous cycle and the second temperature of the target component. The processing module is used to perform weighted processing on the first temperature based on the power consumption change rate and the temperature change rate to determine the first predicted temperature of the target component in the next cycle; The second determining module is used to determine the first rotational speed information of the fan based on the first temperature of the target component, the first predicted temperature, and the target temperature. The transmitting module is used to generate a control signal based on the first rotational speed information and send the control signal to the fan so that the fan adjusts its rotational speed based on the control signal.

11. The fan control device for the heat dissipation system according to claim 10, characterized in that, The formula for weighting the first temperature based on the power consumption change rate and the temperature change rate includes: ； in, Predicted temperature for the target component in the next cycle; The first temperature; The rate of change of power consumption; The rate of temperature change; The time period is denoted by α, which is the temperature change rate weighting coefficient, representing the influence of temperature inertia on the predicted temperature, with an initial value of 1.0; β is the power consumption change rate weighting coefficient, with units of ℃·s / W, used to convert the power consumption change rate into an equivalent temperature rise rate, with an initial value of 0.05℃·s / W.

12. The fan control device for the heat dissipation system according to claim 10 or 11, characterized in that, Also includes: The update module is used to determine the prediction error of the previous cycle based on the first temperature and the second predicted temperature of the target component in the previous cycle. The weighting coefficients for the rate of temperature change and the rate of power consumption change are updated based on the prediction error.

13. The fan control device for the heat dissipation system according to claim 12, characterized in that, The formula for updating the weighting coefficient of the rate of temperature change based on the prediction error includes: ； in, The updated weighting coefficients for the rate of temperature change. This is the weighting coefficient for the current rate of temperature change; The learning rate is set to a fixed decimal of 0.

01. This represents the prediction error; This represents the rate of temperature change in the previous cycle.

14. The fan control device for the heat dissipation system according to claim 12, characterized in that, The formula for updating the power consumption change rate weighting coefficient based on the prediction error includes: ； in, The updated power consumption change rate weighting coefficient. This is the weighting coefficient for the current rate of change in power consumption; The learning rate is set to a fixed decimal of 0.

01. This represents the prediction error; This represents the rate of change in power consumption in the previous cycle.

15. The fan control device for the heat dissipation system according to claim 10, characterized in that, The second determining module is used to determine the current deviation based on the first temperature of the target component and the target temperature, wherein the current deviation represents the degree to which the first temperature deviates from the target temperature; Based on the first predicted temperature and the target temperature, a predicted overshoot is determined, wherein the predicted overshoot represents the overshoot temperature difference introduced when the predicted temperature exceeds the target temperature. The overshoot temperature difference is the difference between the predicted temperature and the target temperature; Based on the current deviation and the predicted overshoot, the first rotational speed of the fan is determined.

16. The fan control device for the heat dissipation system according to claim 15, characterized in that, Based on the current deviation and the predicted overshoot, the formula for determining the first rotational speed of the fan includes: ； in, The duty cycle of the modulation pulse width signal corresponding to the fan speed represents the fan speed. This is a proportionality coefficient with a value of 5% / ℃, representing the intensity of the response to the current deviation. This is the current deviation. To predict overshoot; The prediction coefficient, with a value of 3% / ℃, represents the degree of importance attached to the overshoot temperature difference.

17. The fan control device for the heat dissipation system according to claim 10, characterized in that, The sending module is used to determine the second rotational speed information of the fan based on the first rotational speed information and the rotational speed smoothing strategy; The fan control signal is generated based on the second rotation speed information.

18. The fan control device for the heat dissipation system according to claim 10, characterized in that, Also includes: The third determining module is used to obtain fan speed feedback information; If the feedback information determines that the fan speed is not within the fan speed threshold range, a fan malfunction is identified, and an alarm is triggered.

19. An electronic device, characterized in that, include: Baseboard management controller, heat dissipation system; The heat dissipation system includes a circuit board, a target component mounted on the circuit board, and a fan. The baseboard management controller is used to execute the fan control method of the heat dissipation system as described in any one of claims 1-9.

20. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the fan control method of the cooling system as described in any one of claims 1-9.

21. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the processor, they implement the fan control method of the heat dissipation system according to any one of claims 1 to 9.

Images