A thermal management system for an all-in-one computer

Abstract

The present invention belongs to the field of thermal management of all-in-one computers, and discloses a thermal management system for an integrated all-in-one computer, comprising: a relationship modeling module for modeling the power consumption and heat of the integrated all-in-one computer, generating an associated expression of power consumption and heat, and obtaining a power consumption-heat model; a data collection and implementation module for importing a real-time data set into the power consumption-heat model to obtain a converted data set; a trend prediction and mode selection module for judging the state of the all-in-one computer according to a curve and selecting an operating mode according to the state; a thermal management module for designing the operating mode under a thermal management program, and the design content includes a dynamic frequency and voltage regulation unit and a power consumption and heat dissipation collaborative optimization unit; a regulation management module regulates the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit; and realizes the thermal management of the integrated all-in-one computer.

Description

Technical Field

[0001] The present invention relates to the field of thermal management of all-in-one computers, and more specifically, to a thermal management system for an integrated all-in-one computer. Background Art

[0002] Due to the compact structure and high integration of integrated all-in-one computers, thermal management is particularly important. The internal hardware of all-in-one computers is highly integrated, but the body design is usually thin and the internal space is limited. Due to the narrow space, the size and efficiency of the heat dissipation components are restricted, and heat is not easily discharged effectively. Without effective thermal management, heat will accumulate inside, leading to an increase in the temperature of the hardware.

[0003] All-in-one computers usually carry high-performance processors (CPUs) and graphics cards (GPUs), and these components generate a large amount of heat when running high-load tasks. Thermal management can prevent the hardware from experiencing performance degradation or damage due to high temperature, especially when running tasks that require high computing power. Thermal management can avoid performance fluctuations caused by excessive temperature and ensure the stable operation of the system.

[0004] Excessive temperature may cause the surface of the body to heat up, affecting the user experience. High temperature may also cause the fan speed to increase, generating greater noise and further affecting the user's perception. Through thermal management optimization, the temperature of the body shell and fan noise can be reduced, improving the user experience.

[0005] In view of this, the present invention proposes a thermal management system for an integrated all-in-one computer to solve the above problems. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art and to achieve the above object, the present invention provides the following technical solution: A thermal management system for an integrated all-in-one computer, comprising: a relationship modeling module: for modeling the power consumption and heat of the integrated all-in-one computer, generating an associated expression of power consumption and heat, and obtaining a power consumption-heat model;

[0007] A data collection and implementation module: for collecting a real-time data set, importing the real-time data set into the power consumption-heat model, and obtaining a converted data set;

[0008] A trend prediction and mode selection module: for drawing a trend prediction curve based on the converted data set, judging the state of the all-in-one computer according to the curve, and selecting an operating mode according to the state;

[0009] A thermal management module: for designing the operating mode under the thermal management program, and the design content includes a dynamic frequency and voltage regulation unit and a power consumption and heat dissipation collaborative optimization unit;

[0010] Regulation and management module: Regulate the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit;

[0011] Visualization module: Display the analysis report on the computer terminal page and interact with the user through the interface.

[0012] Preferably, the method for modeling the power consumption and heat of the integrated all-in-one computer, generating the correlation expression of power consumption and heat, and obtaining the power consumption-heat model includes:

[0013] The power consumption-heat model is jointly constructed based on the power consumption expression, the heat correlation expression, and the temperature difference and heat dissipation efficiency influence formula;

[0014] The power consumption expression is:

[0015] + ;

[0016] The heat correlation expression is ;

[0017] The temperature difference and heat dissipation efficiency influence formula , ;

[0018] Among them, represents the total power consumption of the all-in-one computer, represents the total power consumption of the CPU and GPU, represents the total power consumption of other components, represents the capacitance of the th component, represents the voltage of the th component, represents the frequency of the th component, represents the leakage current related to the temperature on the th component, , represents the power consumption coefficient of the memory, represents the load of the memory, represents the power consumption coefficient of the storage device, represents the load of the storage device, represents the baseline power consumption of the motherboard and other fixed components, represents the comprehensive temperature of each component in the all-in-one computer, represents the ambient temperature, represents the th component's thermal resistance, represents the th component's power consumption, represents the heat dissipation efficiency, represents the comprehensive thermal resistance of each component inside the all-in-one computer, represents the heat dissipation area, represents the convective heat transfer coefficient, represents the difference between the comprehensive temperature of the components inside the all-in-one computer and the ambient temperature;

[0019] and the leakage current , is the leakage current at the selected reference temperature and supply voltage, is the voltage when the all-in-one computer is working, , , , , and are constants.

[0020] Preferably, the real-time data set includes the capacitance, frequency, voltage, temperature and thermal resistance of each component of the CPU and GPU, as well as the ambient temperature;

[0021] According to the real-time data set, use the power consumption-heat model to calculate the temperature of each component, the total power consumption and the heat dissipation efficiency of the all-in-one computer;

[0022] Form a conversion data set with the temperature of each component, the total power consumption and the heat dissipation efficiency of the all-in-one computer.

[0023] Preferably, the method of drawing a trend prediction curve according to the conversion data set, judging the state of the all-in-one computer according to the curve, and selecting an operation mode according to the state includes:

[0024] Preset a cycle time, discretize the current cycle time into time points, and obtain the conversion data set at each time point with a fixed sampling frequency;

[0025] Taking the chronological time points as the abscissa, and the temperature of each component and the total power consumption of the all-in-one computer in the conversion data set as the abscissa and ordinate respectively, draw a two-dimensional coordinate system, and perform curve fitting on each two-dimensional coordinate system to obtain the curve expressions of the temperature of each component and the total power consumption of the all-in-one computer over the cycle time;

[0026] Calculate the state performance coefficient according to the curve expression , if is greater than the boundary threshold, it is judged that the all-in-one computer has an overheating phenomenon, and the thermal management program is started. If is less than or equal to the boundary threshold, it is judged that the all-in-one computer does not have an overheating phenomenon, and it continues to work normally;

[0027] Among them, and are weight coefficients, and the value range is within 0 to 1, is the number of discrete time points within the cycle time, is the index of the time point, represents the value at the time point and represents the value at the time point and represents the integral of the curve expression within the cycle time.

[0028] Preferably, the operating modes under the thermal management program include a performance - priority mode, a quiet - priority mode, and a balanced mode;

[0029] The performance - priority mode means setting high - threshold power consumption and temperature under high load, the quiet - priority mode means restricting the fan speed and power consumption, and the balanced mode means balancing between performance and noise;

[0030] An adaptive mode - switching mechanism is introduced, specifically , where is the safety threshold of the ambient temperature, is the average value of the current component load, is the preset load threshold, is the low - load threshold.

[0031] Preferably, the design method of the dynamic frequency and voltage regulation unit includes:

[0032] Defining a frequency dynamic regulation mechanism and a voltage dynamic regulation mechanism for each component of the all - in - one computer;

[0033] Frequency dynamic regulation mechanism , where represents the frequency after the component is dynamically adjusted, , , and are the preset frequencies under various temperature and load states, and the values are arranged from high to low, , and are the preset multi - level temperature thresholds, and the values are arranged from low to high, represents the real - time temperature of the component, represents the real - time load of the component, and are the preset load thresholds;

[0034] Voltage dynamic regulation mechanism , where represents the voltage after the CPU or GPU is dynamically adjusted, is the reference voltage of the CPU or GPU, is the reference temperature of the CPU or GPU, is the sensitivity coefficient of temperature to voltage.

[0035] Preferably, the design method of the power consumption and heat dissipation co-optimization unit includes:

[0036] The power consumption and heat dissipation co-optimization unit includes a power consumption-heat dissipation adjustment mechanism and a co-optimization mechanism;

[0037] The power consumption-heat dissipation adjustment mechanism includes power consumption distribution and an optimization target. The power consumption distribution is and the optimization target is ;

[0038] Among them, is the dynamically allocated power consumption of the CPU or GPU, is the maximum power consumption of the CPU or GPU, is the dynamic adjustment coefficient, is the fan speed, is the total heat generated by the all-in-one computer, is the fan noise, is the temperature fluctuation penalty coefficient, refers to the change rate of the temperature of the component in the all-in-one computer;

[0039] The co-optimization mechanism is , is the base PWM frequency, is the frequency adjustment amplitude, is the phase;

[0040] The power consumption and heat dissipation co-optimization unit introduces a dynamic weight adjustment mechanism. The dynamic weight adjustment mechanism is Among them, represents the dynamic user experience weight, is the base weight;

[0041] is the environmental factor weight, and ; is the load factor weight, and ;

[0042] Among them, and are the base weights of the environmental factor and the load factor respectively, and are the sensitivity coefficients of the environment and the load respectively, is the reference temperature, is the number of loads on all current components, is the maximum load on all components, Indicates the average temperature change of each load, which is the time interval over the cycle time.

[0043] Preferably, the method for regulating the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit includes:

[0044] Obtain the dynamically adjusted frequency and voltage according to the dynamic frequency and voltage regulation unit, obtain the dynamically allocated power consumption and fan speed of the CPU or GPU according to the power consumption and heat dissipation collaborative optimization unit, and use the obtained frequency and voltage, power consumption and fan speed to regulate the all-in-one computer.

[0045] Preferably, the method for displaying the analysis report on the computer terminal page and performing interface interaction with the user includes:

[0046] The user clicks, queries, or downloads the analysis report through the computer terminal page.

[0047] Preferably, the analysis report includes the status of the all-in-one computer, the selected operating mode, the obtained frequency and voltage, and the power consumption and fan speed.

[0048] The technical effects and advantages of a thermal management system for an integrated all-in-one computer according to the present invention:

[0049] 1. Precise power consumption and heat modeling

[0050] By constructing a power consumption-heat model, simultaneous modeling of the power consumption expression, heat correlation expression, and temperature difference and heat dissipation efficiency formula is achieved, and the dynamic relationship between power consumption and heat in the all-in-one computer can be accurately described.

[0051] The model considers the power consumption characteristics of various hardware components such as the CPU, GPU, memory, storage device, and motherboard, covering various physical parameters such as capacitance, voltage, frequency, temperature, and thermal resistance. Key influencing factors such as leakage current, ambient temperature, and heat dissipation area are quantified to ensure the accuracy of the model. The model is driven by real-time data, enabling it to reflect the operating state of the all-in-one computer in real time and adapt to the dynamically changing power consumption and heat characteristics.

[0052] 2. Efficient real-time data acquisition and conversion

[0053] Data acquisition covers a variety of key hardware parameters, ensuring a comprehensive perception of the system state. The real-time data acquisition and conversion ability provides a basis for rapid response to subsequent prediction and regulation. Through model calculation, data conversion and analysis are automatically completed, reducing human intervention.

[0054] 3. Trend prediction and operating mode selection

[0055] Based on the conversion dataset, a trend prediction curve is plotted, and through curve fitting and the calculation of the state performance coefficient, the operating state of the all-in-one computer (such as whether it overheats) is judged, so as to select the performance priority, quiet priority or balanced mode. Through the trend prediction curve, the operating state of the all-in-one computer is judged in advance to avoid overheating or performance fluctuations.

[0056] It supports performance priority, quiet priority and balanced modes, and can dynamically adjust the operating mode according to the application scenario and user needs. By introducing parameters such as environmental temperature and load conditions, the mode switching mechanism is dynamically adjusted, improving the intelligence and adaptability of mode selection.

[0057] 4. Optimized Design of Thermal Management Program

[0058] Through the design of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit in the thermal management module, the collaborative optimization of power consumption and heat dissipation is realized, improving the heat dissipation efficiency and operating stability of the all-in-one computer.

[0059] For different temperature and load conditions, dynamically adjust the frequency and voltage of each component to reduce energy consumption and heat. The combination of the power consumption distribution mechanism and the fan noise suppression mechanism realizes the global optimization of power consumption and heat dissipation, taking into account performance and user experience. Achieve a balance among performance, noise and temperature fluctuations to meet the needs of different users.

[0060] 5. Dynamic Frequency and Voltage Regulation

[0061] The dynamic frequency and voltage regulation unit dynamically adjusts the frequency and voltage of each component according to the real-time temperature and load status, optimizing the relationship between power consumption and heat.

[0062] By dynamically adjusting the frequency and voltage, power consumption and heat generation are reduced, and the energy efficiency ratio is improved. Avoid component overheating or damage caused by too high frequency or voltage. Support the regulation strategy of multi-level temperature and load thresholds to adapt to different operating scenarios.

[0063] 6. Collaborative Optimization of Power Consumption and Heat Dissipation

[0064] The power consumption and heat dissipation collaborative optimization unit dynamically distributes power consumption and adjusts the fan speed through the power consumption distribution mechanism and the collaborative optimization mechanism, optimizing the relationship between power consumption and heat dissipation. By dynamically adjusting the power consumption distribution coefficient, the power consumption of the CPU and GPU is reasonably distributed to avoid overheating of a single component. The fan speed optimization mechanism effectively reduces noise and improves the user experience. Dynamically adjust the weight according to the user experience, environmental factors and load factors to optimize the collaborative effect of power consumption and heat dissipation.

[0065] Generally speaking, through the collaborative work of multiple modules, the global optimization of power consumption, heat and performance is realized, improving the overall performance and user experience of the all-in-one computer.

[0066] By establishing an accurate power consumption - heat model and integrating modules such as real - time data acquisition, trend prediction, dynamic regulation, and collaborative optimization, the power consumption management and heat control capabilities of all - in - one computers have been comprehensively improved. Its main advantages include high accuracy, high real - time performance, flexibility, and intelligence, which can significantly improve the operation stability, heat dissipation efficiency, and user experience of all - in - one computers, and have significant engineering application value. Brief Description of the Drawings

[0067] Figure 1 It is a schematic structural diagram of a thermal management system for an integrated all - in - one computer of the present invention;

[0068] Figure 2 It is a schematic step diagram of a thermal management method for an integrated all - in - one computer of the present invention. Detailed Embodiments

[0069] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0070] Embodiment 1

[0071] Please refer to Figure 1 and Figure 2 As shown, a thermal management system for an integrated all - in - one computer in this embodiment includes:

[0072] For integrated all - in - one computers (such as iMac or other All - in - One computers), due to their compact structure and high integration, thermal management is particularly important. The following are several key reasons why all - in - one computers need thermal management:

[0073] 1. Difficult heat dissipation due to limited space

[0074] The internal hardware of all - in - one computers (such as CPUs, GPUs, storage devices, power supplies, etc.) is highly integrated, but the body design is usually thin and the internal space is limited. Due to the narrow space, the size and efficiency of heat dissipation components (such as fans, heat sinks) are restricted, and heat is not easily discharged effectively. If effective thermal management is not carried out, heat will accumulate inside, resulting in an increase in the temperature of the hardware.

[0075] 2. High - performance hardware generates a large amount of heat

[0076] All-in-one computers usually come with high-performance processors (CPUs) and graphics cards (GPUs). These components generate a large amount of heat when running high-load tasks such as video editing, graphics rendering, gaming, etc. Thermal management can prevent the hardware from experiencing performance degradation (such as throttling) or damage due to high temperatures.

[0077] 3. Protect the hardware and extend the device lifespan

[0078] Running for a long time in a high-temperature environment will accelerate the aging of the hardware and shorten the device lifespan. For example, high temperatures may damage electronic components such as capacitors and transistors, and even cause solder joints to come off (such as the solder joint cracking problem of BGA chips). Good thermal management can effectively reduce the hardware temperature and extend the service life of the device.

[0079] 4. Ensure stable performance

[0080] When the hardware temperature is too high, modern computers will reduce the operating frequency of the CPU or GPU through "thermal throttling" to reduce heat generation. This throttling mechanism will directly affect the performance of the computer, especially when running tasks that require high computing power. Thermal management can avoid performance fluctuations caused by excessive temperatures and ensure stable system operation.

[0081] 5. Prevent the user experience from being affected

[0082] Excessive temperature may cause the surface of the fuselage to heat up, affecting the user experience. High temperatures may also cause the fan speed to increase, generating greater noise and further affecting the user's perception. Through thermal management optimization, the temperature of the fuselage shell and fan noise can be reduced, enhancing the user experience.

[0083] 6. Avoid safety hazards

[0084] In extreme cases, overheating may cause hardware failures and even pose a safety hazard of causing a fire. Thermal management can prevent the device from overheating and reduce potential safety risks.

[0085] Therefore, thermal management is crucial for all-in-one computers. It can not only ensure stable performance, extend the hardware lifespan, but also enhance the user experience and avoid safety hazards. Due to its design limitations, all-in-one computers require more efficient and innovative thermal management solutions to achieve efficient heat dissipation within a limited space, thus meeting the dual requirements of modern users for performance and aesthetics.

[0086] In the process of thermal management of an all-in-one computer, due to its compact structure, complex design, and high performance requirements, active cooling (such as a fan) is currently the most common cooling method, enabling heat to be quickly transferred to the outside of the body through heat pipes, heat sinks, or air ducts, and then dissipating heat through the outlet air. However, when operating at high loads, the increase in fan speed will generate obvious noise, affecting the user experience. In addition, the all-in-one computer will generate a large amount of heat under high loads, while increasing power consumption, and the power consumption itself will further exacerbate the heating, forming a vicious cycle. How to find a balance between high performance and high power consumption, ensure performance output while controlling heat generation, and enable the power management and cooling system to work together to avoid thermal management failure caused by uncontrolled power consumption is an important issue faced by thermal management.

[0087] In order to design a thermal management solution for an all-in-one computer that can effectively control heat and power consumption while ensuring high performance, starting from the perspectives of hardware, software, and system coordination, combining thermodynamics, electronics, and intelligent control technologies, and comprehensively considering factors such as power consumption distribution, heat dissipation path optimization, fan noise control, and user experience. The following is a detailed implementation plan:

[0088] Relationship modeling module: used to model the power consumption and heat of an all-in-one computer, generate an associated expression of power consumption and heat, and obtain a power consumption-heat model;

[0089] Step S1: Model the power consumption and heat of an all-in-one computer;

[0090] The methods for modeling the power consumption and heat of an all-in-one computer, generating an associated expression of power consumption and heat, and obtaining a power consumption-heat model include:

[0091] The power consumption-heat model is jointly constructed based on the power consumption expression, heat correlation expression, and temperature difference and heat dissipation efficiency influence formula;

[0092] The power consumption expression is + ;

[0093] The heat correlation expression is ;

[0094] The temperature difference and heat dissipation efficiency influence formula , ;

[0095] Among them, represents the total power consumption of the all-in-one computer, represents the total power consumption of the CPU and GPU, represents the total power consumption of other components, represents the capacitance of the th component, represents the The voltage of a component represents the frequency of the th component, and represents the leakage current related to the temperature on the th component, represents the power consumption coefficient of the memory, represents the load of the memory, represents the power consumption coefficient of the storage device, represents the load of the storage device, represents the baseline power consumption of the motherboard and other fixed components, represents the combined temperature of the components within the all-in-one computer, represents the ambient temperature, represents the thermal resistance of the th component, represents the power consumption of the th component, represents the heat dissipation efficiency, represents the combined thermal resistance of the components within the all-in-one computer, represents the heat dissipation area, represents the convective heat transfer coefficient, represents the difference between the combined temperature of the components within the all-in-one computer and the ambient temperature;

[0096] And the leakage current is the leakage current at the selected reference temperature and supply voltage, is the voltage when the all-in-one computer is working, , , , , and are constants.

[0097] The power consumption of the CPU and GPU has a non-linear relationship at different frequencies and voltages and will dynamically adjust the frequency and voltage according to the load, having dynamic power consumption characteristics. Therefore, we need to collect frequency and voltage data in real time to calculate the power consumption more accurately. According to the power consumption formula of the CMOS circuit, the dynamic power consumption can be expressed as: , where is the capacitive load (related to the design of the chip), is the supply voltage (Volt), is the operating frequency (Hz).

[0098] In addition, the static power consumption also needs to be considered, which is mainly caused by the leakage current Generated, static power consumption Related to voltage and temperature Related , Is the leakage current related to temperature The leakage current, leakage current Related to temperature The relationship can be expressed by the following formula: , Is the leakage current at a certain (selectable by personnel) reference temperature and supply voltage Is the operating supply voltage , , , , And Are empirically determined technical constants

[0099] Therefore, the total power consumption of the CPU and GPU Is expressed as: , Since the CPU and GPU support dynamic frequency and voltage scaling (DVFS) during actual application, the actual power consumption changes with the load, frequency, and voltage. Considering the power consumption of multiple components inside the CPU and GPU, the actual energy consumption , Is the capacitance of the th component Is the voltage of the th component Is the frequency of the th component Is the leakage current of the th component related to temperature th component

[0100] For other components (such as memory, storage devices, motherboard, etc.), their power consumption is usually linearly related to the load and can be modeled as: ,

[0101] Is the power consumption coefficient of the memory Is the load of the memory Is the power consumption coefficient of the storage device Is the load of the storage device Is the baseline power consumption of the motherboard and other fixed components. This formula considers the linear relationship between the power consumption of different components and the load, as well as the baseline power consumption of the fixed components

[0102] Combining the power consumption of the CPU, GPU, and other components, the total system power consumption + .

[0103] According to the law of conservation of energy, most of the power consumption of the system is ultimately converted into heat, which can be expressed as , where \(t\) is the running time of the computer. However, due to the existence of heat dissipation paths in the system (such as fans, radiators, etc.), the heat will not accumulate infinitely. The temperature change of the system is controlled by the heat conduction equation , where \(\Delta T\) is the difference between the system temperature and the ambient temperature, \(R\) is the thermal resistance of the system (K / W), representing the heat dissipation efficiency. \(A\) is the heat dissipation area (m²), and \(T_{a}\) is the ambient temperature.

[0104] For different components (CPU, GPU, memory, etc.), the heat generation and temperature are affected by their respective power consumption and heat dissipation characteristics. Assuming that the thermal resistance of each component is different, ultimately, the temperatures of the components within the system can be comprehensively represented by the superposition and heat conduction model: , where \(T_{sys}\) is the comprehensive temperature of the components within the system, \(R_{i}\) is the thermal resistance of the \(i\)-th component, and \(P_{i}\) is the power consumption of the \(i\)-th component. The ambient temperature \(T_{a}\) has a direct impact on the system temperature, especially in the case of low heat dissipation efficiency. Considering that the heat dissipation efficiency is inversely proportional to the ambient temperature, the heat dissipation efficiency \(\eta\) can be modeled as: .

[0105] The power consumption - heat model is a more accurate power consumption and heat model designed after considering more complex factors (such as the frequencies, voltages, ambient temperatures of the CPU and GPU, and complex heat transfer processes). This model combines the effects of dynamic frequency and voltage scaling (DVFS), component - level thermal characteristics, and ambient temperature, and is modeled through physical and statistical methods.

[0106] Data acquisition and implementation module: used to acquire real - time data sets, import the real - time data sets into the power consumption - heat model, and obtain the converted data sets;

[0107] Step S2: Import the real - time data sets into the power consumption - heat model to obtain the converted data sets;

[0108] The real - time data sets include the capacitance, frequency, voltage, temperature, and thermal resistance of each component of the CPU and GPU, as well as the ambient temperature;

[0109] According to the real - time data sets, use the power consumption - heat model to calculate the temperatures of each component, the total power consumption of the all - in - one computer, and the heat dissipation efficiency;

[0110] Form a conversion data set from the temperatures of each component, the total power consumption of the all-in-one computer, and the heat dissipation efficiency.

[0111] When performing thermal management on the computer, hardware optimization is also indispensable, including optimizing the design of heat pipes and heat sinks. For example, using heat pipe materials with high thermal conductivity (such as copper, graphene composite materials) to improve the heat transfer efficiency; and increasing the effective heat dissipation area of the heat sink, while optimizing the heat sink structure to enhance the air flow performance. Or optimizing the air duct design and fans. For example, the air duct design adopts a partitioned heat dissipation strategy to isolate the heat dissipation paths of heat sources (CPU, GPU) from other modules to avoid heat accumulation.

[0112] Trend prediction and mode selection module: used to draw a trend prediction curve based on the conversion data set, judge the state of the all-in-one computer according to the curve, and select an operating mode according to the state.

[0113] Step S3: Draw a trend prediction curve based on the conversion data set, analyze and judge the state of the all-in-one computer according to the curve, and select an operating mode according to the state.

[0114] The method of drawing a trend prediction curve based on the conversion data set, judging the state of the all-in-one computer according to the curve, and selecting an operating mode according to the state includes:

[0115] Preset a cycle time, discretize the current cycle time into time points, and obtain the conversion data set at each time point with a fixed sampling frequency.

[0116] Taking the chronological time points as the abscissa, and taking the temperatures of each component in the conversion data set and the total power consumption of the all-in-one computer as the horizontal and vertical coordinates respectively, draw a two-dimensional coordinate system, and perform curve fitting for each two-dimensional coordinate system. Curve fitting can be obtained using a neural network or an ensemble learning method to obtain the curve expressions of the temperatures of each component and the total power consumption of the all-in-one computer over the cycle time.

[0117] Calculate the state performance coefficient according to the curve expression. , The state performance coefficient comprehensively considers the fluctuations and performance accumulations of temperature and total power consumption within the cycle time, making the judgment more accurate and refined. If is greater than the boundary threshold, it is judged that the all-in-one computer has an overheating phenomenon, and the thermal management program is started. If is less than or equal to the boundary threshold, it is judged that the all-in-one computer does not have an overheating phenomenon, and it continues to work normally.

[0118] The boundary threshold is obtained through data analysis based on the performance of the state performance coefficient formula in experimental data, or can be preset by the management personnel according to experience.

[0119] Among them, and is the weight coefficient, and its value range is within 0 to 1, is the number of discrete time points within the cycle time, is the index of the time point, represents the time point the value at, represents the time point the value at, represents the integral of the curve expression within the cycle time;

[0120] Thermal management module: used to design the operating modes under the thermal management program, and the design content includes a dynamic frequency and voltage regulation unit and a power consumption and heat dissipation collaborative optimization unit;

[0121] Step S4: Design the operating modes under the thermal management program, and design a dynamic frequency and voltage regulation unit and a power consumption and heat dissipation collaborative optimization unit according to the designed operating modes;

[0122] The operating modes under the thermal management program include a performance priority mode, a quiet priority mode, and a balanced mode;

[0123] The performance priority mode means setting high thresholds for power consumption and temperature under high load, that is, allowing higher power consumption and temperature under high load; the quiet priority mode means restricting the fan speed and power consumption; the balanced mode means balancing between performance and noise;

[0124] Introduce a mode adaptive switching mechanism, aiming to reduce the frequency of manual adjustment by the user and automatically select the most suitable working mode for the current conditions. Specifically , where, is the safety threshold of the ambient temperature, exceeding which may affect system stability or user comfort, is the average value of the current component load, reflecting the average workload of the system. is the preset load threshold, used to distinguish different working modes. is the low load threshold, below which the system can switch to a quieter mode.

[0125] The working principle of this mode switching mechanism is as follows:

[0126] Performance priority mode: When the system load is higher than the threshold and the ambient temperature is within the safe range, the system will switch to the performance mode to provide the maximum processing power.

[0127] Balanced mode: When the system load is close to the threshold, the system will switch to the balanced mode, aiming to balance between performance and noise / power consumption.

[0128] Silent Priority Mode: When the ambient temperature exceeds the safety threshold or the system load is below the low load threshold, the system will switch to the silent mode to reduce noise and power consumption and protect the system from damage due to overheating.

[0129] The design method of the dynamic frequency and voltage regulation unit includes:

[0130] Introduce a multi-level dynamic regulation strategy according to the real-time load and temperature, and define a frequency dynamic regulation mechanism and a voltage dynamic regulation mechanism for each component of the all-in-one computer;

[0131] Frequency dynamic regulation mechanism , where represents the frequency after the component is dynamically adjusted, , , and are the preset frequencies in each temperature and load state, and the values are arranged from high to low. They can be pre-formatted by the manufacturer during production or customized by the user. is the maximum frequency, usually corresponding to the highest performance state. is the minimum frequency, usually corresponding to the lowest power consumption state. , are the intermediate frequency states, used to balance performance and power consumption. , and are the preset multi-level temperature thresholds, and the values are arranged from low to high. represents the real-time temperature of the component, represents the real-time load of the component, and are the preset load thresholds;

[0132] Voltage dynamic regulation mechanism , where represents the voltage after the CPU or GPU is dynamically adjusted, is the reference voltage of the CPU or GPU, that is, the voltage of the CPU or GPU under normal working conditions of the device, is the reference temperature of the CPU or GPU, usually a preset safe or optimal working temperature, is the sensitivity coefficient of temperature to voltage, used to quantify the degree of influence of temperature change on voltage adjustment.

[0133] The voltage dynamic regulation mechanism is based on a comprehensive regulation strategy of temperature and leakage current. The purpose is to reduce the power consumption increase caused by leakage current by dynamically adjusting the voltage, indicating that as the temperature of the CPU or GPU increases, the voltage will decrease relative to the reference temperature. This adjustment helps reduce the leakage current because the leakage current usually increases with temperature. By reducing the voltage, the power consumption caused by the leakage current can be reduced, thereby controlling the overall power consumption and heat generation.

[0134] The design method of the power consumption and heat dissipation co-optimization unit includes:

[0135] The goal of power consumption and heat dissipation co-optimization is to dynamically optimize the power consumption distribution and fan noise control while ensuring the thermal management efficiency, improving the user experience.

[0136] The power consumption and heat dissipation co-optimization unit includes a power consumption-heat dissipation adjustment mechanism and a co-optimization mechanism;

[0137] According to the user experience weight to adjust the power consumption distribution priority of the CPU / GPU.

[0138] The power consumption-heat dissipation adjustment mechanism includes power consumption distribution and optimization objectives. The power consumption distribution is , and the optimization objective is ;

[0139] Among them, is the dynamically allocated power consumption of the CPU or GPU, is the maximum power consumption of the CPU or GPU, is the dynamic adjustment coefficient, used to control the impact of power consumption distribution on noise, is the fan speed, is the total heat generated by the all-in-one computer, is the fan noise, is the temperature fluctuation penalty coefficient, used to reduce the impact of rapid temperature fluctuations on system stability, refers to the change rate of the temperature of the component in the all-in-one computer, used to suppress overheating;

[0140] The optimization objective is to refine the balance objective of total power consumption and fan noise into a dynamic optimization problem. The working principle of this model is as follows:

[0141] When is larger, it means that the user prefers to reduce noise more. Therefore, term will be larger, resulting in decreasing, thereby reducing power consumption and heat generation and reducing noise;

[0142] When is smaller, it means that the user prefers to improve performance more. Therefore, The item will be smaller, resulting in close to , thus improving performance.

[0143] In high-load scenarios, increasing the heat dissipation efficiency preferentially can be achieved by increasing the fan speed to ensure that the system does not reduce performance due to overheating. In low-load scenarios, reducing noise preferentially can be achieved by reducing the fan speed to reduce noise and improve the user experience.

[0144] The collaborative optimization mechanism is based on the noise spectrum analysis of the fan, optimizes the frequency distribution of the PWM signal, and avoids generating high-frequency noise. The collaborative optimization mechanism is , is the base PWM frequency, that is, the normal operating frequency of the fan, is the frequency adjustment amplitude, which is used to adjust the PWM frequency to optimize the noise spectrum, is the phase, which is used to mitigate the tinnitus effect caused by frequency changes. The working principle of this model is to superimpose a frequency change of a sine wave on the base frequency to optimize the frequency distribution of the PWM signal. The frequency change of the sine wave can effectively disperse the noise energy, reduce the noise intensity at specific frequencies, and thus reduce the overall noise level. The introduction of the phase

[0145] is to mitigate the tinnitus effect caused by frequency changes. The tinnitus effect refers to the fact that when the frequency changes too fast, the human ear may perceive an uncomfortable sound. By adjusting the phase , the frequency change can be smoothed and the tinnitus effect can be reduced.

[0146] The power consumption and heat dissipation collaborative optimization unit introduces a dynamic weight adjustment mechanism. The dynamic weight adjustment mechanism makes the user experience weight dynamic and adjusts it according to the real-time scenario and user mode. The dynamic weight adjustment mechanism is , where represents the dynamic user experience weight, is the base weight; it is set by the user mode. For example, in the silent mode, this weight may be higher, such as 0.6, to reduce noise and heat; in the performance mode, this weight may be lower, such as 0.5, to optimize performance.

[0147] is the environmental factor weight, which is dynamically adjusted based on the environmental temperature. When the environmental temperature is high, this weight may need to be increased to reduce heat and keep the device operating at a safe temperature, and ; is the load factor weight, which is dynamically adjusted based on the current load and the temperature change rate. High load or rapid temperature change may require an increase in this weight to adjust the device's performance and heat dissipation strategy, and ;

[0148] where and are the base weights of the environmental factor and the load factor respectively, and are the sensitivity coefficients of the environment and the load respectively, indicating the degree of influence of the current environment and load on the weight, is the reference temperature, usually set as the nominal environmental temperature during device design. is the number of loads on all current components, is the maximum load on all components, represents the average temperature change of each load, is the time interval over the cycle time.

[0149] Regulation and management module: Regulates the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit;

[0150] Step S5: Obtain the dynamically adjusted frequency and voltage according to the dynamic frequency and voltage regulation unit, obtain the power consumption and fan speed after dynamic allocation of the CPU or GPU according to the power consumption and heat dissipation collaborative optimization unit, and use the obtained frequency and voltage, power consumption and fan speed to regulate the all-in-one computer;

[0151] The method for regulating the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation collaborative optimization unit includes:

[0152] Obtain the dynamically adjusted frequency and voltage according to the dynamic frequency and voltage regulation unit, obtain the power consumption and fan speed after dynamic allocation of the CPU or GPU according to the power consumption and heat dissipation collaborative optimization unit, and use the obtained frequency and voltage, power consumption and fan speed to regulate the all-in-one computer.

[0153] Step S6: Display the analysis report on the computer terminal page and perform interface interaction with the user;

[0154] Visualization module: Used to arrange the analysis report on the computer terminal page and perform interface interaction with the user;

[0155] The analysis report includes the status of the all-in-one computer, the selected operating mode, the obtained frequency and voltage, as well as the power consumption and fan speed.

[0156] The user can perform interactive operations such as clicking, querying, and downloading the analysis report through the computer terminal page (computer display screen, where the computer refers to a smart device, which can be a mobile phone or a tablet).

[0157] Embodiment 2

[0158] Please refer to Figure 2 As shown, for the parts not described in detail in this embodiment, refer to the description in Embodiment 1. A thermal management method for an all-in-one computer is provided, including:

[0159] Step S1: Model the power consumption and heat of the all-in-one computer;

[0160] Step S2: Import the real-time data set into the power consumption-heat model to obtain a conversion data set;

[0161] Step S3: Draw a trend prediction curve based on the conversion data set, analyze and judge the state of the all-in-one computer according to the curve, and select an operating mode according to the state;

[0162] Step S4: Design the operating mode under the thermal management program, and design a dynamic frequency and voltage regulation unit and a power consumption and heat dissipation collaborative optimization unit according to the designed operating mode;

[0163] Step S5: Obtain the dynamically adjusted frequency and voltage according to the dynamic frequency and voltage regulation unit, obtain the dynamically allocated power consumption and fan speed of the CPU or GPU according to the power consumption and heat dissipation collaborative optimization unit, and use the obtained frequency and voltage, power consumption and fan speed to control the all-in-one computer;

[0164] Step S6: Display the analysis report on the computer terminal page and perform interface interaction with the user.

[0165] Embodiment 3

[0166] This embodiment publicly provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the thermal management system of the all-in-one computer provided above.

[0167] Since the electronic device introduced in this embodiment is the electronic device adopted by the thermal management system of an integrated all-in-one computer in the embodiments of the present application, based on the thermal management system of an integrated all-in-one computer introduced in the embodiments of the present application, those skilled in the art can understand the specific implementation manners and various variations of the electronic device in this embodiment. Therefore, the specific implementation of how the electronic device implements the method in the embodiments of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device adopted by the thermal management system of an integrated all-in-one computer in the embodiments of the present application, it falls within the scope of protection of the present application.

[0168] The above formulas are all calculated by removing the dimension and taking their numerical values. The formula is obtained by collecting a large amount of data and performing software simulation to get a formula that is closest to the actual situation. The preset parameters and threshold selection in the formula are set by those skilled in the art according to the actual situation.

[0169] The above is only the preferred implementation manner of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those ordinary technical users in the technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A thermal management system for an all-in-one computer, characterized in that, include: Relationship modeling module: used to model the power consumption and heat of the integrated all-in-one computer, generate the associated expression of power consumption and heat, and obtain the power consumption-heat model; Data collection and implementation module: used to collect real-time data sets, import the real-time data sets into the power consumption-heat model, and obtain the conversion data sets; Trend prediction and mode selection module: used to draw a trend prediction curve based on the converted data set, and judge the status of the all-in-one computer based on the curve, and select the operating mode according to the status; Thermal management module: used to design the operation mode under the thermal management program, including the dynamic frequency and voltage control unit and the power consumption and heat dissipation coordinated optimization unit; The design method of the dynamic frequency and voltage control unit includes: Define frequency dynamic control mechanism and voltage dynamic control mechanism for each component of the all-in-one computer; Frequency dynamic regulation mechanism Among them, f_tz represents the frequency after the component is dynamically adjusted, and f1, f2, f3, and f4 are the preset frequencies under each temperature T co and load L co state, and the values are arranged from high to low. T l , T2, and T3 are preset multi-level temperature thresholds, and the values are arranged from low to high. T co represents the real-time temperature of the component, and L co represents the real-time load of the component, and L1 and L2 are preset load thresholds; Voltage dynamic regulation mechanism Among them, V_tz represents the voltage after the dynamic adjustment of the CPU or GPU, V_ba is the reference voltage of the CPU or GPU, and T ck is the reference temperature of the CPU or GPU, and mg is the sensitivity coefficient of temperature to voltage; The power consumption and heat dissipation collaborative optimization unit includes a power consumption-heat dissipation adjustment mechanism and a collaborative optimization mechanism; the power consumption-heat dissipation adjustment mechanism includes power consumption allocation and optimization targets, and the power consumption and heat dissipation collaborative optimization unit introduces a dynamic weight adjustment mechanism; Control management module: controls the all-in-one computer according to the output of the dynamic frequency and voltage control unit and the power consumption and heat dissipation collaborative optimization unit; Visualization module: displays the analysis report on the computer page and interacts with the user interface.

2. The thermal management system of the integrated all-in-one computer according to claim 1, wherein, The method for modeling the power consumption and heat of the integrated all-in-one computer, generating a correlation expression of the power consumption and heat, and obtaining a power consumption-heat model includes: The power consumption-heat model is constructed based on the power consumption expression, heat correlation formula, and the formula affecting temperature difference and heat dissipation efficiency; The power consumption expression is: P system = P total + P other = ∑ i C i × V i 2 × f i + V i × I leak,i (T) + α RAM × L RAM + α Sto × L Sto + P base ; The heat correlation formula is T system = ∑ j (T amb + R th,j × P j ); Formula for the influence of temperature difference and heat dissipation efficiency Among them, P system represents the total power consumption of the all-in-one computer, P total represents the total power consumption of the CPU and GPU, P other represents the total power consumption of other components, C i represents the capacitance of the i-th component, V i represents the voltage of the i-th component, f i represents the frequency of the i-th component, I leak,i I(T) represents the leakage current I related to the temperature T on the i-th component leak I(T), α RAM represents the power consumption coefficient of the memory, L RAM represents the load of the memory, α Sto represents the power consumption coefficient of the storage device, L Sto represents the load of the storage device, P base represents the baseline power consumption of the motherboard and other fixed components, T system represents the comprehensive temperature of each component inside the all-in-one computer, T amb represents the ambient temperature, R th,j represents the thermal resistance of the j-th component, P j represents the power consumption of the j-th component, η represents the heat dissipation efficiency, R th represents the comprehensive thermal resistance of each component inside the all-in-one computer, Sr represents the heat dissipation area, represents the convective heat transfer coefficient, ΔT represents the difference between the comprehensive temperature of the components inside the all-in-one computer and the ambient temperature; and the leakage current I s is the leakage current at the selected reference temperature and supply voltage, V dd is the voltage when the all-in-one computer is working, and A, β, α, B, δ, and γ are constants.

3. The thermal management system of the integrated all-in-one computer according to claim 2, characterized in that, The real-time data set includes the capacitance, frequency, voltage, temperature and thermal resistance of each component of the CPU and GPU, as well as the ambient temperature; Based on real-time data sets, use power-thermal models to calculate the temperature of each component, the total power consumption of the AIO PC, and the cooling efficiency; The temperature of each component, the total power consumption of the all-in-one computer, and the cooling efficiency form a conversion data set.

4. The thermal management system of the integrated all-in-one computer according to claim 3, characterized in that, The method of drawing a trend prediction curve according to the converted data set, judging the state of the all-in-one computer according to the curve, and selecting the operation mode according to the state includes: Preset the cycle time, discretize the current cycle time into time points, and acquire the conversion data set at a fixed sampling frequency at each time point; With the time point as the horizontal axis, and the temperature of each component in the conversion data set and the total power consumption of the all-in-one computer as the horizontal and vertical axes, a two-dimensional coordinate system is drawn, and curve fitting is performed for each two-dimensional coordinate system to obtain the curve expression of the temperature of each component and the total power consumption of the all-in-one computer in the cycle time; Calculate the state performance coefficient according to the curve expression If Bx is greater than the boundary threshold, it is determined that the all-in-one computer has an overheating phenomenon, and the thermal management program is started. If Bx is less than or equal to the boundary threshold, it is determined that the all-in-one computer does not have an overheating phenomenon, and normal operation continues; where, θ1 and θ2 are weight coefficients, and their value ranges are within 0 to 1, K is the number of discrete time points within the period, k is the index of the time point, Sz k represents the value at time point k, Sz k+1 represents the value at time point k + 1, ∫F(t) represents the integral of the curve expression within the period.

5. The thermal management system of the integrated all-in-one computer according to claim 4, characterized in that, The operation modes under the thermal management program include performance priority mode, quiet priority mode and balanced mode; Performance priority mode means setting high thresholds for power consumption and temperature under high load, silent priority mode means limiting fan speed and power consumption, and balanced mode means striking a balance between performance and noise. Introduce a mode adaptive switching mechanism, specifically where T ys is the safety threshold of the ambient temperature, L avg is the average value of the current component load, L ys is the preset load threshold, L low is the low load threshold.

6. The thermal management system of the integrated all-in-one computer according to claim 5, characterized in that, The design method of the power consumption and heat dissipation collaborative optimization unit also includes: The power consumption allocation is P CPU / GPU = P max(CPU / GPU) ·(1 - gx·λ), and the optimization objective is min(P system , ω f ) = min(η·Q system + λ·N f + ∑ m μ·ΔT_z m ); Among them, P CPU / GPU is the dynamically allocated power consumption of the CPU or GPU, P max(CPU / GPU) is the maximum power consumption of the CPU or GPU, gx is the dynamic adjustment coefficient, ω f is the fan speed, Q system is the total heat generated by the all-in-one computer, N f is the fan noise, μ is the temperature fluctuation penalty coefficient, ΔT_z m refers to the change rate of the temperature of the m-th component in the all-in-one computer; The collaborative optimization mechanism is f_o base the basic PWM frequency, and Δf_o is the frequency adjustment range, and θ is the phase; The dynamic weight adjustment mechanism is λ = λ base + λ env + λ load , where λ represents the dynamic user experience weight, and λ base is the basic weight; λ env is the weight of the environmental factor, and λ env = λ env-ba ×(1 + q_1×(T amb - T ref )); λ load is the weight of the load factor, and Among them, λ env-ba and λ load-ba are the basic weights of the environmental factor and the load factor respectively, q_1 and q_2 are the sensitivity coefficients of the environment and the load respectively, T ref is the reference temperature, L ss is the number of loads on all current components, L max is the maximum load on all components, W_b represents the average temperature change of each load, and Zq is the time interval on the cycle time.

7. The thermal management system of the integrated all-in-one computer according to claim 6, characterized in that, The method for regulating the all-in-one computer according to the outputs of the dynamic frequency and voltage regulation unit and the power consumption and heat dissipation coordinated optimization unit includes: Obtain the dynamically adjusted frequency and voltage according to the dynamic frequency and voltage regulation unit, obtain the dynamically allocated power consumption and fan speed of the CPU or GPU according to the power consumption and heat dissipation co-optimization unit, and use the obtained frequency and voltage, power consumption and fan speed to regulate the all-in-one computer.

8. The thermal management system of the integrated all-in-one computer according to claim 7, characterized in that, The method of displaying the analysis report on the computer terminal page and performing interface interaction with the user includes: The user clicks, queries or downloads the analysis report through the computer terminal page.

9. The thermal management system of the integrated all-in-one computer according to claim 8, wherein, The analysis report includes the status of the all-in-one computer, the selected operating mode, the obtained frequency and voltage, and the power consumption and fan speed.

Images