Heat dissipation method and device based on magnetic control micro robot dynamic tracking chip hotspot and storage medium

Abstract

The present application relates to the technical field of microelectronic device thermal management, and discloses a heat dissipation method and device based on magnetic control micro-robot dynamic tracking of chip hot spots and a storage medium. The method comprises: obtaining chip surface temperature distribution and heat flow density distribution, determining hot spot area, hot spot center position, hot spot area and maximum temperature; guiding cooling medium to flow through the chip, and placing the micro-robot under an applied magnetic field; establishing a translation force balance model to determine the control parameters of the movement of the micro-robot to the hot spot area; establishing a rotational torque balance model to determine the stable rotation frequency of the micro-robot in the hot spot area; determining the initial heat exchange efficiency according to the ratio of the actual heat exchange amount to the maximum heat exchange amount; adjusting the control parameters according to the deviation of the initial heat exchange efficiency and the preset target efficiency until the maximum temperature of the hot spot is lower than the safety temperature or the heat exchange efficiency reaches the target. The present application realizes dynamic tracking of chip hot spots, spot reinforcement heat dissipation and adaptive closed-loop regulation and control, and improves the heat dissipation efficiency.

Description

Technical Field

[0001] This invention relates to the field of thermal management technology for microelectronic devices, specifically to a heat dissipation method, device, and storage medium based on a magnetically controlled microrobot dynamically tracking hot spots on a chip. Background Technology

[0002] With the development of integrated circuits, power devices, microelectromechanical systems (MEMS), and high-performance computing chips, the power consumption per unit area of ​​chips is constantly increasing, making it easy for high-temperature hotspots to form in localized areas. Hotspots are typically characterized by their small size, high heat flux density, and ability to migrate with changes in load. Traditional heat dissipation methods often rely on overall cooling structures or fixed-position enhanced heat dissipation structures. When the location of a hotspot changes, it is difficult to concentrate cooling capacity on the new hotspot area in a timely manner, leading to excessive local temperature rise, thermal stress concentration, accelerated electromigration, and reduced device reliability.

[0003] In active cooling methods, if a controllable microrobot can be introduced into the cooling medium and driven to move near a hot spot using an external magnetic field, and then its local fluid disturbance can be enhanced through self-rotation or oscillation, the cooling capacity can be transferred from the overall area to the hot spot area. However, the microrobot operates in a low Reynolds number environment within the microscale cooling channel, and its motion is affected by magnetic force, viscous drag, pressure gradient force, gravity, buoyancy, wall forces, and rotational resistance torque. Without a complete force analysis and torque balance model, it is difficult to accurately determine the robot's control parameters and to prevent the microrobot from losing its footing, hitting walls, or deviating from the hot spot area. Therefore, it is necessary to propose a complete chip heat dissipation method and device that includes hot spot identification, force modeling, torque analysis, heat transfer evaluation, and feedback control. Summary of the Invention

[0004] This invention provides a heat dissipation method, device, and storage medium based on a magnetically controlled microrobot dynamically tracking chip hotspots, in order to solve the problems that existing heat dissipation methods are difficult to adapt to the dynamic migration of chip hotspots, have insufficient local reinforcement capabilities, and lack mechanical basis for the magnetic control parameters of the microrobot.

[0005] According to a first aspect of the disclosed embodiments of the present invention, the present invention provides a heat dissipation method based on dynamic tracking of chip hotspots by a magnetically controlled microrobot, employing the following technical solution: A heat dissipation method based on dynamic tracking of chip hotspots using a magnetically controlled microrobot includes: Step 1: Obtain the surface temperature distribution of the chip and surface heat flux density distribution Identify chip hotspot areas Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Step 2: Guide the cooling medium to flow through the chip, and place the magnetically controlled microrobot suspended in the cooling medium under the action of an external magnetic field; Step 3: Based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift experienced by the microrobot in the cooling medium, establish a translational force balance model for the microrobot to determine the control parameters for the microrobot to move from its initial position to the area where the chip hotspot is located. Step 4: Based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, establish a rotational torque balance model for the microrobot to determine the stable rotational frequency of the microrobot above or near the hot spot area. Step 5: Based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity after the microrobot has been running stably, determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, so as to evaluate the local initial heat dissipation effect of the chip hot spot area. Step Six: Based on the deviation between the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized and the preset target heat exchange efficiency, adjust the control parameters to change the translational and rotational motion states of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized reaches the preset target heat exchange efficiency.

[0006] Furthermore, the hotspot area Determined by the following criteria: in, This is an effective heat dissipation area for the chip. and Indicates the effective heat dissipation area of ​​the chip Planar position coordinates within, Indicates in t At any given moment, the position on the chip surface Temperature value at that location, To preset the hotspot temperature threshold, Indicates in t At any given moment, the position on the chip surface The heat flux density at that location, This is a preset heat flux density threshold.

[0007] The hotspot center location Determined according to the weighted centroid formula based on temperature and heat flux density: Among them, the weight function for: in, dA=d d , where is a micro-element representing a tiny surface element within the hotspot region. Indicates hotspot area Integrating the area within the occupied two-dimensional plane region. Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area The sum of weights within, Representative at t Time, location The overall heat hazard weight, Representing hotspot areas Inside t Highest temperature at any time Representing hotspot areas Maximum heat flux density at time t and These are the temperature weight and the heat flux density weight, respectively. .

[0008] Furthermore, the force balance model for the translation of the microrobot, based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force, and shear lift experienced by the microrobot in the cooling medium, is established as follows: When the inertia of the microrobot is neglected under low Reynolds number conditions, the translational motion of the microrobot satisfies quasi-static force equilibrium: Therefore, the control parameters for the microrobot to move from its initial position to the region where the chip hotspot is located satisfy the following equation: in, For the quality of microrobots, For the speed of microrobots, It is a magnetic force. For pressure gradient force, For viscous resistance, For gravity, For buoyancy, For shear lift, Forces acting as wall or boundary constraints. The dynamic viscosity of the cooling medium, For the local velocity of the cooling medium, To account for the translational drag correction factor due to near-wall effects, channel constraints, and the influence of the non-spherical structure of the microrobot, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

[0009] Furthermore, the rotational torque balance model of the microrobot, based on the magnetic driving torque, viscous drag torque, and near-wall damping torque experienced by the microrobot, is established as follows: The viscous drag torque is: Under low Reynolds number quasi-steady-state conditions, the stable rotational frequency of the microrobot above or near the hotspot region satisfies: in, For the rotational inertia of the microrobot, To stabilize the rotation frequency of the microrobot, Magnetic driving torque, For local fluid vorticity-induced torque, For viscous drag torque, For near-wall damping moment, This is the rotational drag correction factor. The local rotational speed of the cooling medium at the location of the microrobot. ， For vector differential operators, The dynamic viscosity of the cooling medium, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

[0010] Furthermore, after the microrobot has been running stably, the actual heat exchange between the cooling medium and the chip hotspot is as follows: in, This refers to the area of ​​the chip's hot spot region. This represents the real-time local convective heat transfer coefficient of the microrobot during stable rotation in the hotspot region after it has achieved stable operation. The average surface temperature of the hot spot area after the microrobot has been operating stably. and These are the inlet and outlet temperatures of the cooling medium, respectively.

[0011] Furthermore, after the microrobot is running stably, the maximum heat transfer between the cooling medium and the chip hotspot is: , In order to be in t At any given moment, the highest temperature within the hotspot area: The initial heat exchange efficiency of the hotspot area after the microrobot has stabilized is: in, This is the mass flow rate of the cooling medium. This is the mass flow rate of the cooling medium. The density of the cooling medium, For volumetric flow rate, This refers to the specific heat capacity at constant pressure of the cooling medium.

[0012] Furthermore, the control parameters are updated in a closed loop using an objective function: in, The safe temperature threshold for the chip. To achieve the preset target heat exchange efficiency, For the real-time location of the microrobot, As the real-time hotspot center location, 、 and For preset weighting coefficients, This represents the positive part of the function.

[0013] According to a second aspect of the disclosed embodiments of the present invention, the present invention provides a heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots, employing the following technical solution: A heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots includes: Temperature detection unit, used to acquire the surface temperature distribution of the chip. ; The heat flux density detection unit is used to obtain the surface heat flux density distribution of the chip. ; The hotspot identification unit, connected to the temperature detection unit and the heat flux density detection unit, is used to determine the chip hotspot region based on the surface temperature distribution and the surface heat flux density distribution. Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Cooling medium guiding unit, used to guide the cooling medium to flow across the chip surface; The magnetic field generating unit is used to generate an external magnetic field in the area where the chip is located; A magnetically controlled microrobot, suspended in the cooling medium, is capable of translational and rotational motion under the action of the external magnetic field. The translation control unit, connected to the magnetic field generating unit, is used to establish a translation force balance model based on the magnetic field force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift experienced by the microrobot in the cooling medium, determine the control parameters for the microrobot to move from the initial position to the chip hot spot area, and control the magnetic field generating unit to drive the microrobot to translate to the hot spot area based on the control parameters. A rotation control unit, connected to the magnetic field generating unit, is used to establish a rotation torque balance model based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, determine the stable rotation frequency of the microrobot above or near the hot spot area, and control the magnetic field generating unit to drive the microrobot to rotate in the hot spot area based on the stable rotation frequency. A heat exchange evaluation unit, connected to the temperature detection unit and the heat flux density detection unit, is used to determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity. The feedback control unit is connected to the heat exchange evaluation unit, the translation control unit, and the rotation control unit, respectively. It is used to send adjustment commands to the translation control unit and / or the rotation control unit according to the deviation between the initial heat exchange efficiency and the preset target heat exchange efficiency, so as to change the translation and / or rotation motion state of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency reaches the preset target heat exchange efficiency.

[0014] According to a third aspect of the embodiments disclosed in this invention, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the chip heat dissipation method as described above.

[0015] Compared with the prior art, the present invention has at least the following beneficial effects: The heat dissipation method and apparatus provided by this invention locates the hot spot of the chip by acquiring the surface temperature distribution and heat flux density distribution of the chip, avoiding the blindness that may exist in traditional heat dissipation methods. Then, a magnetically controlled microrobot is placed under the action of an external magnetic field. Control parameters are used to move the microrobot above the hot spot and stabilize its rotation, achieving dynamic tracking and targeted enhanced heat dissipation of the hot spot. Based on this, this invention calculates the initial heat transfer efficiency of the hot spot area after the microrobot has stabilized, to quantitatively evaluate the initial heat dissipation effect under the current control parameters. According to the deviation between the initial heat transfer efficiency and the preset target heat transfer efficiency, the control parameters are adjusted in a closed-loop feedback manner to continuously optimize the translational and rotational motion of the microrobot until the highest temperature of the hot spot drops below a safe threshold or the heat transfer efficiency reaches the expected target; thus achieving the purpose of effectively dissipating heat from the chip hot spot. The above-mentioned control strategy of "initial evaluation, deviation judgment, closed-loop adjustment, and termination upon reaching the target" enables the heat dissipation system to adaptively respond to the dynamic changes of the chip hot spot, ensuring continuous, stable, and efficient localized enhanced heat dissipation of the chip hot spot while avoiding over-regulation and energy waste. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the heat dissipation method based on dynamic tracking of chip hotspots using a magnetically controlled microrobot, provided by the present invention.

[0017] Figure 2 This is a structural block diagram of a heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots, provided by the present invention.

[0018] Figure 3 This is a schematic diagram of the heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots provided by the present invention.

[0019] Reference numerals: 201, Temperature detection unit; 202, Heat flux density detection unit; 203, Hot spot identification unit; 204, Cooling medium guiding unit; 205, Magnetic field generating unit; 206, Magnetically controlled microrobot; 207, Translation control unit; 208, Rotation control unit; 209, Heat exchange evaluation unit; 210, Feedback control unit. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0021] With the development of integrated circuits, power devices, and high-performance computing chips, the power consumption per unit area of ​​chips is constantly increasing. This leads to the formation of high-temperature hotspots in localized areas, characterized by their small size, high heat flux density, and tendency to migrate with load changes. Traditional chip heat dissipation methods often rely on overall cooling structures or fixed-position enhanced heat dissipation structures. When the location of these hotspots shifts with load changes, cooling capacity cannot be concentrated on the new hotspot area in a timely manner, resulting in excessively high localized temperatures. Furthermore, in active cooling methods, when a controllable microrobot is introduced into the cooling medium, its movement within the microscale channel is affected by various complex torques. Without a complete force analysis and torque balance model, it is difficult to accurately determine the robot's control parameters and to prevent the microrobot from losing its footing, colliding with walls, or deviating from the hotspot area.

[0022] To address this issue, the present invention provides a heat dissipation method based on a magnetically controlled microrobot dynamically tracking chip hotspots, comprising the following steps: Step 1: Obtain the surface temperature distribution of the chip and surface heat flux density distribution Identify chip hotspot areas Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Step 2: Guide the cooling medium to flow through the chip, and place the magnetically controlled microrobot suspended in the cooling medium under the action of an external magnetic field; Step 3: Based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift experienced by the microrobot in the cooling medium, establish a translational force balance model for the microrobot to determine the control parameters for the microrobot to move from its initial position to the area where the chip hotspot is located. Step 4: Based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, establish a rotational torque balance model for the microrobot to determine the stable rotational frequency of the microrobot above or near the hot spot area. Step 5: Based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity after the microrobot has been running stably, determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, so as to evaluate the local initial heat dissipation effect of the chip hot spot area. Step Six: Based on the deviation between the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized and the preset target heat exchange efficiency, adjust the control parameters to change the translational and rotational motion states of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized reaches the preset target heat exchange efficiency.

[0023] For ease of understanding, the following explains some key terms in this embodiment: Chip hotspots are localized high-temperature areas on the chip surface where the temperature or heat flux density is significantly higher than the surrounding area. These areas are typically small in size but have high heat flux density, and their location may dynamically shift with changes in the chip's workload.

[0024] A magnetically controlled microrobot is a miniature device, on the order of micrometers, whose translational and rotational movements can be precisely controlled by an external magnetic field. It is typically suspended in a cooling medium and uses the magnetic field to drive localized, enhanced heat dissipation in hot areas of a chip.

[0025] Cooling medium refers to the fluid used to remove heat from the chip during the heat dissipation process, such as water, oil, or specially designed coolant. As the cooling medium flows over the chip surface, it transfers heat away from the chip through convection heat transfer.

[0026] The translational force balance model is a mathematical model used to describe the balance relationship between various forces (such as magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force, and shear lift) experienced by a magnetically controlled microrobot during translational motion in a cooling medium. This model is used to calculate the control parameters required to drive the microrobot to move precisely.

[0027] The rotational torque balance model is a mathematical model used to describe the balance relationship between various torques (such as magnetic driving torque, viscous drag torque, and near-wall damping torque) experienced by a magnetically controlled microrobot during rotational motion in a cooling medium. This model is used to determine the frequency at which the microrobot can rotate stably above or near hot spots to achieve localized fluid disturbance and enhanced heat transfer.

[0028] Heat exchange efficiency refers to the ratio between the actual heat exchanged between the cooling medium and the chip hot spot to the maximum heat exchangeable heat, and is used to quantify and evaluate the local heat dissipation effect of the chip hot spot area.

[0029] This embodiment provides a heat dissipation method based on a magnetically controlled microrobot dynamically tracking chip hotspots, the specific implementation of which is as follows: In step one, the first step is to acquire and identify hotspot information on the chip. This includes acquiring the surface temperature distribution of the chip. and surface heat flux density distribution In one embodiment, the chip surface temperature distribution is obtained by infrared thermometry, an embedded temperature sensing array, thermal resistance network inversion, or a thermal imaging system. The chip surface heat flux density distribution can be obtained by input power distribution, temperature gradient inversion, a finite element heat conduction model, or a thermal resistance network model.

[0030] In practical applications, hotspots on chip surfaces are often characterized by small size, dynamic location migration, and uneven heat flux density distribution. Without accurate hotspot identification criteria and localization algorithms, microrobots may fail to accurately locate hotspot areas, leading to uneven heat dissipation resource allocation or low cooling efficiency. To address this, this invention further proposes a method for determining hotspot areas. Specifically, the hotspot area is determined by the following criteria: in, This is an effective heat dissipation area for the chip. and Indicates the effective heat dissipation area of ​​the chip Planar position coordinates within, This indicates the position of the chip surface at time t. Temperature value at that location, To preset the hotspot temperature threshold, This indicates the position of the chip surface at time t. The heat flux density at that location, This is a preset heat flux density threshold.

[0031] Hotspot Center Determined according to the weighted centroid formula based on temperature and heat flux density: Among them, the weight function for: in, dA=d d , where is a micro-element representing a tiny surface element within the hotspot region. Indicates hotspot area Integrating the area within the occupied two-dimensional plane region. Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area The sum of weights within, Representative at t Time, location The overall heat hazard weight, Representing hotspot areas Inside t Highest temperature at any time Representing hotspot areas Maximum heat flux density at time t and These are the temperature weight and the heat flux density weight, respectively. .

[0032] Chip hotspot area Calculate as follows: The highest temperature in the hotspot area : in, The highest temperature within the hotspot region at time t. (Comprehensive heat hazard weighting) It is a normalized weighted fusion value of temperature and heat flux density at each location within the hotspot region of the chip, used to quantify the degree of harm that location poses to the overall thermal reliability of the chip. The larger the weight value, the higher the thermal risk at that location, and the stronger the urgent need for local heat dissipation.

[0033] Through the above technical solution, this invention introduces a criterion based on dual thresholds of temperature and heat flux density, achieving accurate definition of hotspot areas on the chip surface. By constructing a comprehensive weighting function, the weighting function can take only temperature weight or simultaneously consider heat flux density weight. Weighted fusion of temperature and heat flux density allows for a more scientific quantification of the thermal risk level at each point on the chip surface. The hotspot center position is calculated using a weighted centroid formula, and the weighted total is calculated using area integrals, ensuring that the hotspot localization process covers the distribution characteristics of the entire hotspot area. This allows the microrobot to accurately move to the core area of ​​the hotspot based on the calculated centroid position, maximizing heat dissipation efficiency. When the hotspot is a multi-discrete region, connected component segmentation can be performed first, followed by calculating the area, highest temperature, and center position of each hotspot separately, prioritizing the hotspot with the highest thermal risk as the control target. This dynamic identification mechanism based on physical field distribution characteristics provides a reliable decision-making basis for the microrobot's autonomous tracking, ensuring that the heat dissipation strategy can be adjusted in real time according to changes in chip load.

[0034] The proposed solution in this application utilizes a magnetically controlled microrobot for chip heat dissipation. However, in a microscale flow environment, the microrobot operates in a low Reynolds number region, and its translational motion is influenced by a combination of physical forces, including magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall forces, and shear lift. Without a complete translational force balance model, it is difficult to calculate the control parameters required for the microrobot to move from its initial position to the hotspot region, causing it to deviate from its expected trajectory during movement and consequently affecting heat dissipation efficiency.

[0035] In response, this application further proposes a heat dissipation method based on a magnetically controlled microrobot dynamically tracking chip hotspots. This method establishes a translational force balance model for the microrobot based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force, and shear lift experienced by the microrobot in the cooling medium. When the inertia of the microrobot is neglected under low Reynolds number conditions, the translational motion of the microrobot satisfies quasi-static force equilibrium: Therefore, the control parameters for the microrobot to move from its initial position to the region where the chip hotspot is located satisfy the following equation: in, For the quality of microrobots, For the speed of microrobots, It is a magnetic force. For pressure gradient force, For viscous resistance, For gravity, For buoyancy, For shear lift, Forces acting as wall or boundary constraints. The dynamic viscosity of the cooling medium, For the local velocity of the cooling medium, To account for the translational drag correction factor due to near-wall effects, channel constraints, and the influence of the non-spherical structure of the microrobot, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

[0036] The above equation, derived from a quasi-static force equilibrium model, relates the various forces acting on the microrobot to externally adjustable parameters. Solving this equation determines the external control parameters required under a given cooling medium environment and target trajectory to ensure the microrobot accurately moves from its initial position to the chip's hot spot area. This equation can be solved using numerical solvers (such as finite element analysis software or MATLAB) to obtain real-time control parameters; alternatively, for specific simplified cases, explicit expressions for the control parameters can be derived analytically, enabling faster calculation and control.

[0037] The aforementioned parameters and variables are key components in constructing the translational force equilibrium model and control parameter equations. They represent the inherent properties of the microrobot, its motion state, the various physical forces it experiences, and the fluid properties of the cooling medium. In particular, the translational drag correction coefficient compensates for the complex hydrodynamic effects caused by proximity to walls, channel geometry, and its non-spherical structure during microscale movement, ensuring the accuracy of model predictions. The inherent properties of the microrobot, such as mass and equivalent radius, can be determined during the design and manufacturing stages; fluid properties such as dynamic viscosity and local velocity of the cooling medium can be obtained through real-time sensor measurements or hydrodynamic simulations; and the calculation of various force terms is based on physical laws and empirical formulas. The translational drag correction coefficient can be determined through experimental calibration, numerical simulation (such as computational fluid dynamics (CFD)), or theoretical derivation (such as corrections to Faxen's law).

[0038] Through the above technical solution, this application constructs a translational force balance model for a microrobot in a cooling medium, and specifically considers the quasi-static force balance under low Reynolds number conditions where inertia is negligible. However, in the low Reynolds number environment of microscale cooling channels, the microrobot is not only affected by translational forces, but its rotational motion above the hot spot area is also constrained by various torques. Without a torque balance model for rotational motion, it is difficult to accurately control the rotational frequency of the microrobot, making it difficult for the microrobot to maintain a stable rotational state above or near the hot spot area, and thus unable to effectively enhance local fluid disturbance to achieve efficient heat dissipation.

[0039] In response, based on the above-mentioned scheme in this application, this application further proposes to establish a rotational torque balance model for the microrobot based on the magnetic driving torque, viscous drag torque, and near-wall damping torque experienced by the microrobot: The viscous drag torque is: Under low Reynolds number quasi-steady-state conditions, the stable rotational frequency of the microrobot above or near the hotspot region satisfies: in, For the rotational inertia of the microrobot, To stabilize the rotation frequency of the microrobot, Magnetic driving torque, For local fluid vorticity-induced torque, For viscous drag torque, For near-wall damping moment, This is the rotational drag correction factor. The local rotational speed of the cooling medium at the location of the microrobot. ， For vector differential operators, The dynamic viscosity of the cooling medium, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

[0040] In the above embodiments of this application, establishing a microrobot rotational torque balance model refers to analyzing all the torques experienced by the microrobot during its rotational motion in a fluid environment and establishing their interrelationships to determine the conditions under which the microrobot can maintain a stable rotational state. This model quantifies the rotational dynamics of the microrobot by balancing the driving torque and the drag torque, thereby predicting or controlling the rotational behavior of the microrobot and providing a theoretical basis for subsequent rotational frequency control.

[0041] In some of the embodiments described above in this application, a microrobot is driven to rotate by a magnetic field to enhance local heat transfer. However, in actual operation, the viscous drag torque and near-wall damping torque experienced by the microrobot will change with the environment. If the magnetic driving torque cannot effectively overcome these drags, the microrobot will not be able to maintain synchronous motion with the rotating magnetic field, resulting in a decrease in rotation frequency and thus failing to maintain the expected enhanced heat transfer effect.

[0042] To address this, this application further proposes a heat dissipation method based on a magnetically controlled microrobot dynamically tracking chip hotspots. When the microrobot rotates synchronously with the rotating magnetic field, it has… When the magnetic driving torque is insufficient to overcome the viscous drag torque and the near-wall damping torque, the microrobot desynchronizes. The critical desynchronization rotational frequency is: in, The rotation frequency of the rotating magnetic field. This is the critical desynchronization rotational frequency of the microrobot. , The amplitude of the rotating magnetic field. The magnetic moment amplitude representing the microrobot, For the maximum local fluid vorticity induced torque, For the maximum near-wall damping moment, when At that time, the microrobot rotates synchronously with the magnetic field; when At this time, the microrobot enters a desynchronized rotation state.

[0043] Specifically, synchronous rotation of a microrobot with a rotating magnetic field refers to a state where the actual rotation frequency of the magnetically controlled microrobot matches the rotation frequency of the externally applied rotating magnetic field. In this state, the microrobot can stably follow the magnetic field, thereby continuously disturbing the surrounding fluid and enhancing local heat transfer. This can be achieved by controlling the frequency and amplitude of the external magnetic field to match the inherent magnetic response characteristics of the microrobot; or by optimizing the microrobot's geometry, material magnetization properties, and its suspension state in the cooling medium to reduce its rotational inertia and resistance, making it easier to synchronize with the rotating magnetic field.

[0044] When the magnetic driving torque is insufficient to overcome the viscous drag torque and near-wall damping torque, the microrobot will become desynchronized. The microrobot will be unable to maintain synchronous rotation with the rotating magnetic field, and its own rotational frequency will be lower than the frequency of the rotating magnetic field. Desynchronization leads to a reduction in the microrobot's disturbance to the fluid, thus affecting heat exchange efficiency. This can be addressed by monitoring the difference between the microrobot's actual rotational frequency and the rotating magnetic field frequency in real time; when the difference exceeds a preset threshold, desynchronization is identified.

[0045] To quantify this critical state, this application introduces a critical desynchronization rotational frequency. This frequency refers to the highest rotational frequency at which the microrobot can maintain synchronous rotation under specific fluid and magnetic field conditions. Once the frequency of the rotating magnetic field exceeds this critical value, the microrobot will desynchronize. Based on the critical desynchronization rotational frequency, an upper limit for the rotational frequency of the rotating magnetic field can be determined. This avoids the microrobot losing synchronization due to excessively high frequencies and reduces local disturbances.

[0046] The rotation frequency of the rotating magnetic field refers to the number of rotations the external magnetic field completes per unit time. It can be controlled by adjusting the current phase and frequency of the coils in the external magnetic field generating unit, for example, by using a multiphase AC drive coil array and changing the frequency of the AC current. The amplitude of the rotating magnetic field refers to the magnitude of its strength, which can be changed by adjusting the driving current of the coils in the external magnetic field generating unit. The magnetic moment amplitude of the microrobot is the basis for the force or torque it experiences in the magnetic field and can be preset during manufacturing by selecting high-permeability materials and magnetizing them. The maximum local fluid vorticity-induced torque refers to the maximum rotational torque generated on the microrobot by the fluid vorticity when the cooling medium flows around it. It can be calculated using computational fluid dynamics (CFD) simulation, combining the flow velocity distribution of the cooling medium and the geometry of the microrobot, to determine the local fluid vorticity and its maximum rotational torque on the microrobot. The maximum near-wall damping moment refers to the additional damping moment generated on the rotation of the microrobot due to the increased fluid viscous resistance caused by the wall effect when the microrobot approaches the cooling channel wall or chip surface. It can be calculated using modified Stokes' law or a more complex fluid dynamics model, taking into account the distance between the microrobot and the wall, the viscosity of the cooling medium, and the rotational speed of the microrobot.

[0047] In some of the embodiments described above in this application, a magnetically controlled microrobot is proposed for chip heat dissipation. However, after the microrobot is running stably, there is a lack of an evaluation index to accurately measure the actual heat exchange effect between the cooling medium and the chip hot spot. This makes it impossible to intuitively judge whether the current heat dissipation strategy is effective, and it is also difficult to support the continuous optimization and adjustment of the microrobot's motion state.

[0048] To address this issue, this application proposes a method for evaluating the actual heat transfer between the cooling medium and the chip hotspots after the microrobot has stabilized in operation. This actual heat transfer can be calculated using the following formula: in, This refers to the area of ​​the chip's hot spot region. This represents the real-time local convective heat transfer coefficient of the microrobot during stable rotation in the hotspot region after it has achieved stable operation. The average surface temperature of the hot spot area after the microrobot has been running stably. and These are the inlet and outlet temperatures of the cooling medium, respectively.

[0049] This actual heat transfer represents the heat actually removed from the hot spot area of ​​the chip by the cooling medium after the microrobot has been operating stably. It is a key indicator for evaluating the performance of the heat dissipation system and directly reflects the quality of the cooling effect. Its calculation can be achieved in various ways. For example, in addition to the calculation model based on the convective heat transfer coefficient given in this application, it can also be indirectly calculated by measuring the temperature rise and mass flow rate of the cooling medium before and after flowing through the hot spot area.

[0050] Based on the above-described embodiments, this application further provides a method for calculating the maximum heat transfer between the cooling medium and the chip hotspot after the microrobot has been operating stably. This maximum heat transfer can be calculated using the following formula: in, In order to be in t At any given moment, the highest temperature within the hotspot area: Based on this, the initial heat exchange efficiency of the hot spot area after the microrobot has been operating stably can be calculated using the following formula: in, This is the mass flow rate of the cooling medium. This is the mass flow rate of the cooling medium. Density of the cooling medium For volumetric flow rate, This refers to the specific heat capacity at constant pressure of the cooling medium.

[0051] In some of the solutions described above in this application, a method for chip hotspot heat dissipation based on magnetically controlled microrobots has been proposed, and an evaluation system with heat exchange efficiency as the core has been established. However, in actual operation, the position and intensity of the chip hotspots change dynamically with the load, and the movement state of the microrobot is also affected by factors such as fluid disturbances. Relying solely on open-loop control makes it difficult to ensure that the microrobot always acts at the center of the hotspot and maintains optimal heat exchange efficiency, resulting in the inability to adaptively adjust the heat dissipation effect according to real-time operating conditions.

[0052] To address this, this application further proposes a closed-loop update mechanism based on an objective function. The control parameters are updated in a closed loop using the objective function: in, The safe temperature threshold for the chip. To achieve the preset target heat exchange efficiency, For the real-time location of the microrobot, As the real-time hotspot center location, 、 and For preset weighting coefficients, This represents the positive part of the function.

[0053] Through the above technical solution, this application achieves closed-loop feedback adjustment of the motion control parameters of the magnetically controlled microrobot. Specifically, by introducing the difference between the real-time position of the microrobot and the real-time hotspot center position, the degree to which the microrobot deviates from the hotspot area can be quantified, thereby driving the translation control unit to perform position correction and ensuring that the microrobot always stays near the hotspot center. Simultaneously, by introducing the deviation between the preset target heat transfer efficiency and the actual heat transfer efficiency, the current heat dissipation status can be evaluated in real time, and the rotation control parameters can be dynamically adjusted accordingly. By changing the rotation frequency of the microrobot, local fluid disturbance is enhanced, thereby achieving closed-loop optimization of heat transfer efficiency. Furthermore, by introducing the difference between the chip's safe temperature threshold and the real-time maximum temperature, and combining it with a positive part function, when the hotspot temperature exceeds the safe threshold, this term generates a significant penalty weight, forcing the system to prioritize the cooling strategy, ensuring the chip's thermal safety under extreme loads. By setting preset weight coefficients, position accuracy, heat transfer efficiency, and temperature safety can be weighted and balanced according to the actual application scenario, making the control strategy highly flexible and adaptable. This effectively solves the shortcomings of a single control logic being unable to cope with complex dynamic thermal environments, thereby significantly improving the adaptability, accuracy, and reliability of chip hotspot heat dissipation.

[0054] Figure 2 This is a structural block diagram of the heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots provided by the present invention. Figure 3 This is a schematic diagram of the heat dissipation device based on the dynamic tracking of chip hotspots by a magnetically controlled microrobot provided by the present invention.

[0055] like Figure 2 , 3 As shown, the device includes: Temperature detection unit 201 is used to acquire the surface temperature distribution of the chip. ; The heat flux density detection unit 202 is used to acquire the surface heat flux density distribution of the chip. ; Hotspot identification unit 203, connected to the temperature detection unit and the heat flux density detection unit, is used to determine the chip hotspot region based on the surface temperature distribution and the surface heat flux density distribution. Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Cooling medium guiding unit 204 is used to guide the cooling medium to flow over the chip surface; The magnetic field generating unit 205 is used to generate an external magnetic field in the area where the chip is located; The magnetically controlled microrobot 206 is suspended in the cooling medium and is capable of translational and rotational motion under the action of the external magnetic field. The translation control unit 207 is connected to the magnetic field generating unit and is used to establish a translation force balance model based on the magnetic field force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift force experienced by the microrobot in the cooling medium, determine the control parameters for the microrobot to move from the initial position to the area where the chip hot spot is located, and control the magnetic field generating unit to drive the microrobot to translate to the hot spot area according to the control parameters. The rotation control unit 208, connected to the magnetic field generating unit, is used to establish a rotation torque balance model based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, determine the stable rotation frequency of the microrobot above or near the hot spot area, and control the magnetic field generating unit to drive the microrobot to rotate in the hot spot area based on the stable rotation frequency. The heat exchange evaluation unit 209 is connected to the temperature detection unit and the heat flux density detection unit, and is used to determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity. The feedback control unit 210 is connected to the heat exchange evaluation unit, the translation control unit, and the rotation control unit, respectively. It is used to send adjustment commands to the translation control unit and / or the rotation control unit according to the deviation between the initial heat exchange efficiency and the preset target heat exchange efficiency, so as to change the translation and / or rotation motion state of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency reaches the preset target heat exchange efficiency.

[0056] In practical implementation, the temperature detection unit and heat flux density detection unit work together to acquire thermal state information of the chip surface, providing multi-dimensional data support for the hotspot identification unit and ensuring accurate location of hotspots and heat flux density characteristics. The cooling medium guiding unit and magnetic field generating unit jointly construct the physical environment for the microrobot's movement, where the cooling medium can be water, deionized water, or an aqueous ethylene glycol solution. The translation control unit, based on a translational force balance model, accurately calculates the control parameters required for the microrobot to move from its initial position to the hotspot area by vector superposition of magnetic force, pressure gradient force, viscous drag, and other forces. These parameters include the strength, frequency, and gradient of the applied magnetic field, thus preventing the microrobot from losing its way or deviating from the target path in low Reynolds number environments. The rotation control unit, based on a rotational torque balance model, analyzes the dynamic balance between magnetic drive torque, viscous drag torque, and near-wall damping torque to determine the stable rotational frequency of the microrobot in the hotspot area. By enhancing local fluid disturbance through self-rotation, it significantly improves convective heat transfer capability. The heat transfer evaluation unit monitors the actual heat transfer between the cooling medium and the chip hotspot in real time and calculates the ratio with the theoretical maximum heat transfer capacity to quantify the initial heat transfer efficiency. The feedback control unit dynamically adjusts the translation and rotation control parameters based on the deviation between the efficiency and the preset target value, forming a closed-loop control mechanism to ensure that the microrobot can adapt to changes in the hot spot position and continuously optimize the heat dissipation effect.

[0057] Through the above technical solutions, this application achieves a complete process integration from hotspot perception, hotspot localization, motion driving to heat transfer evaluation and feedback adjustment. Utilizing translational force balance and rotational torque balance models, the motion trajectory and rotational state of the microrobot in the cooling medium can be effectively predicted and controlled, which helps reduce the risk of the microrobot colliding with walls or losing its footing. Simultaneously, the closed-loop feedback mechanism enables the heat dissipation system to adaptively adjust according to changes in the hotspot location, concentrating cooling capacity on localized high-temperature areas, which helps reduce hotspot temperatures and improve chip reliability and lifespan.

[0058] In addition, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the heat dissipation method in any of the above embodiments.

[0059] The above technical solution solidifies the heat dissipation method based on the dynamic tracking of chip hotspots by magnetically controlled microrobots into a computer program and stores it in a non-transitory computer-readable storage medium, which is then executed by a processor. This enables the efficient and automated execution of processes that originally required a lot of manual intervention or complex physical systems, such as real-time hotspot identification, microrobot force and torque modeling, heat exchange efficiency evaluation, and closed-loop feedback control.

[0060] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A heat dissipation method based on dynamic tracking of chip hotspots using a magnetically controlled microrobot, characterized in that, include: Step 1: Obtain the surface temperature distribution of the chip and surface heat flux density distribution Identify chip hotspot areas Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Step 2: Guide the cooling medium to flow through the chip, and place the magnetically controlled microrobot suspended in the cooling medium under the action of an external magnetic field; Step 3: Based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift experienced by the microrobot in the cooling medium, establish a translational force balance model for the microrobot to determine the control parameters for the microrobot to move from its initial position to the area where the chip hotspot is located. Step 4: Based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, establish a rotational torque balance model for the microrobot to determine the stable rotational frequency of the microrobot above or near the hot spot area. Step 5: Based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity after the microrobot has been running stably, determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, so as to evaluate the local initial heat dissipation effect of the chip hot spot area. Step Six: Based on the deviation between the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized and the preset target heat exchange efficiency, adjust the control parameters to change the translational and rotational motion states of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency of the hot spot area after the microrobot has stabilized reaches the preset target heat exchange efficiency.

2. The heat dissipation method according to claim 1, characterized in that, The hotspot area Determined by the following criteria: in, This is an effective heat dissipation area for the chip. and Indicates the effective heat dissipation area of ​​the chip Planar position coordinates within, Indicates in t At any given moment, the position on the chip surface Temperature value at that location, To preset the hotspot temperature threshold, Indicates in t At any given moment, the position on the chip surface The heat flux density at that location, The preset heat flux density threshold; The hotspot center location Determined according to the weighted centroid formula based on temperature and heat flux density: Among them, the weight function for: in, dA=d d , where is a micro-element representing a tiny surface element within the hotspot region. Indicates hotspot area Integrating the area within the occupied two-dimensional plane region. Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area All position coordinates within Weighted total in direction, Used to determine the hotspot area The sum of the weights within, Representative at t Time, location The overall heat hazard weight, Representing hotspot areas Inside t Highest temperature at any time Representing hotspot areas Maximum heat flux density at time t and These are the temperature weight and the heat flux density weight, respectively. .

3. The heat dissipation method according to claim 1, characterized in that, The force balance model for the translation of the microrobot, based on the magnetic force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force, and shear lift experienced by the microrobot in the cooling medium, is established as follows: When the inertia of the microrobot is neglected under low Reynolds number conditions, the translational motion of the microrobot satisfies quasi-static force equilibrium: Therefore, the control parameters for the microrobot to move from its initial position to the region where the chip hotspot is located satisfy the following equation: in, For the quality of microrobots, For the speed of microrobots, It is a magnetic force. For pressure gradient force, For viscous resistance, For gravity, For buoyancy, For shear lift, Forces acting as wall or boundary constraints. The dynamic viscosity of the cooling medium, For the local velocity of the cooling medium, To account for the translational drag correction factor due to near-wall effects, channel constraints, and the influence of the non-spherical structure of the microrobot, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

4. The heat dissipation method according to claim 1, characterized in that, The rotational torque balance model of the microrobot, based on the magnetic driving torque, viscous drag torque, and near-wall damping torque experienced by the microrobot, is established as follows: The viscous drag torque is: Under low Reynolds number quasi-steady-state conditions, the stable rotational frequency of the microrobot above or near the hotspot region satisfies: in, For the rotational inertia of the microrobot, To stabilize the rotation frequency of the microrobot, Magnetic driving torque, For local fluid vorticity-induced torque, For viscous drag torque, For near-wall damping moment, This is the rotational drag correction factor. The local rotational speed of the cooling medium at the location of the microrobot. ， For vector differential operators, The dynamic viscosity of the cooling medium, The equivalent radius of the magnetically controlled microrobot. Pi is the mathematical constant of a circle.

5. The heat dissipation method according to claim 4, characterized in that, When the microrobot rotates synchronously with the rotating magnetic field, there is When the magnetic driving torque is insufficient to overcome the viscous drag torque and the near-wall damping torque, the microrobot desynchronizes. The critical desynchronization rotational frequency is: in, The rotation frequency of the rotating magnetic field. This is the critical desynchronization rotational frequency of the microrobot. , The amplitude of the rotating magnetic field. The magnetic moment amplitude represents the microrobot. For the maximum local fluid vorticity induced torque, For the maximum near-wall damping moment, when At that time, the microrobot rotates synchronously with the magnetic field; when At this time, the microrobot enters a desynchronized rotation state.

6. The heat dissipation method according to claim 1, characterized in that, After the microrobot operates stably, the actual heat exchange between the cooling medium and the chip hotspot is: in, This refers to the area of ​​the chip's hot spot region. This represents the real-time local convective heat transfer coefficient of the microrobot during stable rotation in the hotspot region after it has achieved stable operation. The average surface temperature of the hot spot area after the microrobot has been operating stably. and These are the inlet and outlet temperatures of the cooling medium, respectively.

7. The heat dissipation method according to claim 6, characterized in that, After the microrobot is running stably, the maximum heat transfer between the cooling medium and the chip hotspot is: in, In order to be in t At any given moment, the highest temperature within the hotspot area: The initial heat exchange efficiency of the hotspot area after the microrobot has stabilized is: in, The mass flow rate of the cooling medium. The mass flow rate of the cooling medium. Density of the cooling medium Volumetric flow rate, This refers to the specific heat capacity at constant pressure of the cooling medium.

8. The heat dissipation method according to claim 1, characterized in that, The control parameters are updated in a closed loop using an objective function: in, The safe temperature threshold for the chip. To achieve the preset target heat exchange efficiency, For the real-time location of the microrobot, As the real-time hotspot center location, 、 and For preset weighting coefficients, This represents the positive part of the function.

9. A heat dissipation device based on a magnetically controlled microrobot dynamically tracking chip hotspots, characterized in that, include: Temperature detection unit, used to acquire the surface temperature distribution of the chip. ; The heat flux density detection unit is used to obtain the surface heat flux density distribution of the chip. ; The hotspot identification unit, connected to the temperature detection unit and the heat flux density detection unit, is used to determine the chip hotspot region based on the surface temperature distribution and the surface heat flux density distribution. Hotspot Center Location Chip hotspot area and the highest temperature in the hotspot area ; Cooling medium guiding unit, used to guide the cooling medium to flow across the chip surface; The magnetic field generating unit is used to generate an external magnetic field in the area where the chip is located; A magnetically controlled microrobot, suspended in the cooling medium, is capable of translational and rotational motion under the action of the external magnetic field. The translation control unit, connected to the magnetic field generating unit, is used to establish a translation force balance model based on the magnetic field force, pressure gradient force, viscous drag, gravity, buoyancy, near-wall force and shear lift experienced by the microrobot in the cooling medium, determine the control parameters for the microrobot to move from the initial position to the chip hot spot area, and control the magnetic field generating unit to drive the microrobot to translate to the hot spot area according to the control parameters. A rotation control unit, connected to the magnetic field generating unit, is used to establish a rotation torque balance model based on the magnetic driving torque, viscous drag torque and near-wall damping torque experienced by the microrobot, determine the stable rotation frequency of the microrobot above or near the hot spot area, and control the magnetic field generating unit to drive the microrobot to rotate in the hot spot area based on the stable rotation frequency. A heat exchange evaluation unit, connected to the temperature detection unit and the heat flux density detection unit, is used to determine the initial heat exchange efficiency of the hot spot area after the microrobot has been running stably, based on the ratio of the actual heat exchange between the cooling medium and the chip hot spot to the maximum heat exchange capacity. The feedback control unit is connected to the heat exchange evaluation unit, the translation control unit, and the rotation control unit, respectively. It is used to send adjustment commands to the translation control unit and / or the rotation control unit according to the deviation between the initial heat exchange efficiency and the preset target heat exchange efficiency, so as to change the translation and / or rotation motion state of the magnetically controlled microrobot until the highest temperature of the hot spot is lower than the preset safe temperature or the initial heat exchange efficiency reaches the preset target heat exchange efficiency.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the heat dissipation method as described in any one of claims 1 to 8.

Images