A notebook computer waste heat recycling system and method based on thermoelectric conversion effect

Abstract

This invention discloses a waste heat recovery and utilization system and method for laptops based on thermoelectric conversion effects, belonging to the field of thermoelectric materials and devices technology. It includes: a distributed thermoelectric array module composed of multiple miniature TEG units distributed in high-heat areas of the laptop, generating electricity directly through temperature differences; an adaptive thermal management module including a thermally conductive silicone layer and miniature heat dissipation fins to optimize the temperature difference between the hot and cold ends; a multi-level power management module including a rectifier circuit, a DC-DC boost chip, and a recovery and storage battery device to convert the unstable low-voltage DC power output from the TEGs into stable electrical energy, which is then stored and supplied to the target device; and an intelligent control module that dynamically adjusts the working area of ​​the miniature thermoelectric power generation units based on temperature sensor data, prioritizing power generation at the location with the largest temperature difference, and matching power supply requirements through a PMIC chip. This invention achieves efficient waste heat conversion and utilization through innovative design, solving the problem of low thermoelectric conversion efficiency in miniaturized devices, and achieving energy saving and environmental protection.

Description

Technical Field

[0001] This invention relates to the field of thermoelectric materials and devices, specifically to a system and method for recovering and utilizing waste heat from laptops based on the thermoelectric conversion effect. Background Technology

[0002] As laptop performance continues to improve, its thermal design power (TDP) becomes increasingly prominent: the TDP of modern CPUs / GPUs continues to rise, with high-performance laptops reaching over 45W. Currently, approximately 60% to 70% of the electrical energy in laptops is ultimately converted into waste heat and dissipated into the environment. Traditional cooling methods (heat pipes + fans) only focus on heat dissipation and do not consider energy recovery. Furthermore, the number of electronic devices worldwide is surging, making energy recycling a crucial issue for sustainable development. Currently, battery technology is progressing slowly, but waste heat recovery can extend battery life. Thermoelectric conversion technology, due to its solid-state nature with no moving parts, high reliability, and suitability for applications with small temperature differences, has become an ideal choice for laptop waste heat recovery.

[0003] Current challenges facing waste heat recovery technologies for laptops based on thermoelectric conversion effects include: during laptop operation, components such as the CPU, GPU, and power module generate a large amount of waste heat, which is not effectively utilized, exacerbating greenhouse gas emissions and thermal pollution, further increasing the environmental burden and resource waste; and long-term use of low-power modules such as built-in camera flashes and sensors in laptops can affect battery life. Summary of the Invention

[0004] To address the aforementioned technical problems, a waste heat recovery and utilization system and method for laptops based on thermoelectric conversion effect is provided. This technical solution solves the problem that during the operation of laptops, components such as the CPU, GPU, and power module generate a large amount of waste heat, which is not effectively utilized; and that low-power modules such as built-in camera flashes and sensors in laptops rely on battery power, which may affect battery life with long-term use.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A waste heat recovery system for laptop computers based on thermoelectric conversion effect, comprising:

[0007] Distributed thermoelectric array module: Composed of multiple miniature TEG units, distributed in the high-heat areas of the laptop, including the CPU heat dissipation module substrate, the bending area in the middle of the heat pipe, and the air outlet guide fins, it generates electricity directly through temperature difference and supplies electrical energy to the functional modules of the device.

[0008] Adaptive thermal management module: includes a thermally conductive silicone layer and micro heat dissipation fins to optimize the temperature difference between the hot and cold ends; the hot end forms a directional thermal channel with the CPU top cover through laser micro-welding, and the cold end has a gradient design for fin spacing;

[0009] Multi-level power management module: including rectifier circuit, DC-DC boost chip and energy recovery storage battery device, converts the unstable low-voltage DC power output by TEG into stable power, stores it and supplies it to the target equipment;

[0010] Intelligent control module: Dynamically adjusts the working area of ​​the micro thermoelectric power generation unit based on temperature sensor data, prioritizes the location with the largest temperature difference for power generation, and matches the power supply demand through PMIC chip.

[0011] Preferably, the distributed thermoelectric array module specifically includes:

[0012] Dynamic thermal impedance matching unit: Each TEG unit integrates a miniature heat flux sensor at the bottom to monitor the heat flux distribution in real time; the tilt angle of the unit is dynamically adjusted by a piezoelectric ceramic actuator to align the heat flow direction with the optimal orientation of the thermoelectric lattice;

[0013] Three-dimensional interconnected power supply unit: Laser-induced graphene wiring is used to form circuits directly on the surface of the heat dissipation module. Each region's TEG unit achieves series-parallel adaptive reconfiguration through topology optimization algorithm to maintain the stability of the output voltage within a predetermined temperature range.

[0014] Failure-tolerant mechanism unit: Built-in TEG unit health assessment model, predicts remaining lifetime through Seebeck coefficient decay rate; when the efficiency of a unit drops to a predetermined threshold, the system automatically switches it to a pure heat conduction channel and maintains total power output through current redistribution of adjacent units.

[0015] Preferably, the distributed thermoelectric array module specifically includes:

[0016] CPU heat dissipation module substrate area: Multiple micro TEG units are deployed in an asymmetrical star-shaped arrangement, with the center point offset from the geometric center of the CPU top cover to match the actual hot spot distribution; thick boron nitride thermally conductive insulating film is filled between the TEG units to achieve electrical isolation while maintaining lateral thermal conductivity;

[0017] The bending area in the middle of the heat pipe: Multiple flexible TEG units are deployed, and the substrate uses copper-graphene composite foil to fit the curved shape of the heat pipe; a micro shape memory alloy spring is integrated to automatically increase the contact pressure when the heat pipe temperature exceeds the predetermined temperature;

[0018] Air outlet guide fin area: Multiple airflow driven TEG units are deployed, and the fin surface is processed with micron-level turbulent structure to reduce the cold end temperature by utilizing exhaust airflow convection.

[0019] Preferably, the adaptive thermal management module specifically includes:

[0020] Thermal conduction unit: The thermally conductive silicone layer uses graphene / boron nitride hybrid filler and adds phase change microcapsules to limit the critical temperature of the phase change microcapsules;

[0021] Cold end heat dissipation unit: The miniature heat dissipation fins are designed with a gradient arrangement according to the near heat source area, the far heat source area, and the transition area. The transition area adopts a biomimetic spiral gradient design.

[0022] Heat flux sensor unit: Through an embedded miniature heat flux sensor, heat flux distribution is monitored in real time. Combined with a shape memory alloy actuator, the fin tilt angle is dynamically adjusted. Computational fluid dynamics simulation is used to guide the optimization of the fin tilt angle, so that the heat dissipation efficiency is automatically optimized according to the load. Phonon engineering is introduced to optimize the silicone filler network and optimize the mid-frequency phonon transmission efficiency.

[0023] Preferably, the multi-level power management module specifically includes:

[0024] Front-end signal processing unit: wide input range rectifier circuit, using zero threshold voltage GaN diode, integrating self-biased synchronous rectification technology; dynamic impedance matching network, using digitally adjustable LC resonant circuit, and tracking the TEG internal resistance change in real time through an impedance analyzer;

[0025] The core energy conversion unit is a DC-DC boost chip for wireless energy transmission based on magnetic coupling resonance, enabling adaptive multi-topology switching; the supercapacitor buffer utilizes graphene / CNT composite electrode materials to increase pulse charge and discharge tolerance.

[0026] Back-end energy storage and distribution unit: The hybrid energy storage device includes short-term energy storage using 3D silicon anode lithium batteries; instantaneous buffering using solid-state micro supercapacitors; and an intelligent power distribution network that prioritizes CPU, USB, and battery charging weights and has wireless recharging capabilities.

[0027] Preferably, the intelligent control module specifically includes:

[0028] Multimodal sensor network unit: The distributed temperature sensing array adopts a hybrid network of MEMS infrared thermopile and fiber Bragg grating sensor to realize the reconstruction of the three-dimensional thermal field on the CPU / GPU surface; it integrates a zero-drift operational amplifier to collect the output characteristics of each TEG unit in real time;

[0029] The core unit for intelligent decision-making employs edge computing, runs a thermal-electric coupling model, and customizes energy-efficiency optimization instructions based on the RISC-V instruction set; it features an adaptive PMIC that integrates digital decoupling algorithms to eliminate crosstalk between channels; and it uses federated learning to achieve cross-device fault prediction.

[0030] Actuator network unit: The contact pressure of the TEG unit is dynamically adjusted through a piezoelectric micro-motion platform, and a biomimetic muscle structure is adopted; based on GaN power devices, the series and parallel dynamic reconfiguration of TEG units is realized, and lossless current commutation is supported;

[0031] Key control strategy unit: The temperature difference priority power generation algorithm is used, combined with multi-objective optimization control. Its constraints include reliability constraints, efficiency constraints, and energy consumption constraints. The NSGA-Ⅲ algorithm is used for Pareto front optimization.

[0032] Furthermore, a method for recovering and utilizing waste heat from laptops based on thermoelectric conversion effects, used to implement the aforementioned waste heat recovery and utilization system for laptops based on thermoelectric conversion effects, includes:

[0033] The three-dimensional temperature distribution of the CPU, GPU and heat pipe surfaces is scanned in real time by an infrared thermal imaging sensor array. Based on the gradient descent algorithm, the efficient temperature recovery area is identified and the corresponding micro TEG unit array is activated first.

[0034] The heat flow path is dynamically adjusted using liquid metal thermal interface material, and the flow rate of nanofluid is controlled by an electromagnetic pump to keep the TEG hot end temperature stable within a predetermined temperature range; the tilt angle of the cold end micro heat sink and the PWM duty cycle of the fan are adjusted simultaneously to keep the cold end temperature below the predetermined temperature.

[0035] The millivolt-level voltage output from the TEG is processed in three stages, including charge pump pre-boost, resonant DC-DC conversion, and frequency-segmented charging and discharging of the hybrid energy storage system.

[0036] Predict the CPU power consumption trend in the near future using an LSTM neural network, and dynamically adjust the series and parallel topology of TEG units.

[0037] When the efficiency degradation of a single TEG cell is detected to exceed a predetermined ratio, it is automatically switched to a pure heat conduction channel and the power supply path is redistributed.

[0038] Preferably, the step of real-time scanning of the three-dimensional temperature distribution of the CPU, GPU, and heat pipe surfaces using an infrared thermal imaging sensor array, identifying efficient temperature recovery areas based on a gradient descent algorithm, and preferentially activating the corresponding micro TEG unit arrays specifically includes:

[0039] A hybrid network of quantum dot infrared detectors and MEMS thermopile is used to form a hexagonal monitoring grid with predetermined spacing at the CPU top cover and heat pipe bends.

[0040] Nanoscale heat flux density calculations were performed on hot spots on the silicon chip surface, TEG installation quality assessment was conducted on the thermal resistance of metal interface contacts, and interface material aging detection was performed on the voids inside the thermal paste layer.

[0041] Based on wavelet transform to filter out fan vibration noise, three-dimensional temperature field reconstruction is performed by moving least squares interpolation, and heat flux vector is calculated by combining Fourier's law and Navier-Stokes equations.

[0042] Based on the adaptive gradient descent-based region selection algorithm, the temperature gradient is calculated, the gradient extremum region is located, and the optimal activation sequence is output through temperature difference threshold filtering. A dynamic load balancing strategy is calculated through a multi-objective optimization function.

[0043] Optionally, the method of dynamically adjusting the heat flow path using liquid metal thermal interface materials and controlling the flow rate of nanofluids through an electromagnetic pump to stabilize the TEG hot-end temperature within a predetermined temperature range; and simultaneously adjusting the tilt angle of the cold-end micro heat sink fins and the fan PWM duty cycle to maintain the cold-end temperature below the predetermined temperature specifically includes:

[0044] Liquid metal thermal interface materials include a base alloy doped with diamond nanoparticles, a magnetic additive modified with a surface silane coupling agent, and an encapsulation layer with a microgroove capillary structure.

[0045] The electromagnetic pump is driven by a combination of Halbach array permanent magnets and micro solenoids;

[0046] The tilt angle of the miniature heat sink fins is driven by shape memory alloy wires, and the intelligent speed control of the fan is accomplished through a multi-parameter coupled control model.

[0047] Optionally, the three-stage processing of the millivolt-level voltage output from the TEG includes charge pump pre-boosting, resonant DC-DC conversion, and frequency-segmented charging and discharging of the hybrid energy storage system; dynamically adjusting the TEG unit series-parallel topology by predicting the CPU power consumption trend in the near future using an LSTM neural network specifically includes:

[0048] The charge pump pre-boost stage uses a zero-threshold voltage MOSFET to construct a cross-coupled charge pump with integrated self-oscillation control;

[0049] The resonant DC-DC converter uses an ultra-thin planar transformer and GaN switching transistors, and is controlled by a dual-mode system of frequency conversion and duty cycle conversion.

[0050] The LSTM neural network predicts power consumption through a power prediction model. The time-domain features of this model include CPU utilization, core temperature, and instruction throughput, while the frequency-domain features include the energy distribution from 0.1 to 100 Hz obtained from FFT decomposition.

[0051] TEG topology dynamic reconfiguration is based on GaN matrix switches, with topology modes including: fully parallel connection in high temperature difference steady state; 3 series and 2 parallel connection in transient load impact; and fully series connection in low temperature environment.

[0052] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0053] This invention proposes to directly utilize waste heat from inside a laptop computer to generate electricity through a distributed thermoelectric array module, significantly improving energy efficiency and reducing energy waste. Effective waste heat recovery and utilization help reduce thermal pollution and carbon emissions, enhancing the product's environmental performance. An adaptive thermal management module optimizes the temperature difference between the hot and cold ends, improving thermoelectric conversion efficiency and reducing the burden on traditional cooling systems, thus improving the overall performance and operational stability of the laptop. Lowering thermal resistance and optimizing heat dissipation design helps reduce damage to electronic components caused by overheating, thereby extending the product's lifespan. The ingenious integration of the thermoelectric conversion module and the heat sink not only improves thermoelectric conversion efficiency but also enhances system stability and durability. The ultra-thin design of the thermoelectric module makes it suitable for integration into ultra-thin laptops, meeting the demands of modern electronic products for thinness and lightness. The intelligent control module dynamically adjusts the working area based on real-time temperature data, prioritizing power generation at the location with the largest temperature difference and matching power supply requirements through a PMIC chip, achieving optimal energy utilization. The integrated and compact design of this invention gives it a wide range of applications, not only for laptops but also for other electronic devices requiring efficient heat dissipation and energy recovery. Attached Figure Description

[0054] Figure 1 This is an internal framework diagram of a waste heat recovery and utilization system for laptop computers based on the thermoelectric conversion effect.

[0055] Figure 2 This is an internal framework diagram of a distributed thermoelectric array module;

[0056] Figure 3 This is an internal framework diagram of the adaptive thermal management module;

[0057] Figure 4 This is an internal framework diagram of the intelligent control module;

[0058] Figure 5 This is a flowchart of a method for recovering and utilizing waste heat from laptops based on the thermoelectric conversion effect;

[0059] Figure 6 A diagram of a laptop computer system architecture;

[0060] Figure 7 This is a structural diagram of the heat dissipation module;

[0061] Figure 8 This is a schematic diagram illustrating the implementation principle.

[0062] Figure 9 This is a schematic diagram illustrating the working principle;

[0063] Figure 10 LED boost driver circuit diagram for providing supplemental lighting for the camera;

[0064] Figure 11 A circuit diagram for brightness detection that powers the fill light for a camera. Detailed Implementation

[0065] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0066] Reference Figure 1 As shown, a waste heat recovery system for laptop computers based on thermoelectric conversion effect includes:

[0067] Distributed thermoelectric array module: Composed of multiple miniature TEG units, distributed in the high-heat areas of the laptop, including the CPU heat dissipation module substrate, the bending area in the middle of the heat pipe, and the air outlet guide fins, it generates electricity directly through temperature difference and supplies electrical energy to the functional modules of the device.

[0068] Adaptive thermal management module: includes a thermally conductive silicone layer and micro heat dissipation fins to optimize the temperature difference between the hot and cold ends; the hot end forms a directional thermal channel with the CPU top cover through laser micro-welding, and the cold end has a gradient design for fin spacing;

[0069] Multi-level power management module: including rectifier circuit, DC-DC boost chip and energy recovery storage battery device, converts the unstable low-voltage DC power output by TEG into stable power, stores it and supplies it to the target equipment;

[0070] Intelligent control module: Dynamically adjusts the working area of ​​the micro thermoelectric power generation unit based on temperature sensor data, prioritizes the location with the largest temperature difference for power generation, and matches the power supply demand through PMIC chip.

[0071] It should be noted that the thermo-electric-mechanical co-optimization maximizes the temperature gradient at the hot end and homogenizes the mechanical stress distribution by using the spatial-temperature mapping relationship between the liquid metal thermal interface and the micro TEG array; the electromagnetic-thermal co-simulation is optimized using the COMSOL multiphysics model, in which the angle θ between the thermoelectric unit spacing and the heat flow direction is controlled at 15±5°.

[0072] The efficiency of the energy conversion chain is improved through end-to-end loss control, including optimized thermal interface conduction using laser micro-welding + nano-silver seed layer; optimized thermoelectric conversion using gradient ZT materials; and optimized power management using resonant DC-DC + GaN matrix switch.

[0073] Reference Figure 2 As shown, the distributed thermoelectric array module specifically includes:

[0074] Dynamic thermal impedance matching unit: Each TEG unit integrates a miniature heat flux sensor at the bottom to monitor the heat flux distribution in real time; the tilt angle of the unit is dynamically adjusted by a piezoelectric ceramic actuator to align the heat flow direction with the optimal orientation of the thermoelectric lattice;

[0075] Three-dimensional interconnected power supply unit: Laser-induced graphene wiring is used to form circuits directly on the surface of the heat dissipation module. Each region's TEG unit achieves series-parallel adaptive reconfiguration through topology optimization algorithm to maintain the stability of the output voltage within a predetermined temperature range.

[0076] Failure-tolerant mechanism unit: Built-in TEG unit health assessment model, predicts remaining lifetime through Seebeck coefficient decay rate; when the efficiency of a unit drops to a predetermined threshold, the system automatically switches it to a pure heat conduction channel and maintains total power output through current redistribution of adjacent units.

[0077] CPU heat dissipation module substrate area: Multiple micro TEG units are deployed in an asymmetrical star-shaped arrangement, with the center point offset from the geometric center of the CPU top cover to match the actual hot spot distribution; thick boron nitride thermally conductive insulating film is filled between the TEG units to achieve electrical isolation while maintaining lateral thermal conductivity;

[0078] The bending area in the middle of the heat pipe: Multiple flexible TEG units are deployed, and the substrate uses copper-graphene composite foil to fit the curved shape of the heat pipe; a micro shape memory alloy spring is integrated to automatically increase the contact pressure when the heat pipe temperature exceeds the predetermined temperature;

[0079] Air outlet guide fin area: Multiple airflow driven TEG units are deployed, and the fin surface is processed with micron-level turbulent structure to reduce the cold end temperature by utilizing exhaust airflow convection.

[0080] It should be noted that the dynamic thermal impedance matching unit includes:

[0081] The miniature heat flux sensor employs a hybrid sensing method combining a MEMS thin-film thermopile and a fiber Bragg grating, with a heat flux measurement range of 1–500 W / cm². 2 ;

[0082] The piezoelectric ceramic actuator utilizes a hinged amplification mechanism (5 times the stroke) to improve displacement accuracy to ±50nm, utilizes a multi-layer co-fired ceramic structure to achieve a maximum thrust of 8N, and reduces power consumption to 5mW / axis through a self-powered energy recovery design.

[0083] The lattice orientation optimization algorithm, based on a density functional theory pre-calculation database, calculates the optimal tilt angle according to the heat flow vector, driving the piezoelectric ceramic to rotate.

[0084] The three-dimensional interconnected power supply unit includes:

[0085] The process parameters for laser-induced graphene wiring are as follows: laser wavelength 355nm (ultraviolet), power density 10.5 W / cm 2 The line width / spacing is 30μm / 50μm; the sheet resistance of the conductivity is <0.1Ω / sq and the bending resistance is >1000 cycles (curvature radius 3mm);

[0086] The topology optimization algorithm includes: maximizing the output current (>500mA) through full parallel connection, ΔT>25℃; increasing the voltage to 3.3V±5% through 3 series and 2 parallel connection, 5℃<ΔT<15℃; and suppressing output voltage ripple and transient loads through dynamic switching (<100ns).

[0087] TEG unit health management of the failure-tolerant mechanism unit, online detection of Seebeck coefficient decay rate (ΔS / S0), and prediction of remaining lifetime (error <5%).

[0088] Health assessment model:

[0089] In the formula, Remaining-Life is the remaining lifespan; S(t) is the health state at time t; S0 is the initial health state, usually at time t = 0; α is a constant representing the influence coefficient of temperature on the health state; T(τ) is the temperature at time τ; T ref The reference temperature is usually a standard or base temperature; α = 0.025 (Bi2Te3 aging coefficient), T_ref = 80℃.

[0090] The current redistribution strategy includes: for failed TEG cells, current compensation is provided by adjacent cells, and the driving voltage of parallel cells is increased to adjust the DC-DC duty cycle in order to maintain total power output;

[0091] In extreme cases, phase change material melting protection and liquid metal diversion are used to cope with instantaneous 100°C thermal shock, and PTC thin film preheating and series topology boost are used to cope with -20°C low temperature start-up.

[0092] The CPU heat dissipation module substrate area features an asymmetrical star-shaped arrangement, offset 3mm from the CPU hotspot at the center point, with radial branches spaced at 22.5° intervals. The boron nitride insulating film has a thickness of 0.05mm and a thermal conductivity of 15W / mK (in-plane) and 5W / mK (normal).

[0093] The flexible TEG unit in the bending zone of the heat pipe uses copper-graphene composite foil, and its bending fatigue life is >10. 5 The shape memory alloy spring has a trigger temperature of 60℃ (NiTiNOL-Cu variant) and a pressure gradient of 0.5MPa / ℃ (60-90℃ range). (R=25mm)

[0094] The turbulent structure design of the air outlet guide fin area includes: groove depth of 50μm near the TEG end and 80μm far from the TEG end; surface roughness of Ra=1.6μm near the TEG end and Ra=3.2μm far from the TEG end; and airflow acceleration ratio of 1.8x near the TEG end and 1.2x far from the TEG end.

[0095] Reference Figure 3 As shown, the adaptive thermal management module specifically includes:

[0096] Thermal conduction unit: The thermally conductive silicone layer uses graphene / boron nitride hybrid filler and adds phase change microcapsules to limit the critical temperature of the phase change microcapsules;

[0097] Cold end heat dissipation unit: The miniature heat dissipation fins are designed with a gradient arrangement according to the near heat source area, the far heat source area, and the transition area. The transition area adopts a biomimetic spiral gradient design.

[0098] Heat flux sensor unit: Through an embedded miniature heat flux sensor, heat flux distribution is monitored in real time. Combined with a shape memory alloy actuator, the fin tilt angle is dynamically adjusted. Computational fluid dynamics simulation is used to guide the optimization of the fin tilt angle, so that the heat dissipation efficiency is automatically optimized according to the load. Phonon engineering is introduced to optimize the silicone filler network and optimize the mid-frequency phonon transmission efficiency.

[0099] It should be noted that the graphene / boron nitride hybrid filler material ratio and properties in the hot-end conduction unit include: graphene nanosheets with a content of 25wt%, improving in-plane thermal conductivity (≥180W / mK); hexagonal boron nitride (h-BN) with a content of 15wt%, enhancing electrical insulation (breakdown voltage >5kV / mm); and silicone matrix (PDMS) with a content of 60wt%, providing flexibility and interfacial adhesion.

[0100] The core materials of the phase change microcapsule are paraffin (C22H46) and a silica shell (thickness 100±10nm). Its critical temperature is 80±2℃ (matching the CPU TJmax threshold), which can absorb 15J / g transient thermal shock and reduce the peak hot spot temperature by 8-12℃ in actual tests.

[0101] The gradient fin structure parameters in the cold-end heat dissipation unit are as follows: the fin density in the near-heat source area is 120 fins, the height is 8mm, and the surface is treated with laser micro-texturing (Ra=1.6μm); the fin density in the transition area is 90 fins, the height is 6.5mm, and the surface is treated with biomimetic spiral gradient (pitch 2mm); the fin density in the far-heat source area is 60 fins, the height is 5mm, and the surface is treated with hydrophobic nano-coating (θ=152°).

[0102] The biomimetic spiral gradient design is derived from the aerodynamic structure of eagle feathers. Its spiral angle changes continuously from 18 to 25°, and the airflow acceleration ratio is 1.8x at the near end to 1.2x at the far end. The wind resistance optimization effect is 35% lower than that of traditional straight fins.

[0103] The miniature heat flow sensor in the heat flow sensor unit uses a MEMS thin-film (Pt1000 temperature sensing element) with a measurement range of 1–500 W / cm². 2 Response time < 1ms, 4 sensor nodes deployed per square centimeter;

[0104] The phase transformation temperature of the driving material in the shape memory alloy actuator is adjustable (55-90℃), containing NiTiNOL-Cu (Cu added 3wt%); the strain recovery rate of the displacement output is >99.5%, with 0-15° tilt angle adjustment; the response speed is driven by pulse current (5A / pulse), 50ms (heating) / 80ms (cooling);

[0105] Governing equations:

[0106] In the formula, PWM% is the percentage of pulse width modulation, representing the magnitude of the control output; K p The proportional gain represents the strength of the proportional control; ΔT is the temperature deviation, i.e., the difference between the setpoint and the actual value; K i K represents the integral gain, indicating the strength of the integral control; ∫ΔTdt is the integral term, representing the accumulation of temperature deviation over time; K d The differential gain represents the strength of the differential control. The differential term represents the rate of change of the temperature deviation; the coefficient is tuned using the Ziegler-Nichols method: K p =0.8,K i =0.05,K d =0.3; Forced power off when SMA temperature is detected to be >120℃.

[0107] Computational fluid dynamics simulation optimization, multi-scale modeling, macroscopic model solving RANS equations (k-ωSST turbulence model), boundary condition wind speed 2±0.5m / s (corresponding to laptop fan operating conditions); microscopic model using LBM (lattice Boltzmann method) to analyze the fin surface boundary layer, mesh size is the first layer mesh near the wall y+≈1.

[0108] Phonon transmission path optimization involves constructing an h-BN "thermal bridge" network, induced alignment by an electric field (field strength 1kV / mm), and enhanced interfacial phonon coupling by an aminosilane coupling agent, resulting in a 60% improvement in mid-frequency phonon (1-10THz) transmission efficiency.

[0109] Reference Figure 4As shown, the intelligent control module specifically includes:

[0110] Multimodal sensor network unit: The distributed temperature sensing array adopts a hybrid network of MEMS infrared thermopile and fiber Bragg grating sensor to realize the reconstruction of the three-dimensional thermal field on the CPU / GPU surface; it integrates a zero-drift operational amplifier to collect the output characteristics of each TEG unit in real time;

[0111] The core unit for intelligent decision-making employs edge computing, runs a thermal-electric coupling model, and customizes energy-efficiency optimization instructions based on the RISC-V instruction set; it features an adaptive PMIC that integrates digital decoupling algorithms to eliminate crosstalk between channels; and it uses federated learning to achieve cross-device fault prediction.

[0112] Actuator network unit: The contact pressure of the TEG unit is dynamically adjusted through a piezoelectric micro-motion platform, and a biomimetic muscle structure is adopted; based on GaN power devices, the series and parallel dynamic reconfiguration of TEG units is realized, and lossless current commutation is supported;

[0113] Key control strategy unit: The temperature difference priority power generation algorithm is used, combined with multi-objective optimization control. Its constraints include reliability constraints, efficiency constraints, and energy consumption constraints. The NSGA-Ⅲ algorithm is used for Pareto front optimization.

[0114] It should be noted that the three-dimensional thermal field reconstruction in the multimodal sensing network unit achieves high-frequency dynamic temperature tracking through MEMS infrared thermopile; precise positioning of micro hotspots is achieved through fiber optic grating sensors; and lossless acquisition of millivolt-level TEG signals is achieved through zero-drift operational amplifiers.

[0115] The thermal field modeling algorithm is based on spatial interpolation using the moving least squares (MLS) method, and integrates multispectral data (3-5μm + 8-12μm bands).

[0116] TEG characteristic monitoring and output impedance analysis were performed. The Seebeck coefficient was calculated in real time using a frequency sweep method (10Hz-100kHz). Health assessment was conducted by establishing an exponential decay model (R0) between ΔS / S0 and aging time. 2 >0.99).

[0117] The edge computing architecture of the core unit for intelligent decision-making includes:

[0118] The RISC-V custom instruction set includes THERMAL_FMA, which is a fusion instruction for thermal-electric coupling matrix operations; NSGA3_OPT, which is a hardware acceleration instruction for multi-objective optimization; and FED_UPDATE, which is a federated learning parameter aggregation instruction.

[0119] Thermo-electric coupling model:

[0120] In the formula, The gradient operator is used to calculate the rate of change in space; k is the thermal conductivity, representing the material's ability to conduct heat. ρ is the temperature gradient, representing the rate of temperature change in space; J is the resistivity, representing the degree to which a material impedes the flow of electric current; α is the thermoelectric conversion coefficient, representing the conversion efficiency between electrical energy and thermal energy; and T is the temperature. Indicates heat conduction; ρJ 2 Joule heating refers to the heat generated when an electric current passes through a resistor. This represents the thermoelectric effect, i.e., the current caused by the temperature gradient; the solution method is the finite volume method.

[0121] The biomimetic muscle structure in the actuator network unit includes: the strain velocity is bio-inspired by octopus tentacle muscles, the energy density is bio-inspired by insect flight muscles, and the self-sensing ability is bio-inspired by human proprioceptors; the dynamic range of contact pressure control is 0.1-5N (resolution 0.01N), and the overshoot is <3% (step response test).

[0122] GaN topology reconfiguration switches feature lossless commutation technology, zero-voltage switching (ZVS) timing control, and a dead time of <2ns. The full parallel to 3-series-2-parallel topology mode is suitable for game load surge scenarios with a switching time of 80ns, while the full series to full parallel topology mode is suitable for low-temperature startup with a switching time of 120ns.

[0123] In the key control strategy unit, the temperature difference priority power generation algorithm selects the three units with the largest temperature difference based on morphological gradient detection.

[0124] Multi-objective optimization control, solved using the Pareto front, objective function:

[0125] max(ηTEG), min(T) hotspot ),min(P control )

[0126] In the formula, ηTEG is the thermoelectric power generation efficiency, representing the efficiency with which a thermoelectric generator (TEG) converts heat energy into electrical energy; T hotspot Hot spot temperature refers to the point with the highest temperature in the system; P control To control power;

[0127] The constraints for reliability, efficiency, and energy consumption are as follows: chip junction temperature ≤ 95℃, ΔT ≥ 10℃, and system power consumption ≤ 50mW.

[0128] NSGA-Ⅲ parameters: population size 50, number of reference points 12;

[0129] Reference Figure 5As shown, a method for recovering and utilizing waste heat from laptop computers based on the thermoelectric conversion effect.

[0130] The three-dimensional temperature distribution of the CPU, GPU and heat pipe surfaces is scanned in real time by an infrared thermal imaging sensor array. Based on the gradient descent algorithm, the efficient temperature recovery area is identified and the corresponding micro TEG unit array is activated first.

[0131] The heat flow path is dynamically adjusted using liquid metal thermal interface material, and the flow rate of nanofluid is controlled by an electromagnetic pump to keep the TEG hot end temperature stable within a predetermined temperature range; the tilt angle of the cold end micro heat sink and the PWM duty cycle of the fan are adjusted simultaneously to keep the cold end temperature below the predetermined temperature.

[0132] The millivolt-level voltage output from the TEG is processed in three stages, including charge pump pre-boost, resonant DC-DC conversion, and frequency-segmented charging and discharging of the hybrid energy storage system.

[0133] Predict the CPU power consumption trend in the near future using an LSTM neural network, and dynamically adjust the series and parallel topology of TEG units.

[0134] When the efficiency degradation of a single TEG cell is detected to exceed a predetermined ratio, it is automatically switched to a pure heat conduction channel and the power supply path is redistributed.

[0135] It should be noted that the hot end utilizes the magnetohydrodynamic properties of liquid metal to achieve directional heat flow guidance, and dynamically adjusts the heat flux distribution through an electromagnetic pump; the cold end optimizes the fin tilt angle and fan PWM based on an aerodynamic-thermal coupling model to maintain a power generation threshold of ΔT > 15℃.

[0136] Energy conversion chain efficiency: η = η TEG ×η DC-DC ×η storage =7% × 94% × 98% ≈ 6.5%

[0137] In the formula, η is the overall energy conversion efficiency, η TEG For the conversion efficiency of a thermoelectric generator, η DC-DC For the efficiency of a DC-DC converter, η storage For the efficiency of energy storage systems.

[0138] Reference Figure 6 The diagram shows the system structure of a laptop computer. The top left corner shows the front view of the laptop, and the bottom left corner shows the bottom view. The diagram on the right shows the internal structure of the laptop, including the location of key components such as the TEG module, motherboard, DC-DC boost chip, battery, DCover, cooling module, CPU chip, GPU chip, DCover, and regenerative battery.

[0139] Reference Figure 7 The diagram shows the internal structure of the heat dissipation module. Key components include the TEG hot end (contacting the heat pipes to absorb heat), the TEG cold end (contacting the fins to dissipate heat), the heat pipes (conducting heat from the heat source to the TEG hot end), the heat dissipation fins (increasing the heat dissipation area and improving efficiency), and the fan (controlled by PWM for forced convection cooling). The TEG unit is embedded within the heat dissipation module, integrated with the heat pipes and vapor chamber, avoiding the need for additional space.

[0140] Reference Figure 8 The diagram shown illustrates the implementation principle and requires further consideration. Figure 6 , 7 As shown in the diagram, the layout and connection of the TEG module inside the laptop are as follows: the TEG module, DC-DC boost chip, and energy recovery storage battery are connected in series. The working process is that the heat pipe conducts the heat generated by the heat-generating components such as the CPU to the hot end of the TEG module. The TEG module uses the thermoelectric effect to convert this heat into electrical energy. The generated electrical energy is converted into a suitable voltage and current by the boost chip. The converted electrical energy is used to power low-power modules such as cameras, fill lights, and sensors inside the laptop.

[0141] Reference Figure 9 The diagram shows the working principle and application scenarios.

[0142] Scenario 1: Camera fill light power supply, which directly drives the LED fill light in the camera area by recovering electrical energy, solving the problem of extra power consumption for video calls in low-light environments.

[0143] Scenario 2: Power supply for sensors and peripherals, providing auxiliary power for keyboard backlight, touchpad, and biometric module (fingerprint / facial recognition).

[0144] Reference Figure 10 The diagram shows the LED boost drive circuit for powering the camera's supplementary lighting. This circuit is used to increase the input voltage to drive the LED, and achieves precise current control and brightness adjustment through a feedback resistor.

[0145] U3 (LED boost driver chip): This is the core component of the circuit, responsible for boosting the input voltage and controlling the brightness of the LED; L1 (inductor): Used for energy storage and voltage conversion;

[0146] D1 (Diode): Used to prevent reverse current flow and protect the circuit;

[0147] C1 (input capacitor): Used to smooth the input voltage and reduce voltage fluctuations;

[0148] C2 (output capacitor): Used to smooth the output voltage and ensure stable operation of the LED;

[0149] R1, R2, R3 (resistors): used for current detection and feedback control;

[0150] Working principle: The input voltage is initially smoothed and boosted through inductor L1 and capacitor C1; LED boost driver chip U3 controls the charging and discharging process of inductor L1, thereby increasing the voltage; diode D1 ensures unidirectional current flow and prevents reverse current from damaging the circuit; the output voltage is further smoothed through capacitor C2 to drive the LED; resistors R1, R2, and R3 are used for current detection and feedback to ensure stable circuit operation.

[0151] Reference Figure 11 The diagram shows a brightness detection circuit powered by a camera fill light. This circuit converts light signals into electrical signals and outputs them through an analog-to-digital converter (ADC).

[0152] Photoresistor (R1): Used to detect ambient light intensity; its resistance value changes with light intensity.

[0153] Operational amplifier (U1): Used to amplify the voltage changes generated by the photoresistor;

[0154] Resistors (R2, R3): Used to set the gain and bias of the operational amplifier;

[0155] Reference voltage (V_REF_2V5): Provides a stable 2.5V reference voltage;

[0156] Analog-to-digital converter (ADC): Converts analog voltage signals into digital signals for further processing.

[0157] Working principle: The resistance of the photoresistor R1 changes with the ambient light intensity, causing the voltage across it to change; the operational amplifier U1 is configured as a voltage follower or amplifier to amplify the voltage change of the photoresistor; the amplified voltage signal is further processed by resistors R2 and R3, and then input to the analog-to-digital converter (ADC); the ADC converts the analog voltage signal into a digital signal and outputs it to a microcontroller or other digital device for processing.

[0158] High-precision brightness detection is achieved using an operational amplifier and an analog-to-digital converter. The gain and sensitivity of the circuit can be changed by adjusting the values ​​of resistors R2 and R3.

[0159] In summary, the advantages of this invention are: it proposes an innovative thermoelectric waste heat recovery system, which is integrated inside a laptop computer and achieves efficient conversion and utilization of waste heat through a unique design, thereby achieving the goals of energy saving and environmental protection;

[0160] The ingenious integration of the thermoelectric conversion module and the heat sink not only effectively reduces thermal resistance and improves thermoelectric conversion efficiency, but also extends the product's lifespan and enhances system stability and durability. The thermoelectric module converts waste heat into electrical energy, which not only alleviates the original heat dissipation burden on laptops but also improves the overall performance of the product. The integrated design of this invention is compact and highly integrated, making it ideal for ultra-thin products such as laptops, and it has extremely broad market application prospects.

[0161] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A waste heat recovery and utilization system for laptop computers based on thermoelectric conversion effect, characterized in that, include: Distributed thermoelectric array module: Composed of multiple miniature TEG units, distributed in the high-heat areas of the laptop, including the CPU heat dissipation module substrate, the bending area in the middle of the heat pipe, and the air outlet guide fins, it generates electricity directly through temperature difference and supplies electrical energy to the functional modules of the device. Adaptive thermal management module: includes a thermally conductive silicone layer and micro heat dissipation fins to optimize the temperature difference between the hot and cold ends; The hot end forms a directional heat channel with the CPU top cover through laser micro-welding, while the cold end features a gradient design for fin spacing. Multi-level power management module: including rectifier circuit, DC-DC boost chip and energy recovery storage battery device, converts the unstable low-voltage DC power output by TEG into stable power, stores it and supplies it to the target equipment; Intelligent control module: dynamically adjusts the working area of ​​the micro thermoelectric power generation unit based on temperature sensor data, prioritizes the location with the largest temperature difference to generate electricity, and matches the power supply demand through the PMIC chip; The distributed thermoelectric array module specifically includes: Dynamic thermal impedance matching unit: Each TEG unit integrates a miniature heat flux sensor at the bottom to monitor the heat flux distribution in real time; the tilt angle of the unit is dynamically adjusted by a piezoelectric ceramic actuator to align the heat flow direction with the optimal orientation of the thermoelectric lattice; Three-dimensional interconnected power supply unit: Laser-induced graphene wiring is used to form circuits directly on the surface of the heat dissipation module. Each region's TEG unit achieves series-parallel adaptive reconfiguration through topology optimization algorithm to maintain the stability of the output voltage within a predetermined temperature range. Failure-tolerant mechanism unit: Built-in TEG unit health assessment model, predicts remaining lifetime through Seebeck coefficient decay rate; when the efficiency of a unit drops to a predetermined threshold, the system automatically switches it to a pure heat conduction channel and maintains total power output through current redistribution of adjacent units; The distributed thermoelectric array module specifically includes: CPU heat dissipation module substrate area: Multiple micro TEG units are deployed in an asymmetrical star-shaped arrangement, with the center point offset from the geometric center of the CPU top cover to match the actual hot spot distribution; thick boron nitride thermally conductive insulating film is filled between the TEG units to achieve electrical isolation while maintaining lateral thermal conductivity; The bending area in the middle of the heat pipe: Multiple flexible TEG units are deployed, and the substrate uses copper-graphene composite foil to fit the curved shape of the heat pipe; a micro shape memory alloy spring is integrated to automatically increase the contact pressure when the heat pipe temperature exceeds the predetermined temperature; Air outlet guide fin area: Multiple airflow driven TEG units are deployed, and the fin surface is processed with a micron-level turbulent structure to reduce the cold end temperature by utilizing exhaust airflow convection. The adaptive thermal management module specifically includes: Thermal conduction unit: The thermally conductive silicone layer uses graphene / boron nitride hybrid filler and adds phase change microcapsules to limit the critical temperature of the phase change microcapsules; Cold end heat dissipation unit: The miniature heat dissipation fins are designed with a gradient arrangement according to the near heat source area, the far heat source area, and the transition area. The transition area adopts a biomimetic spiral gradient design. Heat flux sensor unit: Through an embedded miniature heat flux sensor, heat flux distribution is monitored in real time. Combined with a shape memory alloy actuator, the fin tilt angle is dynamically adjusted. Computational fluid dynamics simulation is used to guide the optimization of the fin tilt angle, so that the heat dissipation efficiency is automatically optimized according to the load. Phonon engineering is introduced to optimize the silicone filler network and optimize the mid-frequency phonon transmission efficiency.

2. The waste heat recovery system for laptop computers based on thermoelectric conversion effect according to claim 1, characterized in that, The multi-level power management module specifically includes: Front-end signal processing unit: wide input range rectifier circuit, using zero threshold voltage GaN diode, integrating self-biased synchronous rectification technology; dynamic impedance matching network, using digitally adjustable LC resonant circuit, and tracking the TEG internal resistance change in real time through an impedance analyzer; The core energy conversion unit is a DC-DC boost chip for wireless energy transmission based on magnetic coupling resonance, enabling adaptive multi-topology switching; the supercapacitor buffer utilizes graphene / CNT composite electrode materials to increase pulse charge and discharge tolerance. Back-end energy storage and distribution unit: The hybrid energy storage device includes short-term energy storage using 3D silicon anode lithium batteries; instantaneous buffering using solid-state micro supercapacitors; and an intelligent power distribution network that prioritizes CPU, USB, and battery charging weights and has wireless recharging capabilities.

3. The waste heat recovery and utilization system for laptop computers based on thermoelectric conversion effect according to claim 2, characterized in that, The intelligent control module specifically includes: Multimodal sensor network unit: The distributed temperature sensing array adopts a hybrid network of MEMS infrared thermopile and fiber Bragg grating sensor to realize the reconstruction of the three-dimensional thermal field on the CPU / GPU surface; it integrates a zero-drift operational amplifier to collect the output characteristics of each TEG unit in real time; The core unit for intelligent decision-making employs edge computing, runs a thermal-electric coupling model, and customizes energy-efficiency optimization instructions based on the RISC-V instruction set; it features an adaptive PMIC that integrates digital decoupling algorithms to eliminate crosstalk between channels; and it uses federated learning to achieve cross-device fault prediction. Actuator network unit: The contact pressure of the TEG unit is dynamically adjusted through a piezoelectric micro-motion platform, and a biomimetic muscle structure is adopted; based on GaN power devices, the series and parallel dynamic reconfiguration of TEG units is realized, and lossless current commutation is supported; Key control strategy unit: The temperature difference priority power generation algorithm is used, combined with multi-objective optimization control. Its constraints include reliability constraints, efficiency constraints, and energy consumption constraints. The NSGA-Ⅲ algorithm is used for Pareto front optimization.

4. A method for recovering and utilizing waste heat from a laptop computer based on the thermoelectric conversion effect, comprising a system for recovering and utilizing waste heat from a laptop computer based on the thermoelectric conversion effect according to any one of claims 1-3, characterized in that, include: The three-dimensional temperature distribution of the CPU, GPU and heat pipe surfaces is scanned in real time by an infrared thermal imaging sensor array. Based on the gradient descent algorithm, the efficient temperature recovery area is identified and the corresponding micro TEG unit array is activated first. The heat flow path is dynamically adjusted using liquid metal thermal interface material, and the flow rate of nanofluid is controlled by an electromagnetic pump to keep the TEG hot end temperature stable within a predetermined temperature range; the tilt angle of the cold end micro heat sink and the PWM duty cycle of the fan are adjusted simultaneously to keep the cold end temperature below the predetermined temperature. The millivolt-level voltage output from the TEG is processed in three stages, including charge pump pre-boost, resonant DC-DC conversion, and frequency-segmented charging and discharging of the hybrid energy storage system. Predict the CPU power consumption trend in the near future using an LSTM neural network, and dynamically adjust the series and parallel topology of TEG units. When the efficiency degradation of a single TEG cell is detected to exceed a predetermined ratio, it is automatically switched to a pure heat conduction channel and the power supply path is redistributed.

5. The method for recovering and utilizing waste heat from laptop computers based on thermoelectric conversion effect according to claim 4, characterized in that, The process of using an infrared thermal imaging sensor array to scan the three-dimensional temperature distribution of the CPU, GPU, and heat pipe surfaces in real time, identifying efficient temperature recovery areas based on a gradient descent algorithm, and preferentially activating the corresponding micro TEG unit arrays specifically includes: A hybrid network of quantum dot infrared detectors and MEMS thermopile is used to form a hexagonal monitoring grid with predetermined spacing at the CPU top cover and heat pipe bends. Nanoscale heat flux density calculations were performed on hot spots on the silicon chip surface, TEG installation quality assessment was conducted on the thermal resistance of metal interface contacts, and interface material aging detection was performed on the voids inside the thermal paste layer. Based on wavelet transform to filter out fan vibration noise, three-dimensional temperature field reconstruction is performed by moving least squares interpolation, and heat flux vector is calculated by combining Fourier's law and Navier-Stokes equations. Based on the adaptive gradient descent-based region selection algorithm, the temperature gradient is calculated, the gradient extremum region is located, and the optimal activation sequence is output through temperature difference threshold filtering. A dynamic load balancing strategy is calculated through a multi-objective optimization function.

6. The method for recovering and utilizing waste heat from laptop computers based on thermoelectric conversion effect according to claim 5, characterized in that, The method of dynamically adjusting the heat flow path using liquid metal thermal interface materials and controlling the flow rate of nanofluids through an electromagnetic pump to stabilize the TEG hot end temperature within a predetermined temperature range; and simultaneously adjusting the tilt angle of the cold end micro heat sink fins and the fan PWM duty cycle to maintain the cold end temperature below the predetermined temperature specifically includes: Liquid metal thermal interface materials include a base alloy doped with diamond nanoparticles, a magnetic additive modified with a surface silane coupling agent, and an encapsulation layer with a microgroove capillary structure. The electromagnetic pump is driven by a combination of Halbach array permanent magnets and micro solenoids; The tilt angle of the miniature heat sink fins is driven by shape memory alloy wires, and the intelligent speed control of the fan is accomplished through a multi-parameter coupled control model.

7. The method for recovering and utilizing waste heat from a laptop computer based on thermoelectric conversion effect according to claim 6, characterized in that, The three-stage processing of the millivolt-level voltage output from the TEG includes charge pump pre-boosting, resonant DC-DC conversion, and frequency-segmented charging and discharging of the hybrid energy storage system; the dynamic adjustment of the TEG unit series-parallel topology by predicting the CPU power consumption trend in the near future using an LSTM neural network specifically includes: The charge pump pre-boost stage uses a zero-threshold voltage MOSFET to construct a cross-coupled charge pump with integrated self-oscillation control; The resonant DC-DC converter uses an ultra-thin planar transformer and GaN switching transistors, and is controlled by a dual-mode system of frequency conversion and duty cycle conversion. The LSTM neural network predicts power consumption through a power prediction model. The time-domain features of this model include CPU utilization, core temperature, and instruction throughput, while the frequency-domain features include the energy distribution from 0.1 to 100 Hz obtained from FFT decomposition. TEG topology dynamic reconfiguration is based on GaN matrix switches, with topology modes including: fully parallel connection in high temperature difference steady state; 3 series and 2 parallel connection in transient load impact; and fully series connection in low temperature environment.

Images