A power device air-cooled radiator model generation method and optimization method

Abstract

The application belongs to the technical field of power device heat dissipation design, and discloses a power device air-cooled radiator model, an optimization method and a performance calculation method. Step one: determining radiator structure parameters; establishing a radiator model; step two: determining the radiator optimization variables as the radiator fin gap b and the radiator fin thickness d; step three: selecting an optimization target; step four: taking the weight m of the radiator as the optimization target and the highest temperature Tmax on the radiator as the constraint function, performing optimization calculation on the above-mentioned radiator mathematical model based on the optimization design program of Matlab programming, and obtaining the radiator structure with the temperature not exceeding the constraint value and the lightest weight; and step five: taking the highest temperature Tmax on the radiator as the optimization target and the weight m of the radiator as the constraint function, performing optimization calculation on the above-mentioned radiator mathematical model based on the optimization design program of Matlab programming, and obtaining the radiator structure with the weight not exceeding the constraint value and the lowest temperature.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power device heat dissipation design, and particularly relates to a power device air-cooled heat sink model generation method and optimization method. BACKGROUND

[0002] With the gradual increase of the power of the aircraft power supply system and the lightweight development of the power electronic conversion device, the power device heat dissipation technology has become one of the key technologies to be solved. Forced air cooling is one of the main forms of power module heat dissipation design of the motor controller at present. Compared with the liquid cooling heat dissipation technology, the air-cooled heat dissipation structure is simple, easy to manufacture and install, and therefore is widely used in the heat dissipation of power devices of controllers and other equipment. The IGBT power module air-cooled heat sink mainly includes a substrate and fins, and the factors affecting the heat dissipation effect of the heat sink include the substrate, the heat sink structure, the fin size and the fin gap. Usually, the selection of the heat sink mainly adopts engineering experience supplemented by Icepak or Flotherm heat design software for preliminary simulation analysis to obtain the heat sink structure parameters meeting the requirements, and sometimes multiple iteration simulations are needed to obtain satisfactory results. This method not only requires high heat design engineering experience of the designer, but also requires the designer to be familiar and thorough with the use of the heat dissipation design software to obtain the final result, which increases the design cost and has poor operability. SUMMARY

[0003] The application provides a power device air-cooled heat sink model generation method and optimization method, improves the reliability of the power device air-cooled heat dissipation, simplifies the process and steps of the air-cooled heat dissipation design, and realizes rapid research and development and design.

[0004] Technical scheme one

[0005] A power device air-cooled heat sink model, comprising: a heat sink pressure drop calculation mathematical model and a thermal resistance calculation mathematical model.

[0006] The pressure drop calculation mathematical model comprises:

[0007] Heat sink pressure drop ;

[0008] wherein is the air density, which is a constant, is the rib gap inlet flow rate, ;

[0009] f app is the apparent friction coefficient,

[0010] ;

[0011] d his the hydraulic diameter of the flow channel, ;

[0012] Re f is the Reynolds number for pressure drop calculation, ;

[0013] wherein is the air volume of the fan, is the kinematic viscosity of air, specified at ambient temperature is a constant;

[0014] Kc 1 is the inlet resistance loss coefficient ;

[0015] Kc 2 is the outlet resistance loss coefficient ;

[0016] wherein, L is the length of the heat sink, W is the width of the heat sink, t is the thickness of the substrate, H is the height of the fin, b is the gap of the fin, d is the thickness of the fin, and n is the equivalent number of the fin, Ls is the length of the power module, Ws is the width of the power module, and x is the distance of the power module; wherein , the proportion of the heat sink slot ;

[0017] The thermal resistance calculation mathematical model comprises: the heat resistance Rh of the heat sink is the sum of the heat resistance Rsp of the substrate of the heat sink and the heat resistance Rcov of the air side convection heat transfer of the fin;

[0018] The heat resistance Rsp of the substrate of the heat sink ;

[0019] The heat resistance Rcov of the air side convection heat transfer ;

[0020] wherein the dimensionless number ;

[0021] The dimensionless number ;

[0022] The Biot number is calculated ;

[0023] The empirical parameter ;

[0024] The equivalent convection heat transfer coefficient ;

[0025] The fin efficiency wherein , is the heat conductivity of the fin of the heat sink;

[0026] The air convection heat transfer coefficient ;

[0027] wherein, k is the thermal conductivity of air, which is a constant, Pr is the Prandtl number, which can be calculated from the specific heat capacity, dynamic viscosity and thermal conductivity of air;

[0028] equivalent radius of the heat sink substrate ;

[0029] equivalent radius of the heat source ;

[0030] dimensionless characteristic radius ;

[0031] dimensionless substrate thickness ;

[0032] Re L is the Reynolds number for thermal resistance calculation, .

[0033] Further, the equivalent radius of the heat source in the heat sink substrate spreading thermal resistance model is r s related to the distance x between the power module and the heat sink, and is expressed by a quadratic polynomial as follows:

[0034] ;

[0035] Take a plurality of different x values, and establish a thermal simulation model according to the heat sink structure parameters to perform finite element simulation. According to the air-cooled heat sink temperature simulation results, the least square method is used to obtain a correction coefficient , so that the error between the calculation results of the mathematical model and the simulation results is minimized, and the equivalent radius of the heat source is determined according to the correction coefficient r s .

[0036] Further, when the heat sink structure parameters are determined, the heat sink pressure drop calculation model is wherein is a function of the heat sink pressure drop with respect to q;

[0037] For the selected fan, the P-Q curve is wherein P is the fan pressure, is a function of the fan pressure with respect to q, and the fan P-Q curve and the heat sink pressure drop calculation mathematical model are solved simultaneously to make , then , and the fan working point air volume q and pressure P are obtained.

[0038] Further, the highest temperature on the heat sink is wherein Q is the heat source power consumption, is the ambient temperature;​

[0039] When the heat sink structure parameters are determined, the heat sink thermal resistance calculation mathematical model is , wherein is the function of the heat sink thermal resistance about q, the heat sink thermal resistance is calculated according to the air volume q of the fan working point, and then the highest temperature of the heat sink is obtained .

[0040] Technical solution two

[0041] A power device air-cooled heat sink optimization method, comprising the following steps:

[0042] Step one: determining the heat sink structure parameters: heat sink length L, heat sink width W, substrate thickness t, rib height H, rib gap b, rib thickness d and rib equivalent number n; determining the heat source structure parameters: power module length Ls, power module width Ws, power module distance x; step two: determining the heat sink optimization variables as the heat sink rib gap b and the heat sink rib thickness d;

[0043] Step three: selecting the optimization target, if the optimization target is the weight of the heat sink, then entering step four, if the optimization target is the highest temperature of the heat sink, then entering step five;

[0044] Step four: taking the weight m of the heat sink as the optimization target, taking the highest temperature Tmax on the heat sink as the constraint function, performing optimization calculation on the above heat sink mathematical model based on the optimization design program of Matlab programming, and obtaining the heat sink structure with the temperature not exceeding the constraint value and the lightest weight;

[0045] Step five: taking the highest temperature Tmax on the heat sink as the optimization target, taking the weight m of the heat sink as the constraint function, performing optimization calculation on the above heat sink mathematical model based on the optimization design program of Matlab programming, and obtaining the heat sink structure with the temperature not exceeding the constraint value and the lowest weight.

[0046] Further, in the step two, when the heat sink structure parameters are determined, the heat sink pressure drop calculation model is , is the function of the heat sink pressure drop about q, b and d;

[0047] The fan P-Q curve and the heat sink pressure drop calculation mathematical model are solved together, so that ,

[0048] The air volume q of the fan working point is the function of the rib gap b and the rib thickness d, , is the function of the air volume about b and d;

[0049] When the heat sink structure parameters are determined, the heat sink thermal resistance calculation mathematical model is , is a function of q, b and d; the highest temperature Tmax on the heat sink is , wherein Q is the heat source power consumption, is the ambient temperature; substituting into, we get , is a function of b and d.

[0050] Further, in the fourth step, the nonlinear programming function fmincon in Matlab is used as the optimization function, and the constraint conditions for the variables are determined by the heat sink structure parameters and the heat sink performance parameters, and the constraint conditions are as follows:

[0051]

[0052] In the formula, is the optimization variable; is the lower limit of the optimization variable value, is the upper limit of the optimization variable value, is the fin gap b and the fin thickness d of the heat sink; is a function of b and d, is the temperature constraint value.

[0053] Further, in the fifth step, the nonlinear programming function fmincon in Matlab is used as the optimization function, and the constraint conditions for the variables are determined by the heat sink structure parameters and the heat sink performance parameters, and the constraint conditions are as follows:

[0054]

[0055] In the formula, is the optimization variable; is the lower limit of the optimization variable value, is the upper limit of the optimization variable value, is the fin gap b and the fin thickness d of the heat sink; is a function of b and d, is the temperature constraint value.

[0056] Technical solution three

[0057] A power device air-cooled heat sink heat exchange performance calculation method, for a heat sink with known structure parameters, fan curve and environmental conditions, using the modified mathematical model, the highest temperature of the heat sink, the heat sink pressure drop and the heat sink thermal resistance parameters are calculated.

[0058] This invention proposes a model, optimization method, and performance calculation method for a power device air-cooled heatsink, achieving two main functions: first, given heatsink structural parameters, the heat transfer performance can be quickly obtained by modifying the mathematical model, allowing for an evaluation of whether the designed heatsink meets requirements; second, under constrained conditions, the structural parameters of the heatsink can be obtained by calling an optimization algorithm, enabling optimized heatsink design. This invention improves the accuracy of heatsink heat transfer performance calculation by constructing a mathematical model, enabling rapid calculation, verification, and optimization of heatsink heat dissipation performance. It reduces the engineering experience requirements for designers, lowers the cost and difficulty of power device heat dissipation design, and has practical engineering application value. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the heat sink structural parameter optimization design process;

[0060] Figure 2 This is a diagram of the radiator and heat source structure;

[0061] Figure 3 A schematic diagram of the mathematical model construction process. Detailed Implementation

[0062] According to one aspect of the present invention, a power device air-cooled heat sink model, such as Figure 1 As shown, the structural parameters of the radiator are determined, and a mathematical model of the radiator is constructed. The model includes a mathematical model for calculating the radiator pressure drop and a mathematical model for calculating the thermal resistance.

[0063] like Figure 2 As shown, the heat sink structural parameters include heat sink length L, heat sink width W, substrate thickness t, fin height H, fin gap b, fin thickness d, and equivalent number of fins n. The heat source structural parameters include power module length Ls, power module width Ws, and power module distance x. Radiator slot percentage .

[0064] (1) Pressure drop calculation model

[0065] Hydraulic diameter of flow channel ;

[0066] ;

[0067] In pressure drop calculations, the characteristic length is taken as the hydraulic diameter d of the flow channel. h hour,

[0068] Reynolds number for voltage drop calculation ;

[0069] in For fan airflow, For air kinematic viscosity, constant at ambient temperature ;

[0070] Inlet resistance loss coefficient ;

[0071] Outlet resistance loss coefficient ;

[0072] Apparent friction coefficient

[0073] ;

[0074] Radiator pressure drop ;

[0075] Wherein is air density, constant, is fin gap inlet flow velocity, ;

[0076] Therefore, according to the above calculation, when the radiator structure parameters are determined, the radiator pressure drop calculation model is , wherein is the function of radiator pressure drop about q;

[0077] For the selected fan, its P-Q curve , wherein P is the fan pressure, is the function of fan pressure about q, together with the fan P-Q curve and the radiator pressure drop calculation mathematical model, make , that is , the fan working point is obtained by solving the air volume q and the pressure P.

[0078] (2) Thermal resistance calculation model

[0079] The radiator thermal resistance Rh is the sum of the radiator substrate diffusion thermal resistance Rsp and the fin air side convective heat transfer thermal resistance Rcov. Analyze the relationship between the radiator structure parameters, fluid motion parameters and the radiator thermal resistance, and construct the radiator thermal resistance calculation mathematical model.

[0080] Radiator substrate equivalent radius ;

[0081] Heat source equivalent radius ;

[0082] Dimensionless characteristic radius ;

[0083] Dimensionless substrate thickness ;

[0084] In the thermal resistance calculation, the characteristic length is the fin length L, and the thermal resistance calculation uses Reynolds number ;

[0085] Air convection heat transfer coefficient ;

[0086] wherein, is the thermal conductivity of air, is a constant, is the Prandtl number, which can be calculated from the specific heat capacity, dynamic viscosity and thermal conductivity of air;

[0087] Fin efficiency wherein , is the fin heat transfer coefficient of the heat sink;

[0088] Equivalent convection heat transfer coefficient ;

[0089] Calculate the Biot number ;

[0090] Empirical parameter ;

[0091] Dimensionless number ;

[0092] Dimensionless number ;

[0093] Heat sink baseboard diffusion thermal resistance ;

[0094] Air side convection heat transfer thermal resistance ;

[0095] Heat sink thermal resistance ;

[0096] Maximum temperature on the heat sink is wherein Q is the heat source power consumption, is the ambient temperature;

[0097] Therefore, according to the above calculation, when the heat sink structure parameters are determined, the heat sink thermal resistance calculation mathematical model is wherein is the function of the heat sink thermal resistance about q, according to the air volume q of the fan working point, the heat sink thermal resistance can be calculated, and then the maximum temperature of the heat sink is obtained .

[0098] Preferably, the heat sink mathematical model is modified;

[0099] The calculation of the baseboard diffusion thermal resistance R sp , the non-circular baseboard and the heat source need to be converted into a circular area equivalent,

[0100] And in actual application, the equivalent radius r sThe size of the power module is related to the distance x between the power modules, and cannot be accurately calculated, so a quadratic polynomial about x is used to represent the influence, i.e. The maximum temperature of the heat sink is a function about x. A plurality of different x values are taken, and a thermal simulation model is established according to the heat sink structure parameters to perform finite element simulation, and according to the temperature simulation result of the air-cooled heat sink, a correction coefficient is obtained by using the least square method , so that the error between the calculation result of the mathematical model and the simulation result is minimized.

[0101] According to another aspect of the present application, a power device air-cooled heat sink optimization method is shown as follows Figure 3 , the method comprises the following steps:

[0102] Step 1: determining the heat sink structure parameters: heat sink length L, heat sink width W, substrate thickness t, rib height H, rib gap b, rib thickness d and rib equivalent number n; determining the heat source structure parameters: power module length Ls, power module width Ws, power module distance x; and establishing a power device air-cooled heat sink mathematical model;

[0103] Step 2: determining the heat sink optimization variables as the heat sink rib gap b and the heat sink rib thickness d; when the other heat sink structure parameters are known, the heat sink pressure drop calculation model is , which is a function about q, b and d;

[0104] The fan P-Q curve and the heat sink pressure drop calculation mathematical model are combined, so that , i.e.

[0105] The air volume of the fan working point is a function about the rib gap b and the rib thickness d, i.e. , which is a function about b and d;

[0106] The heat sink thermal resistance calculation mathematical model is , and the maximum temperature Tmax on the heat sink is , which is substituted into , to obtain , which is a function about b and d;

[0107] Step 3: selecting an optimization target, if the optimization target is the heat sink weight, then entering step 4, if the optimization target is the maximum temperature of the heat sink, then entering step 5;

[0108] ​Step four: taking the weight m of the heat sink as the optimization target, taking the highest temperature Tmax on the heat sink as the constraint function, performing optimization calculation on the heat sink mathematical model based on the optimization design program programmed by Matlab, and obtaining the heat sink structure with the temperature not exceeding the constraint value and the lightest weight; specifically:

[0109] The optimization function adopted is the nonlinear programming function fmincon in Matlab, and the constraint conditions for the variables are determined by the heat sink structure parameters and the heat sink performance parameters, that is:

[0110]

[0111] In the formula, is the optimization variable; is the lower limit of the optimization variable, is the upper limit of the optimization variable, is the fin gap b and the fin thickness d of the heat sink; is the function of the weight about b and d, is the temperature constraint value.

[0112] Step five: taking the highest temperature Tmax on the heat sink as the optimization target, taking the weight m of the heat sink as the constraint function, performing optimization calculation on the heat sink mathematical model based on the optimization design program programmed by Matlab, and obtaining the heat sink structure with the temperature not exceeding the constraint value and the lightest weight. Specifically

[0113] The optimization function adopted is the nonlinear programming function fmincon in Matlab, and the constraint conditions for the variables are determined by the heat sink structure parameters and the heat sink performance parameters, that is:

[0114]

[0115] In the formula, is the optimization variable; is the lower limit of the optimization variable, is the upper limit of the optimization variable, is the fin gap b and the fin thickness d of the heat sink; is the function of the temperature about b and d, is the temperature constraint value.

[0116] According to another aspect of the present application, a heat exchange performance calculation method of a power device air-cooled heat sink is provided. In the case that all the structure parameters, fan curves and environmental conditions of a given heat sink are known, the highest temperature of the heat sink, the pressure drop of the heat sink, the heat resistance of the heat sink and other performance parameters can be calculated more accurately by using the modified mathematical model of the heat sink.

[0117] Example 1: Optimization of the structure parameters of the heat sink

[0118] For a heat sink, the structure parameters are: heat sink width W = 200 mm, heat sink length L = 300 mm, substrate thickness t = 7 mm, fin height H = 30 mm. The heat source structure parameters are: length Ls = 50 mm, width Ws = 50 mm, heat source distance x = 15 mm. The total power consumption of the heat source is 345 watts, and the ambient temperature is 20℃.

[0119] The optimization variables of the heat sink are determined as the heat sink fin gap b and the heat sink fin thickness d;

[0120] Taking the minimum weight m of the heat sink as the optimization objective and the maximum temperature Tmax on the heat sink not exceeding 55℃ as the constraint function, the above heat sink mathematical model is subjected to an optimization design program based on Matlab programming, and optimization calculation is performed to obtain the heat sink fin gap b of 5.86 mm and the heat sink fin thickness d of 1 mm that meet the requirements.

[0121] Example 1: Optimization of the structure parameters of the heat sink

[0122] For a heat sink, the structure parameters are: heat sink width W = 200 mm, heat sink length L = 300 mm, substrate thickness t = 7 mm, fin height H = 30 mm. The heat source structure parameters are: length Ls = 50 mm, width Ws = 50 mm, heat source distance x = 15 mm. The total power consumption of the heat source is 345 watts, and the ambient temperature is 20℃.

[0123] The optimization variables of the heat sink are determined as the heat sink fin gap b and the heat sink fin thickness d;

[0124] Taking the minimum temperature Tmax on the heat sink as the optimization objective and the weight m of the heat sink not exceeding 3 kg as the constraint function, the above heat sink mathematical model is subjected to an optimization design program based on Matlab programming, and optimization calculation is performed to obtain the heat sink fin gap b of 2 mm and the heat sink fin thickness d of 1.07 mm that meet the requirements.

[0125] Example 3: Calculation of the heat exchange performance of the heat sink

[0126] For a heat sink, the structure parameters are: heat sink width W = 200 mm, heat sink length L = 300 mm, substrate thickness t = 6 mm, fin height H = 31 mm, fin gap b = 1.5 mm, fin thickness d = 1 mm. The heat source structure parameters are: length Ls = 50 mm, width Ws = 50 mm, heat source distance x = 15 mm. The total power consumption of the heat source is 345 watts, and the ambient temperature is 20℃.

[0127] The highest temperature of the heat sink calculated by the modified calculation model is 49.26℃, the highest temperature of the heat sink simulated by the software is 50.38℃, and the error is 2.23%.

[0128] Example 4: heat exchange performance calculation of heat sink

[0129] For a heat sink, the structural parameters are: heat sink width W=200mm, heat sink length L=300mm, substrate thickness t=8mm, fin height H=29mm, fin gap b=2mm, and fin thickness d=2mm. The heat source structural parameters are: length Ls=50mm, width Ws=50mm, and heat source distance x=15mm. The total power consumption of the heat source is 345W, and the ambient temperature is 20℃.

[0130] The highest temperature of the heat sink calculated by the modified calculation model is 52.67℃, the highest temperature of the heat sink simulated by the software is 51.52℃, and the error is 2.24%.

[0131] Example 5: heat exchange performance calculation of heat sink

[0132] For a heat sink, the structural parameters are: heat sink width W=200mm, heat sink length L=290mm, substrate thickness t=10mm, fin height H=27mm, fin gap b=3mm, and fin thickness d=1mm. The heat source structural parameters are: length Ls=50mm, width Ws=50mm, and heat source distance x=15mm. The total power consumption of the heat source is 345W, and the ambient temperature is 20℃.

[0133] The highest temperature of the heat sink calculated by the modified calculation model is 47.895℃, the highest temperature of the heat sink simulated by the software is 48.41℃, and the error is 1.07%.

[0134] The technical solutions described in the present application, or the technical solutions designed by those skilled in the art inspired by the technical solutions of the present application, can achieve the above technical effects, and are within the protection scope of the present application.

Claims

1. A method for generating a power device air-cooled heatsink model, characterized in that, The model comprises a radiator pressure drop calculation mathematical model and a thermal resistance calculation mathematical model; The pressure drop calculation mathematical model comprises: Radiator pressure drop where p air is the density of air, a constant, V ch is the rib gap inlet flow velocity, f app is the apparent friction coefficient, d h for the flow passage hydraulic diameter, Re f Reynolds number for pressure drop calculation, where q is the fan air volume, k yn is the kinematic viscosity of air, k yn is a constant; Kc1 is the inlet resistance loss coefficient Kc2 is the outlet resistance loss coefficient where L is the length of the heat sink, W is the width of the heat sink, t is the thickness of the substrate, H is the height of the fin, b is the gap between the fins, d is the thickness of the fin, and n is the equivalent number of fins, Ls is the length of the power module, Ws is the width of the power module, and x is the distance of the power module; wherein Heat sink slot ratio The heat resistance calculation mathematical model comprises: a heat sink heat resistance calculation mathematical model R h is a heat sink substrate diffusion heat resistance model R sp and a fin air side convection heat transfer heat resistance model R cov sum; Heat spreader substrate spreading thermal resistance model Ribbed air-side convection heat transfer thermal resistance model where the dimensionless number Dimensionless number calculating the biel number empirical parameters equivalent convective heat transfer coefficient Fins efficiency wherein λ h is the heat conductivity of the fins of the heat sink; Air convection heat transfer coefficient Where, λ air ρ is the thermal conductivity of air, which is a constant, and Pr is the Prandtl number, which is calculated from the specific heat capacity, dynamic viscosity and thermal conductivity of air. Equivalent radius of a heat spreader substrate Heat source equivalent radius Dimensionless characteristic radius Dimensionless substrate thickness Re L Reynolds number for heat resistance calculation, 2. The method of claim 1, wherein: The heat source equivalent radius r in the heat spreader substrate spreading resistance model s The distance x between the power modules is related to the quadratic polynomial as follows: Take multiple different x values, and according to the heat sink structure parameter establishes the thermal simulation model to carry out the finite element simulation, according to the air-cooled radiator temperature simulation result, adopts the least square method to obtain correction coefficient a2, a1, a0, so that the error between the calculation results of mathematical model and simulation results is minimum, according to the correction coefficient determines the equivalent radius r of heat source s .

3. The method of claim 2, wherein: When the heat sink structure parameters are determined, the heat sink pressure drop calculation model is ΔP h = f h (q), where f h is a function of the heat sink pressure drop with respect to q; For a selected fan, its PQ curve is P = f fan (q), where P is the fan pressure, f fan Let the fan pressure be a function of q, and combine this with the fan PQ curve: P = f fan (q) and the mathematical model for calculating the radiator voltage drop, such that P = ΔP h Then f h (q)=f fan (q) is used to solve for the air volume q and pressure P at the fan's operating point.

4. The method of claim 3, wherein: Maximum temperature on the heat sink T max T = T max = Q * R h + T a where Q is the heat source power dissipation, T a is the ambient temperature; When the heat sink structure parameters are determined, the heat sink thermal resistance calculation mathematical model is R h = f R (q), wherein f R is a function of the heat sink thermal resistance about q, the heat sink thermal resistance is calculated according to the air volume q of the fan working point, and then the highest temperature T max of the heat sink is obtained.

5. A method of optimizing a power device air-cooled heat sink, the method comprising: A method for optimizing the radiator model generated by the radiator model generation method of claim 1 or 2, the method comprising the following steps: ​ Step one: determining radiator structure parameters: radiator length L, radiator width W, substrate thickness t, fin height H, fin gap b, fin thickness d and fin equivalent number n; determining heat source structure parameters: power module length Ls, power module width Ws, power module distance x; Step two: determining the radiator optimization variables as the radiator fin gap b and the radiator fin thickness d; Step three: selecting an optimization target, if the optimization target is the radiator weight, then entering step four, if the optimization target is the radiator maximum temperature, then entering step five; Step four: taking the weight m of the radiator as the optimization target, taking the maximum temperature Tmax on the radiator as the constraint function, performing optimization design programming based on Matlab on the above radiator model, performing optimization calculation, and obtaining a radiator structure with a temperature not exceeding a constraint value and a lightest weight; Step five: taking the maximum temperature Tmax on the radiator as the optimization target, taking the weight m of the radiator as the constraint function, performing optimization design programming based on Matlab on the above radiator model, performing optimization calculation, and obtaining a radiator structure with a weight not exceeding a constraint value and a lowest temperature.

6. The method of claim 5, wherein: In the step two, when the structure parameters of the heat sink are determined, the pressure drop calculation model of the heat sink is ΔP h = f h (q, b, d), f h is a function of q, b and d. Simultaneous fan P-Q curve P = f fan (q) and radiator pressure drop calculation mathematical model, P = ΔP h , Solving for the fan operating point air volume q is a function of the fin gap b and the fin thickness d, q = f(b,d) r (b,d), f r is a function of b and d; When the heat sink structure parameters are determined, the heat sink thermal resistance calculation mathematical model is R h = f R (q, b, d), f R is a function of the heat sink thermal resistance with respect to q, b, and d; the highest temperature Tmax on the heat sink is T max = Q * R h + T a , wherein Q is the heat source power consumption, T a is the ambient temperature; substituting q = f r (b, d) into it, T max = f T (b, d), f T is a function of the highest temperature on the heat sink with respect to b and d.

7. The method of claim 6, wherein: In the step four, the nonlinear programming function fmincon in Matlab is used as the optimization function, and the constraint conditions for the variables are determined by the radiator structure parameters and the heat dissipation performance parameters, and the constraint conditions are as follows: In the formula, x i is an optimization variable; x min is a lower limit of the value of the optimization variable, x max is an upper limit of the value of the optimization variable, x is the rib gap b and the rib thickness d of the heat sink. f m T1 is a temperature constraint value.

8. The method of claim 6, wherein: In the step five, the nonlinear programming function fmincon in Matlab is used as the optimization function, and the constraint conditions for the variables are determined by the radiator structure parameters and the heat dissipation performance parameters, and the constraint conditions are as follows: In the formula, x i is an optimization variable; x min is a lower limit of the value of the optimization variable, x max is an upper limit of the value of the optimization variable, x is the rib gap b and the rib thickness d of the heat sink. f T m1 is a weight constraint value, and m2 is a temperature constraint value.

Images