A two-stage non-linear radiation surface space atmospheric density sensor heating cover

Abstract

The application provides a two-section nonlinear radiation surface space atmospheric density sensor heating cover, the top end of which is provided with an opening for the entry and exit of space atmospheric molecules; the radiation surface is divided into two sections, and the generatrix of each section of the radiation surface is a curve, the shape of which follows different nonlinear functions; the bottom end of the heating cover is provided with a mounting seat for fixing with a spacecraft body. The generatrix of each section of the radiation surface of the heating cover is designed according to a self-guided online machine learning method, the undetermined coefficients of the function analytical expressions followed by the generatrix of each section of the radiation surface are selected as factor variables, and the heating time is taken as a target variable, and the best variable design value is obtained by combining a heuristic algorithm with a deep neural network. The two-section nonlinear radiation surface space atmospheric density sensor heating cover provided by the application can enhance the radiation heat transfer characteristics between the sensor front end quartz crystal balance and the sensor, reduce the heating time required for the desorption of space atmospheric molecules, and thus reduce the power consumption of the spacecraft in orbit flight.

Description

Technical Field

[0001] This invention belongs to the field of space atmospheric density measurement technology, specifically relating to a heating cover for a two-section nonlinear radiating surface space atmospheric density sensor. Background Technology

[0002] Space atmospheric density sensors are primarily used to measure and assess the surface density of rarefied gases in space, aiding in the analysis and calculation of drag experienced by spacecraft during orbital flight. The sensor heats a heating shroud via electrical heating, and the radiating surface of the shroud heats a quartz crystal microbalance (QCM) through radiative heat transfer, facilitating the thermal desorption of atmospheric molecules. The surface density of the space atmosphere is then calculated by measuring the frequency drift of the crystal oscillator. The radiative heat transfer characteristics between the radiating surface and the QCM are directly related to the configuration of the inner radiating surface. By optimizing the structural topology of the inner radiating surface, the radiative heat transfer per unit time can be effectively enhanced, allowing the QCM surface temperature to reach the target value more quickly. This effectively reduces the electrical heating required for the space atmospheric density sensor, thus meeting the low power consumption requirements of spacecraft during orbital operation. Currently, many space atmospheric density sensor heating shrouds use a single-segment linear radiating surface, which fails to maximize the radiative heat transfer per unit time to the QCM surface, thus preventing further reduction in heating time and lowering the power consumption of the spacecraft during orbital operation.

[0003] Current space atmospheric density sensors utilize thin-film heating elements, heating shrouds, QCM crystal coatings, and quartz plates. The original heating shroud's radiating surface had a generatrix of a straight line with a certain slope. This single-segment linear radiating surface failed to maximize the heat transfer per unit time to the QCM surface, thus preventing further reductions in heating time and spacecraft power consumption during orbit. Summary of the Invention

[0004] The purpose of this invention is to provide a two-section nonlinear radiative surface space atmospheric density sensor heating cover that can effectively enhance the radiative heat transfer capability between the heating cover and the QCM, reduce the heating time required for the desorption of atmospheric molecules from the crystal oscillator surface, and reduce the power consumption of the spacecraft during on-orbit operation.

[0005] To achieve the above objectives, the present invention provides a heating cover for a two-section nonlinear radiating surface space atmospheric density sensor, characterized in that it includes a housing, an atmospheric molecule inlet at the top of the housing, and a mounting base at the bottom of the housing for mounting and fixing the housing; the outer surface of the housing is in contact with an electric heating element, the atmospheric molecule inlet is the first radiating surface, and the lower end of the atmospheric molecule inlet to the bottom of the housing is the second radiating surface, wherein the generatrix of the first radiating surface and the generatrix of the second radiating surface are both curves.

[0006] The shape of the generatrix of the first radiating surface follows a nonlinear function y = a(e bx -1), the shape of the generatrix of the second radiating surface follows the nonlinear function y=cln(1+dx), where a, b, c, d are undetermined coefficients.

[0007] The design method for the busbars of the first and second radiating surfaces includes the following steps:

[0008] Step 1: Randomly generate m0 groups of a, b, c, d, H using the central composite design method. n Values ​​as sample space, H n Let a be the vertical distance between the boundary line of the first and second radiating surfaces and the top of the heating shroud, subject to the constraint condition: a(e bHn -1)=cln(1+dH n ), cln(1+19.5d)=3.4, 0≤H n ≤19.5, a,b,c,d>0;

[0009] Step 2: Establish a numerical model for heat transfer of the space atmospheric density sensor based on the generated variable sample space, calculate the corresponding heating time τ, and obtain the implicit function τ0 = F(a, b, c, d, H) through a heuristic algorithm. n );

[0010] Step 3: Based on the sample space in Step 1 and the implicit function in Step 2, establish a self-guided online machine learning DNN network model. Through training, obtain the prediction objective function τ1 = f(a, b, c, d, H) between the target and factor variables. n );

[0011] Step 4: Based on the function argmin f(a, b, c, d, H) n Inversely solve for the new variable values ​​a', b', c', d', H. n ', and calculate the actual heating time τ according to step 2, if The predicted value τ1 then satisfies the convergence requirement, and the corresponding a', b', c', d', and H are output. n ' and τ; if the convergence requirement is not met, compare the current a', b', c', d', H n 'Whether it is equal to the existing variable value in the sample space, if so, add it along with the corresponding τ value as a new sample to the DNN network model for continued training; if not, in a', b', c', d', H n Add a perturbation at the location, calculate the τ value according to step 2, and then add it to the DNN network model until τ1 meets the convergence requirement;

[0012] Step 5: Based on the final output a', b', c', d', H n 'Calculate the actual heating time τ according to step 2.

[0013] This invention enhances the radiative heat transfer per unit time to the QCM surface by altering the geometry of the heating shroud's radiating surface, thereby increasing the product of the radiating surface area and the radiative heat transfer angle coefficient between the radiating surface and the QCM. Compared to the original atmospheric density sensor heating shroud, the heating time is reduced by approximately 18.6%. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall structure of the present invention.

[0015] Figure 2 This is a design flowchart as described in the example of the present invention.

[0016] Figure 3 This is the analysis flowchart described in the example of the present invention. Detailed Implementation

[0017] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0018] See Figure 1 The present invention includes a housing 4, an atmospheric molecule inlet 2 formed at the top 1 of the housing 4, and a mounting base 8 located at the bottom of the housing 4 for mounting and fixing the housing 4; wherein the outer surface of the housing 4 is in contact with the electric heating element, the atmospheric molecule inlet 2 is a first radiating surface 6, and the lower end of the atmospheric molecule inlet 2 to the bottom of the housing is a second radiating surface 7, the generatrix 3 of the first radiating surface 6 and the generatrix 5 of the second radiating surface 7 are both curves, and the shape of the generatrix 3 of the first radiating surface 6 follows the nonlinear function y = a(e bx -1), the shape of the generatrix 5 of the second radiation surface 7 follows the nonlinear function y=cln(1+dx), where a, b, c, d are undetermined coefficients.

[0019] In this embodiment, the function followed by the generatrix 3 of the first radiating surface 6 is 1.81(e 0.79x -1), the generatrix 5 of the second radiating surface 7 follows the function 1.72ln(1+45.99x), and the distance between the boundary line of the two radiating surfaces and the top surface 1 of the heating cover is 0.5mm, which is used as an example for illustration.

[0020] like Figure 2 As shown, the design of the optimal configuration of each radiating surface of a two-segment nonlinear radiating surface space atmospheric density sensor heating hood is carried out according to the following steps:

[0021] Step 1: Define factor variables and constraints, and randomly generate the initial sample space using the central composite design method;

[0022] Step 2: Establish a numerical model of heat transfer from a space atmospheric density sensor based on the sample space and calculate the corresponding heating time; and obtain an implicit function based on the variables and heating time in the sample space using a self-guided machine learning optimization method; establish a deep neural network (DNN) model based on the variable values ​​and heating time, and then train it based on the implicit function to obtain an objective function based on the minimum heating time, and search for new variables and predict the corresponding heating time based on the objective function.

[0023] Step 3: Repeat step 2. Stop the calculation when the heating time meets the convergence requirement, and extract the minimum heating time predicted by the neural network model and the corresponding variable value; calculate the actual heating time.

[0024] Further, see Figure 3 :

[0025] Step 1: Randomly generate m0 groups of a, b, c, d, H using the central composite design method. n Values ​​as sample space, H n Let a be the vertical distance between the boundary line of the first and second radiating surfaces and the top of the heating shroud, subject to the constraint condition: a(e bHn -1)=cln(1+dH n ), cln(1+19.5d)=3.4, 0≤H n ≤19.5, a,b,c,d>0;

[0026] Step 2: Establish a numerical model for heat transfer of the space atmospheric density sensor based on the generated variable sample space, calculate the corresponding heating time τ, and obtain the implicit function τ0 = F(a, b, c, d, H) through a heuristic algorithm. n );

[0027] Step 3: Based on the sample space in Step 1 and the implicit function in Step 2, establish a self-guided online machine learning DNN network model. Through training, obtain the prediction objective function τ1 = f(a, b, c, d, H) between the target and factor variables. n );

[0028] Step 4: Based on the function argmin f(a, b, c, d, H) n Inversely solve for the new variable values ​​a', b', c', d', H. n ', and calculate the actual heating time τ according to step 2, if The predicted value τ1 then satisfies the convergence requirement, and the corresponding a', b', c', d', and H are output. n ' and τ; if the convergence requirement is not met, compare the current a', b', c', d', H n'Whether it is equal to the existing variable value in the sample space, if so, add it along with the corresponding τ value as a new sample to the DNN network model for continued training; if not, in a', b', c', d', H n Add a perturbation at the location, calculate the τ value according to step 2, and then add it to the DNN network model until τ1 meets the convergence requirement;

[0029] Step 5: Based on the final output a', b', c', d', H n 'Calculate the actual heating time τ according to step 2.

[0030] In this embodiment, the heating time required for the optimized two-section nonlinear atmospheric density sensor heating hood is reduced by approximately 18.6% compared to the original structure.

Claims

1. A heating shroud for a two-section nonlinear radiating surface space atmospheric density sensor, characterized in that, Includes a shell (4) and an atmospheric molecule inlet (2) opened at the top (1) of the shell (4) and a mounting base (8) located at the bottom of the shell (4) for mounting and fixing the shell (4); the outer surface of the shell (4) is in contact with the electric heating element, the atmospheric molecule inlet (2) is a first radiation surface (6), the lower end of the atmospheric molecule inlet (2) to the bottom of the shell is a second radiation surface (7), the generatrix (3) of the first radiation surface (6) and the generatrix (5) of the second radiation surface (7) are both curves; The design method for the busbar (3) of the first radiating surface (6) and the busbar (5) of the second radiating surface (7) includes the following steps: Step 1: Randomly generate according to the central composite design method m 0 groups a , b , c , d , H n Values ​​as sample space, H n The vertical distance between the boundary line of the first radiating surface (6) and the second radiating surface (7) and the top of the heating shroud (1) is subject to the following constraints: a (e bHn -1)= c ln(1+ dH n ), c ln(1+19.5 d =3.4, 0 ≤ H n ≤ 19.5, a , b , c , d >0; Step 2: Establish a numerical model for heat transfer of the space atmospheric density sensor based on the generated variable sample space, and calculate the corresponding heating time. τ Implicit functions are obtained through heuristic algorithms. τ 0= F ( a , b , c , d , H n ); Step 3: Based on the sample space in Step 1 and the implicit function in Step 2, build a self-guided online machine learning DNN network model, and obtain the prediction objective function between the target and factor variables through training. τ 1= f ( a , b , c , d , H n ); Step 4: Based on the function argmin f ( a , b , c , d , H n Inverse solution to obtain new variable values a ', b ', c ', d ', H n ', and calculate the actual heating time according to step 2. τ ,like Then the predicted value τ 1. Meet the convergence requirements and output the corresponding... a ', b ', c ', d ', H n 'and τ ; If the convergence requirement is not met, compare with the current... a ', b ', c ', d ', H n 'Whether it is equal to the variable value already existing in the sample space; if so, include it along with the corresponding value.' τ The values ​​are used as new samples to supplement the DNN network model for continued training; if none are found, then... a ', b ', c ', d ', H n Add a perturbation at ' and calculate according to step 2'. τ The value is then added to the DNN network model until... τ 1. Until the convergence requirement is met; Step 5: Based on the final output a ', b ', c ', d ', H n 'Calculate the actual heating time according to step 2.' τ .

2. The heating shroud for the two-section nonlinear radiating surface space atmospheric density sensor according to claim 1, characterized in that, The shape of the generatrix (3) of the first radiating surface (6) follows a nonlinear function. y = a (e bx -1), the shape of the generatrix (5) of the second segment of the radiating surface (7) follows a nonlinear function. y = c ln(1+ dx ),in, a , b , c , d These are coefficients to be determined.

Images