CN105760586A - Fuzzy temperature response subordinating degree function solving method based on point collocation theory - Google Patents

Abstract

The invention discloses a fuzzy temperature response subordinating degree function solving method based on a point collocation theory. The method comprises the following steps that a heat transfer structure is discretized to build a finite element equation of a steady-state heat conduction problem; a fuzzy finite element equation of the heat conduction problem is obtained by considering the fuzzy uncertainty of model input parameters; a cut set level is selected, fuzzy variables are converted into interval variables through cut set calculation, and then a set of interval finite element equations are obtained; approximate representation is performed on interval temperature responses in the interval finite element equations through a Legendre polynomial; the change range of the interval temperature responses is solved according to the sparse grid point collocation theory and the function smoothness; all the interval temperature responses at the cut set level are recombined through a fuzzy resolution theorem, and then a fuzzy temperature response subordinating degree function is obtained. According to the method, the steady-state heat conduction problem containing fuzzy uncertain parameters can be systematically solved, and the calculating efficiency of a point collocation method is further improved while the calculating precision is guaranteed.

Description

A kind of based on joining a theoretical Fuzzy temperature response membership function method for solving

Technical field

The invention belongs to mechanical engineering field, be specifically related to a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving.

Background technology

In nature and various production technical field, temperature contrast the heat energy transmission caused is a kind of extremely general physical phenomenon.In manufacturing process particularly in sophisticated products such as industrial equipment and electronic device such as Aero-Space, electromotor, derived energy chemicals, how effectively to realize heat transmission, to improve the thermodynamic property of structure, it has also become an importance of product design.And such issues that premise and crucial first have to determine the regularity of distribution of structure temperature field exactly.Along with the development of computer technology, improve constantly with the numerical technology status in heat transfer problem is studied that Finite Element Method is representative, and in engineering field, play more and more important effect.

Existing it be both for deterministic models about thermoanalytical many researchs and carry out, it does not have consider the uncertainty of mode input parameter.In Practical Project, limitation due to the restriction of manufacturing process, measurement error and cognition, the physical parameters such as the material properties of structure, external load and boundary condition are inevitably subject to the impact of multiple uncertain factor so that the temperature field of structure also shows certain uncertainty.For complication system, even if only small uncertain input, final system response is likely to and can produce obvious disturbance.Stochastic Analysis Method based on probability theory has been achieved for part achievement in research in uncertain thermodynamic study, but the substantial amounts of sample information of needs of setting up of probabilistic model determines the probability density function of parameter in advance.Starting stage in numerical analysis, it is thus achieved that enough sample datas often spend relatively big or cost prohibitive, which limits the further genralrlization of Stochastic calculus model and method.And in fuzzy uncertainty is analyzed, although the numerical value of the concept of some things or parameter is difficult to determine, but a scope substantially can be determined according to experimental data or subjective experience.Consequently, it is possible to fuzzy model shows very strong convenience and economy in uncertainties model.Fuzzy theory in structure static analysis achieved with some achievements in research, but just at the early-stage in thermodynamics field.It addition, to join the application in non-probability problem of spectral analysis method based on a technology also or blank out.Therefore, how to apply and join a theory fuzzy heat transfer problem is carried out numerical solution, be a study hotspot of current sphere of learning, for making up the deficiency of existing Heat Transfer Numerical, there is important engineer applied and be worth.

Summary of the invention

The technical problem to be solved is: overcome the deficiency that prior art exists in thermal conduction under steady state solves, take into full account the fuzzy uncertain factor in heat transfer problem, a theory is joined based on approximation by polynomi-als thought and sparse grid, propose the numerical computation method of a kind of quickly effective predictive fuzzy temperature-responsive membership function, systematization can solve the structure temperature field prediction problem containing fuzzy uncertain variable, while ensureing computational accuracy, reduce further the calculating of tradition sampling approach and expend.

The present invention solves that the technical scheme that above-mentioned technical problem adopts is: a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving, comprise the following steps:

Step one: for thermal conduction under steady state, chooses cell type according to version and carries out discrete to heat transfer structure, obtain the finite element equation of this heat transfer model；Having multiple cell type in finite element analysis, have close ties with version, need particular problem to make a concrete analysis of, this is known about by those skilled in the art；

Step 2: take into full account the fuzzy uncertainty inputting parameter in heat transfer model, and introduce fuzzy variable it is characterized, the fuzzy finite element equation of heat conduction problem is set up according to the finite element equation in step one；

Step 3: choose Truncated set level, utilizes cut set computing, the fuzzy variable in step 2 is converted into interval variable, and then the fuzzy finite element equation in step 2 can be rewritten as a class interval finite element equation；

Step 4: utilize Legnedre polynomial that the silicon carbide response of interval Finite Element Method equation in step 3 is carried out approximate representation, obtain temperature-responsive approximate expression；

Step 5: join a theory according to sparse grid, expansion coefficient in temperature-responsive approximate expression in step 4 is solved, and utilize the excursion of the slickness computation interval temperature-responsive of function, and then the silicon carbide response that all Truncated set level are corresponding can be obtained；

Step 6: utilize fuzzy resolution theorem, responds the silicon carbide under all Truncated set level obtained in step 5 and recombinates, obtain the membership function of Fuzzy temperature response.

Wherein, in described step 3, choosing of Truncated set level is not changeless, determines quantity and the size of required Truncated set level according to the distribution pattern of fuzzy uncertain variable membership degree function.The distribution pattern of membership function includes the multiple distribution form such as Triangle-Profile, trapezoidal profile, can choose multiple Truncated set level it is analyzed in 0 to 1 scope.Due to the difference of problem, the quantity of Truncated set level and size are also different, for instance the quantity of linear problem Truncated set level just can be lacked a little, and the quantity of nonlinear problem Truncated set level will be more.

Wherein, utilizing Legnedre polynomial that silicon carbide response is carried out approximate representation in described step 4, polynomial exponent number is not changeless, requires to choose according to approximation accuracy.Blocking exponent number more high, approximation accuracy is more high.

Wherein, in described step 5, sparse grid joins choosing of a scheme is not changeless, considers to calculate and expends and the requirement of computational accuracy is chosen and joined a level.Joining a level more high, computational accuracy is more high, and it is more big to calculate consuming.

Above steps specifically includes procedure below:

Step one: for thermal conduction under steady state, chooses cell type according to version and carries out discrete to heat transfer structure, obtain the finite element equation of this heat transfer model:

KT=F

Wherein T is temperature vector, and K is conduction of heat matrix, and F is thermal force vector.

Step 2: take into full account the fuzzy uncertainty inputting parameter in heat transfer model, introduces n fuzzy variable α₁,α₂,...,α_nIt is characterized, and is designated as the form α=(α of vector_i)_n=(α₁,α₂,...,α_n), the fuzzy finite element equation of heat conduction problem can be set up thus according to the finite element equation in step one:

K (α) T (α)=F (α)

Step 3: choose Truncated set level λ in 0 to 1 scope, utilizes the cut set computing can by the fuzzy variable α in step 2_iIt is converted into interval variable

${\mathrm{\α}}_{i,\mathrm{\λ}}^I=\[{\underset{\‾}{\mathrm{\α}}}_{i,\mathrm{\λ}},{\stackrel{\‾}{\mathrm{\α}}}_{i,\mathrm{\λ}}\]={\mathrm{\α}}_{i,\mathrm{\λ}}^c+{\mathrm{\Δ\α}}_{i,\mathrm{\λ}}^I={\mathrm{\α}}_{i,\mathrm{\λ}}^c+{\mathrm{\Δ\α}}_{i,\mathrm{\λ}}{\mathrm{\δ}}_i^I,i=1,2,...,n$

Whereinα _i,λWithFor representing interval variableLower bound and the upper bound,WithIt is called interval midpoint and radius,For standard interval variable Represent interval variableThe fluctuation change at some place wherein.All interval variables under λ Truncated set level are designated as the form of vectorAnd then the fuzzy finite element equation in step 2 can be rewritten as a class interval finite element equation:

$K\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)T\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)=F\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)$

WhereinFor silicon carbide response vector.

Step 4: extraction step three silicon carbide response vectorIn certain concrete componentRepresent that certain node temperature responds, utilize Legnedre polynomial, it is possible to node temperature is respondedCarry out approximate representation, obtain temperature-responsive approximate expression；

$T\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)\≈T_N\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)=\underset{\left|i\right|\≤N}{\Σ}{T}_i{\mathrm{\Φ}}_i\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)$

WhereinFor Legendre's orthogonal polynomial substrate function, T_iFor corresponding expansion coefficient, i=(i₁,i₂,...,i_n) represent multidimensional index, and meet | i |=i₁+i₂+...+i_n, N polynomial blocks exponent number for this.

According to chaos polynomial theory, the number Available Variables number n of this polynomial expansion item and block exponent number N and be expressed as

Step 5: join a theory according to sparse grid, expansion coefficient in temperature-responsive approximate expression in step 4 is solved, and utilize the excursion of the slickness computation interval temperature-responsive of function, and then the silicon carbide response that all Truncated set level are corresponding can be obtained.First, selected entirety joins a little horizontal s, makes L=s+n, joins a set Θ according to the sparse grid of tensor product and Smolyak algorithm construction higher dimensional space and is:

$\mathrm{\Θ}=\underset{L-n+1\≤\left|i\right|\≤L}{\∪}({\mathrm{\Θ}}_1^{i_1}\×{\mathrm{\Θ}}_2^{i_2}\×...\×{\mathrm{\Θ}}_n^{i_n})$

Wherein i_jJ=1,2 ..., what n represented jth dimension space (correspond to jth interval variable) joins a level,Represent jth interval variableCorrespondence joins a little horizontal i_jAll set joining some composition, concrete joins a quantityWith join a positionIt is respectively as follows:

The joining a quantity M and can approximate representation be of whole higher dimensional space:

$M=\dim \left(\mathrm{\Θ}\right)=\dim \left(\underset{L-n+1\≤\left|i\right|\≤L}{\∪}\right({\mathrm{\Θ}}_1^{i_1}\×{\mathrm{\Θ}}_2^{i_2}\×...\×{\mathrm{\Θ}}_n^{i_n}\left)\right)\≈\frac{2^s}{s!}{n}^s$

A set Θ will be joined be designated asForm, be used for represent that all sparse grids are joined a little

Make the interval Finite Element Method equation set up in step 3 join a place at above-mentioned all sparse grids set up, be based further on the polynomial expression in step 4 and can obtain about expansion coefficient T_iAn equation group:

WhereinRepresentative polynomial basis functionJoining a littleThe value at place,Represent node temperature responseJoining a littleThe value at place.

Utilize least square method to solve above-mentioned equation group, obtain expansion coefficient T_iA class value, in its generation, is returned in the polynomial expression in step 4, utilizes functionSlickness, it is determined that it is about the extreme point of interval variable, and then approximate obtains temperature-responsiveLower boundAnd the upper bound

$\begin{array}{cc}\underset{\‾}{T}\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)\≈{\mathrm{minT}}_N\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)& \stackrel{\‾}{T}\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)\≈{\mathrm{maxT}}_N\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)\end{array}$

Silicon carbide response under this Truncated set level is expressed as range format have

Selected all Truncated set level are repeated aforesaid operations, and then the silicon carbide response that all Truncated set level are corresponding can be obtained.

Step 6: utilize fuzzy resolution theorem, responds the silicon carbide under all Truncated set level obtained in step 5 and recombinates, it is possible to obtain the membership function T (α) of fuzzy temperature-responsive under fuzzy parameter vector α impact:

$T\left(\mathrm{\α}\right)=\underset{j=1,...,m}{\∪}\left\{{\mathrm{\λ}}_j{T}^I\left({\mathrm{\α}}_{{\mathrm{\λ}}_j}^I\right)\right\}$

Wherein m is the quantity of selected Truncated set level.

Present invention advantage compared with prior art is in that:

(1) compared with analyzing method with traditional thermal conduction under steady state, proposed computation model and computational methods take into full account the fuzzy uncertainty of material properties in Practical Project, external load and boundary condition, and temperature field analysis is had prior directive significance by result of calculation.

(2) utilize the slickness of Legnedre polynomial function, can quickly determine its extreme point about interval variable and then the approximate bound obtaining temperature-responsive.

(3) based on joining a theory, it is possible to the calculation procedure making full use of original deterministic models makes further processing and amendment without to it, it is ensured that the portability of calculating.It addition, utilize Smolyak algorithm construction to join a set, it is effectively reduced and joins a quantity, reduce further calculating and expend.

(4) present invention is simple to operate, and it is convenient to implement, and is ensureing to improve on the basis of computational accuracy the computational efficiency of point collocation.

Accompanying drawing explanation

Fig. 1 is a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving flow chart of the present invention；

Fig. 2 is the FEM (finite element) model schematic diagram of the three-dimensional sandwich structure heat conduction problem of the present invention；

Fig. 3 is the membership function schematic diagram of observation station 6 place Fuzzy temperature response；

Fig. 4 is the membership function schematic diagram of observation station 7 place Fuzzy temperature response.

Detailed description of the invention

Below in conjunction with drawings and Examples, the present invention will be further described.

The present invention is applicable to the structure temperature field prediction of the thermal conduction under steady state containing fuzzy uncertain parameter.Embodiment of the present invention for the temperature field prediction of certain three-dimensional sandwich structure, illustrates described a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving.It addition, the Fuzzy temperature number of responses value calculating method of this three-dimensional sandwich structure can be generalized in other thermal conduction under steady state temperature field predictions containing fuzzy parameter.

A kind of based on joining a calculating process of theoretical Fuzzy temperature response membership function method for solving as shown in Figure 1, FEM (finite element) model based on structure, set up the finite element discretization equation of steady state heat transfer problem, take into full account the fuzzy uncertainty inputting parameter in heat transfer model, introduce fuzzy variable it is characterized, and based on cut set computing, it is further processed, rewriting fuzzy finite element equation is a class interval finite element equation, join the point methods Legendre polynomial expansion coefficient to temperature-responsive according to sparse grid to solve, obtain the interval excursion of temperature-responsive, and utilize fuzzy resolution theorem that it is recombinated, finally give the membership function of Fuzzy temperature response.Following several step can be divided into carry out:

Step one: set up the FEM (finite element) model of three-dimensional sandwich structure heat conduction problem, as shown in Figure 2: outer panels 1 and wainscot 2 all discrete be 640 quadrilateral units, web 3 is divided into 720 quadrilateral units, and outer surface 4 bears hot-fluid load q_s, inner boundary and surrounding generation surface heat exchanging, selected it is numbered four nodes of 5～8 observation station as this structure temperature field.Set up the finite element equation of this heat transfer model:

KT=F

Wherein T is temperature vector, and K is conduction of heat matrix, and F is thermal force vector.

Step 2: in this heat transfer model, upper and lower panel adopts the coefficient of heat conduction to be k₁Structural material, and the coefficient of heat conduction of web material is k₂, ambient temperature is T_e, surface film thermal conductance is designated as h.Due to the error of the restriction of material fabrication process and measurement, all mode input parameters all contain certain fuzzy uncertainty, and membership function meets angular distribution rule, i.e. k₁=(100,120,140) W/ (m DEG C), k₂=(174,204,234) W/ (m DEG C), q_s=(8,10,12) W/m², T_e=(15,20,25) DEG C, h=(9,10,11) W/ (m²·℃).All fuzzy variable vector symbol are collectively expressed as α=(α_i)₅=(k₁,k₂,q_s,T_e, h), the fuzzy finite element equation of heat conduction problem can be set up thus according to the finite element equation in step one:

K (α) T (α)=F (α)

Step 3: choose 11 Truncated set level λ in 0 to 1 scope_j=(j-1) × 0.1j=1 ..., 11, convenient in order to represent, unified to be designated as λ.Utilize the cut set computing can by the fuzzy variable α in step 2_iIt is converted into interval variable

${\mathrm{\α}}_{i,\mathrm{\λ}}^I=\[{\underset{\‾}{\mathrm{\α}}}_{i,\mathrm{\λ}},{\stackrel{\‾}{\mathrm{\α}}}_{i,\mathrm{\λ}}\]={\mathrm{\α}}_{i,\mathrm{\λ}}^c+{\mathrm{\Δ\α}}_{i,\mathrm{\λ}}^I={\mathrm{\α}}_{i,\mathrm{\λ}}^c+{\mathrm{\Δ\α}}_{i,\mathrm{\λ}}{\mathrm{\δ}}_i^I,i=1,2,...,5$

Whereinα _i,λWithFor representing interval variableLower bound and the upper bound,WithIt is called interval midpoint and radius,For standard interval variable Represent interval variableThe fluctuation change at some place wherein.All interval variables under λ Truncated set level are designated as the form of vectorAnd then the fuzzy finite element equation in step 2 can be rewritten as a class interval finite element equation:

$K\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)T\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)=F\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)$

WhereinFor silicon carbide response vector.

Step 4: extraction step three silicon carbide response vectorThe node temperature response of middle corresponding observation station, usesRepresent.Characteristic distributions according to interval variable, selects Legnedre polynomial as approximating function, blocks exponent number and be set as N=3, it is possible to node temperature is respondedCarry out approximate representation, obtain temperature-responsive approximate expression:

$T\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)\≈T_N\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)=\underset{\left|i\right|\≤N}{\Σ}{T}_i{\mathrm{\Φ}}_i\left({\mathrm{\α}}_{\mathrm{\λ}}^I\right)$

WhereinFor Legendre's orthogonal polynomial substrate function, T_iFor corresponding expansion coefficient, i=(i₁,i₂,...,i₅) represent multidimensional index, and meet | i |=i₁+i₂+...+i₅, now the number of polynomial expansion item is

Step 5: join a theory according to sparse grid, expansion coefficient in temperature-responsive approximate expression in step 4 is solved, and utilize the excursion of the slickness computation interval temperature-responsive of function, and then the silicon carbide response that all Truncated set level are corresponding can be obtained.Firstly, for Truncated set level λ₈=0.7, select and join a little horizontal s=2, make L=s+n=7, joining a set Θ according to the sparse grid of tensor product and Smolyak algorithm construction higher dimensional space is:

$\mathrm{\Θ}=\underset{3\≤\left|i\right|\≤7}{\∪}({\mathrm{\Θ}}_1^{i_1}\×{\mathrm{\Θ}}_2^{i_2}\×...\×{\mathrm{\Θ}}_5^{i_5})$

Wherein i_jJ=1,2 ..., 5 represent jth dimension space (correspond to jth interval variable) join a level,Represent jth interval variableCorrespondence joins a little horizontal i_jAll set joining some composition, concrete joins a quantityWith join a positionIt is respectively as follows:

A quantity of joining of whole higher dimensional space is M=61, will join a set Θ and be designated asForm, be used for represent that all sparse grids are joined a little

Make the interval Finite Element Method equation set up in step 3 join a place at above-mentioned all sparse grids set up, be based further on the polynomial expression in step 4 and can obtain about expansion coefficient T_iAn equation group:

WhereinRepresentative polynomial basis functionJoining a littleThe value at place,Represent node temperature responseJoining a littleThe value at place.

Utilize least square method to solve above-mentioned equation group, obtain expansion coefficient T_iA class value, in its generation, is returned in the polynomial expression in step 4, utilizes the slickness of function, it is determined that it is about the extreme point of interval variable, and then approximate obtain Truncated set level λ₈=0.7 time temperature-responsiveLower boundAnd the upper bound

$\begin{array}{cc}\underset{\‾}{T}\left({\mathrm{\α}}_{0.7}^I\right)\≈{\mathrm{minT}}_N\left({\mathrm{\α}}_{0.7}^I\right)& \stackrel{\‾}{T}\left({\mathrm{\α}}_{0.7}^I\right)\≈{\mathrm{maxT}}_N\left({\mathrm{\α}}_{0.7}^I\right)\end{array}$

Silicon carbide response under this Truncated set level is expressed as range format have

Same, for Truncated set level λ₆=0.5, set and join a little horizontal s=3, repeat aforesaid operations, obtain Truncated set level λ₆The lower bound of=0.5 time temperature-responsiveAnd the upper boundIt is designated as range format to have

Other selected Truncated set level are repeated aforesaid operations, and then the silicon carbide response that all Truncated set level are corresponding can be obtained.

Above-mentioned λ₈=0.7 and λ₆Shown in the result of calculation such as table 2 and table 3 of=0.5 two lower four observation station place temperature-responsive of Truncated set level.It is 10 with sample number⁶Traditional Monte Carlo sampling approach contrast it can be seen that the calculating error of the inventive method is less than 1%, computational accuracy can fully meet engineering demand.It addition, from sample size, the calculating of the inventive method expends and is far smaller than monte carlo method, is more suitable for actual complex engineering problem.

Table 2 Truncated set level λ₈=0.7 time observation station place silicon carbide response

Table 3 Truncated set level λ₆=0.5 time observation station place silicon carbide response

Step 6: utilize fuzzy resolution theorem, responds the silicon carbide under all Truncated set level obtained in step 5 and recombinates, it is possible to obtain the membership function T (α) of fuzzy temperature-responsive under fuzzy parameter vector α impact

$T\left(\mathrm{\α}\right)=\underset{j=1,...,11}{\∪}\left\{{\mathrm{\λ}}_j{T}^I\left({\mathrm{\α}}_{{\mathrm{\λ}}_j}^I\right)\right\}$

For observation station 6 and 7, the membership function of Fuzzy temperature response is as shown in Figure 3 and Figure 4, abscissa represents that temperature value, vertical coordinate represent the degree of membership size that temperature value is corresponding, and solid line and dotted line represent Monte Carlo sampling approach and the calculated result of the inventive method respectively.It can be seen that the membership function of temperature-responsive is the same with mode input parameter, still it is similar to and meets angular distribution rule.It addition, the calculated membership function curve of the inventive method is fine with the reference value degree of agreement that the sampling of tradition Monte Carlo obtains, result of calculation is genuine and believable.Can solving the structure thermal conduction under steady state containing Indistinct Input parameter by the inventive method, computational accuracy is high, calculates and expends less, and this function is that general business software institute is irrealizable.

The above-described presently preferred embodiments of the present invention that is only, the present invention is not limited solely to above-described embodiment, and all local made within the spirit and principles in the present invention are changed, equivalent replacement, improvement etc. should be included within protection scope of the present invention.

Claims (4)

1. one kind responds a membership function method for solving based on joining a theoretical Fuzzy temperature, it is characterised in that comprise the following steps:

Step one: for thermal conduction under steady state, chooses cell type according to version and carries out discrete to heat transfer structure, obtain the finite element equation of this heat transfer model；

Step 2: take into full account the fuzzy uncertainty inputting parameter in heat transfer model, and introduce fuzzy variable it is characterized, the fuzzy finite element equation of heat conduction problem is set up according to the finite element equation in step one；

Step 3: choose Truncated set level, utilizes cut set computing, the fuzzy variable in the fuzzy finite element equation in step 2 is converted into interval variable, and then the fuzzy finite element equation in step 2 is rewritten as a class interval finite element equation；

Step 4: utilize Legnedre polynomial that the silicon carbide in interval Finite Element Method equation in step 3 is responded and carry out approximate representation, obtain temperature-responsive approximate expression；

Step 5: join a theory according to sparse grid, expansion coefficient in temperature-responsive approximate expression in step 4 is solved, and utilize the excursion of the slickness computation interval temperature-responsive of function, thus obtain the silicon carbide response that all Truncated set level are corresponding；

Step 6: utilize fuzzy resolution theorem, responds the silicon carbide under all Truncated set level obtained in step 5 and recombinates, obtain the membership function of Fuzzy temperature response.

2. according to claim 1 a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving, it is characterized in that: in described step 3, choosing of Truncated set level is not changeless, determine quantity and the size of required Truncated set level according to the distribution pattern of fuzzy uncertain variable membership degree function.

3. according to claim 1 a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving, it is characterized in that: described step 4 utilizes Legnedre polynomial silicon carbide response is carried out approximate representation, the polynomial exponent number that blocks not is changeless, requires to choose according to approximation accuracy.

4. according to claim 1 a kind of based on joining a theoretical Fuzzy temperature response membership function method for solving, it is characterized in that: in described step 5, sparse grid joins choosing of a scheme is not changeless, consider to calculate and expend and the requirement of computational accuracy is chosen and joined a level.