1. bookAHEAD OF PRINT
Detalles de la revista
License
Formato
Revista
eISSN
2444-8656
Primera edición
01 Jan 2016
Calendario de la edición
2 veces al año
Idiomas
Inglés
access type Acceso abierto

Higher Mathematics Teaching Curriculum Model Based on Lagrangian Mathematical Model

Publicado en línea: 15 Jul 2022
Volumen & Edición: AHEAD OF PRINT
Páginas: -
Recibido: 11 Jan 2022
Aceptado: 30 Mar 2022
Detalles de la revista
License
Formato
Revista
eISSN
2444-8656
Primera edición
01 Jan 2016
Calendario de la edición
2 veces al año
Idiomas
Inglés
Abstract

This paper refers to the Lagrangian mathematical model and the Rackwitz-Fiessler transform model in modeling advanced mathematics applications. In this paper, the Lagrangian interpolation method and Rackwitz-Fiessler transform are used to calculate the correlation coefficient efficiently. The purpose of this algorithm is to speed up the process of probabilistic modeling. This paper uses integral probability transformation to characterize the output's uncertainty by the production's joint distribution function. The purpose of this process is to improve the reliability of the calculation. Through the algorithm analysis, it is found that the method is reasonable and efficient.

Keywords

MSC 2010

Introduction

Reliability theory is developing and improving rapidly. It has been paid more attention to geotechnical engineering and has been widely used [1]. Scholars go through the linear transformation method and the Rosenblatt transformation. Scholars have done a lot of research to consider the correlation between random variables. Some scholars have introduced the correlated variable independence method based on Cholesky decomposition into the stochastic response surface method. But the technique only works in the case of correlated standard variables. The most widely used is the Rackwitz-Fiessler transform. The transformation takes the marginal cumulative distribution function and covariance matrix of random variables as basic information. This paper adopts the Rackwitz-Fiessler probability transformation and merges this transformation with the Lagrange multiplier method. At the same time, this paper proposes a primary reliability method based on Rackwitz-Fiessler probability transformation and the Lagrange multiplier method [2]. In this paper, the rational polynomial method is introduced into the derivation of the inverse function expression of the distribution function. Finally, we demonstrate the effectiveness of the method proposed in this paper through a numerical example analysis.

Rackwitz-Fiessler transform

We use the principle of equal probability transformation to introduce the standard normal random variable Y = (y1, y2, ⋯, yn), as follows [3]: {Φ(yi)=FI(xi)yi=Φ1[Fi(xi)](i=1,,n) \left\{ {\matrix{ {\Phi \left( {{y_i}} \right) = {F_I}\left( {{x_i}} \right)} \hfill \cr {{y_i} = {\Phi ^{ - 1}}\left[ {{F_i}\left( {{x_i}} \right)} \right]} \hfill \cr } } \right.\left( {i = 1, \cdots ,\,n} \right)

Φ( ) and Φ−1 ( ) are the standard normal cumulative distribution function and the inverse of the cumulative distribution function. We find the following rules based on the Rackwitz-Fiessler distribution theory [4]. When Y is a joint average random vector, we can derive the joint probability density function of X by using the implicit function derivation rule: fx(x)=fx1(x1)fx2(x2)fxn(xn)×φn(y,ρ0)φ(y1)φ(y1)φ(yn) {f_x}\left( x \right) = {f_{x1}}\left( {{x_1}} \right){f_{x2}}\left( {{x_2}} \right) \cdots {f_{xn}}\left( {{x_n}} \right) \times {{{\varphi _n}\left( {y,{\rho _0}} \right)} \over {\varphi \left( {{y_1}} \right)\varphi \left( {{y_1}} \right) \cdots \varphi \left( {{y_n}} \right)}}

φ ( ) is the probability density function of the standard normal distribution. φn (y, ρ0) is a vector with zero equivalents, a matrix of unit standard deviations, and a correlation coefficient [5]. The matrix is the n dimensional standard normal distribution of ρ0 · ρ0 = [ρ0,ij] is the correlation coefficient matrix of the joint average random vector Y. The correlation coefficient ρij of the element ρ0,ij of ρ0 and X has the following relationship: ρij=(xiμiσi)(xjμjσj)φ2(yi,yj,ρ0,ij)dyidyj {\rho _{ij}} = \int_{ - \infty }^\infty {\int_{ - \infty }^\infty {\left( {{{{x_i} - {\mu _i}} \over {{\sigma _i}}}} \right)\left( {{{{x_j} - {\mu _j}} \over {{\sigma _j}}}} \right){\varphi _2}\left( {{y_i},{y_j},{\rho _{0,ij}}} \right)d{y_i}d{y_j}} }

φ2 (yi, yj, ρ0,ij is a two-dimensional standard normal distribution with zero equivalence vector, unit standard deviation matrix, and correlation coefficient matrix ρ0,ij. Analytic relations for ρij and ρ0,ij are difficult to give [6]. Through a large number of numerical studies, we have provided an empirical formula for calculating the equivalent correlation coefficient between two types of probability distributions with a total of 10 probability distributions: ρ0,ij=Fρij {\rho _{0,ij}} = F{\rho _{ij}}

The coefficient F ≥ 1 in the formula. It is the correlation coefficient ρij and the marginal distribution FXi (xi). We can calculate it by the two-dimensional Gaussian-Hermite integral [7]. Then we use the nonlinear root-finding method to find the correlation coefficient ρ0,ij. The correlation coefficient matrix ρ0 = [ρ0,ij] of which we obtain the standard average random vector Y is symmetric. It can perform Cholesky decomposition. ρ0=L0L0T {\rho _0} = {L_0}L_0^T

In the formula L0 is the lower triangular decomposition matrix obtained by the Cholesky decomposition of the correlation coefficient matrix ρ0. So we can get the positive transformation form of Rackwitz-Fiessler as follows: u=L01Y=L01{y1y2yn}=L01{Φ1[F1(x1)]Φ1[F2(x2)]Φ1[Fn(xn)]} u = L_0^{ - 1}Y = L_0^{ - 1}\left\{ {\matrix{ {{y_1}} \cr {{y_2}} \cr \vdots \cr {{y_n}} \cr } } \right\} = L_0^{ - 1}\left\{ {\matrix{ {{\Phi ^{ - 1}}\left[ {{F_1}\left( {{x_1}} \right)} \right]} \cr {{\Phi ^{ - 1}}\left[ {{F_2}\left( {{x_2}} \right)} \right]} \cr \vdots \cr {{\Phi ^{ - 1}}\left[ {{F_n}\left( {{x_n}} \right)} \right]} \cr } } \right\}

Lagrange Multiplier Method for Reliability Analysis
The geometric meaning of reliability indicators

We can obtain the following theorem according to the first-order second-moment theory of reliability analysis [89]. We can prove that this distance is an essential indicator in reliability analysis: reliability indicator β. We therefore assume that the limit state equation with n standard variables is expressed as follows: z=g(x1,x2,,xn) z = g\left( {{x_1},{x_2}, \cdots ,{x_n}} \right)

After normalizing the normal variables, we have the following relationship: yi=(xiμxi)/σxi {y_i} = \left( {{x_i} - {\mu _{xi}}} \right)/{\sigma _{xi}}

μxi, σxi are the mean and standard deviation of the variable, respectively [10]. Then the mathematical model of the expected average space calculation of the reliability index can be expressed as follows: {β=min(i=1nyi2)1/2z=g(x1,x2,,xn)=0 \left\{ {\matrix{ {\beta = \min {{\left( {\sum\limits_{i = 1}^n {y_i^2} } \right)}^{1/2}}} \hfill \cr {z = g\left( {{x_1},{x_2}, \cdots ,{x_n}} \right) = 0} \hfill \cr } } \right.

If each random variable follows a general distribution, we can perform Gaussian transformation.

Lagrange Multiplier Method for Reliability Analysis

Suppose the inverse matrix of the lower triangular decomposition matrix L0 obtained by the Cholesky decomposition of the formula is expressed as follows: L01={a11a12a22a1na2nann} L_0^{ - 1} = \left\{ {\matrix{ {{a_{11}}} & {} & {} & {} \cr {{a_{12}}} & {{a_{22}}} & {} & {} \cr {} & \vdots & {} & {} \cr {{a_{1n}}} & {{a_{2n}}} & \cdots & {{a_{nn}}} \cr } } \right\}

The formula (6) can be rewritten as: {u1u2un}=L01{y1y2yn}={a11y1i=12ai2yii=1nainyi} \left\{ {\matrix{ {{u_1}} \cr {{u_2}} \cr \vdots \cr {{u_n}} \cr } } \right\} = L_0^{ - 1}\left\{ {\matrix{ {{y_1}} \cr {{y_2}} \cr \vdots \cr {{y_n}} \cr } } \right\} = \left\{ {\matrix{ {{a_{11}}{y_1}} \cr {\sum\limits_{i = 1}^2 {{a_{i2}}{y_i}} } \cr \vdots \cr {\sum\limits_{i = 1}^n {{a_{in}}{y_i}} } \cr } } \right\}

Therefore, the standard regular space mathematical model of structural reliability index with correlation variables can be obtained from equation (9) as follows: {β=min(i=1nui2)1/2=min[(a11y1)2+(i=1nainyi)2++(i=1nainyi)2]1/2z=g(x1,x2,,xn)=0 \left\{ {\matrix{ {\beta = \min {{\left( {\sum\limits_{i = 1}^n {u_i^2} } \right)}^{1/2}} = \min {{\left[ {{{\left( {{a_{11}}{y_1}} \right)}^2} + {{\left( {\sum\limits_{i = 1}^n {{a_{in}}{y_i}} } \right)}^2} + \cdots + {{\left( {\sum\limits_{i = 1}^n {{a_{in}}{y_i}} } \right)}^2}} \right]}^{1/2}}} \hfill \cr {z = g\left( {{x_1},{x_2}, \cdots ,{x_n}} \right) = 0} \hfill \cr } } \right.

Ψ=(x1,x2,,xn)=β2=i=1nui2 \Psi = \left( {{x_1},{x_2}, \cdots ,{x_n}} \right) = {\beta ^2} = \sum\limits_{i = 1}^n {u_i^2} , then the formula (12) essentially becomes the problem of solving the extremum value of the function Ψ = (x1, x2,⋯, xn) under the condition z = g(x1, x2,⋯, xn) = 0. Therefore, according to the Lagrange multiplier method, the operation can be constructed as follows: G(x1,x2,,xn)=i=1nui2+λg(x1,x2,,xn) G\left( {{x_1},{x_2},\cdots,{x_n}} \right) = \sum\limits_{i = 1}^n {u_i^2} + \lambda g\left( {{x_1},{x_2},\cdots,{x_n}} \right)

In the formula, ui is calculated according to procedure (11). λ is a constant. We take the first-order partial derivative of equation (13) with respect to x1, x2,⋯, xn. We make it zero [11]. Then we combine it with the equation g(x1, x2,⋯, xn) = 0 to get the following formula: {Ψxn(xn)+λgxn(x1,x2,,xn)=0g(x1,x2,,xn)=0 \left\{ {\matrix{ {{\Psi _{xn}}\left( {{x_n}} \right) + \lambda {g_{xn}}\left( {{x_1},\,{x_2},\,\cdots,\,{x_n}} \right) = 0} \hfill \cr {g\left( {{x_1},\,{x_2},\,\cdots,\,{x_n}} \right) = 0} \hfill \cr } } \right.

x1, x2,⋯, xn and λ can be solved by equation system (14), then x1, x2,⋯, xn is the coordinate (checkpoint coordinate) when the structural reliability index is the minimum value [12]. We solve the reliability index β of the structure from (i=1nui2)1/2 {\left( {\sum\limits_{i = 1}^n {u_i^2} } \right)^{1/2}} .

Rational Polynomial Method for Derivative Solving

Ψ / ∂xi in normal and lognormal distributions can be found directly. Since expressions for other distribution factor functions Φ−1 [F(x)] are difficult to give, it is difficult to solve the derivatives instantly. Because it is difficult for us to solve the derivative of the implicit or complex explicit function directly, we can use the rational polynomial method. The reasonable polynomial method is divided into the 3-point formulated, and the 5-point prepared [1314]. Since the 5-point formula has the best accuracy, the calculation example in this paper is solved by the 5-point formula method.

The value of the derivative of the function z = g(x1, x2,⋯, xn) at x* needs to be solved further. First, assume that the 5 x(i) (i = 1, 2, ⋯, 5) are known values. Where x(i) ∈ [x* − 3σx, x* + 3σx]. At this point we get 5 function values: z11=g(x1(1),x2,,xn),z12=g(x1(2),x2,,xn),,z15=g(x1(5),x2,,xn) {z_{11}} = g\left( {x_1^{\left( 1 \right)},{x_2}, \cdots ,{x_n}} \right),\,{z_{12}} = g\left( {x_1^{(2)},{x_2}, \cdots ,{x_n}} \right), \cdots ,{z_{15}} = g\left( {x{\kern 1pt} _1^{\left( 5 \right)},{x_2}, \cdots ,{x_n}} \right)

The function of the function concerning the variable x can be approximated by the rational polynomial method as a function continued fraction form: z=g(x1(1),x2,,xn)a1+xx(1)a2+xx(2)a3++xx(4)a5 z = g\left( {x_1^{\left( 1 \right)},{x_2}, \cdots ,{x_n}} \right) \approx {a_1} + {{x - {x^{\left( 1 \right)}}} \over {{a_2} + {{x - {x^{\left( 2 \right)}}} \over {{a_3} + \cdots + {{x - {x^{\left( 4 \right)}}} \over {{a_5}}}}}}}

(16) can also be written as: z=g(x1,x2,,xn)ϕ1(x) z = g\left( {{x_1},{x_2}, \cdots ,{x_n}} \right) \approx {\phi _1}\left( x \right)

In: ϕi(x)=ai+(xx(i))/ϕi+1(x)(i=1,2,,4) \matrix{ {{\phi _i}\left( x \right) = {a_i} + \left( {x - {x^{\left( i \right)}}} \right)/{\phi _{i + 1}}\left( x \right)} \hfill \cr {\left( {i = 1,2, \cdots ,4} \right)} \hfill \cr } ϕ5(x)=a5 {\phi _5}\left( x \right) = {a_5}

The coefficient a1, a2,⋯, a5 in the above formula can be obtained from Table 1.

Rational polynomial coefficients for variable x

a1 a2 a3 a4 a5
a1 = z11 / / / /
a21 = z12 a2=x(2)x(1)a21a1 {a_2} = {{{x^{\left( 2 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{21}} - {a_1}}} / / /
a31 = z13 a32=x(3)x(1)a31a1 {a_{32}} = {{{x^{\left( 3 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{31}} - {a_1}}} a3=x(3)x(2)a31a2 {a_3} = {{{x^{\left( 3 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{31}} - {a_2}}} / /
a41 = z14 a42=x(4)x(1)a41a1 {a_{42}} = {{{x^{\left( 4 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{41}} - {a_1}}} a43=x(4)x(2)a42a2 {a_{43}} = {{{x^{\left( 4 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{42}} - {a_2}}} a4=x(4)x(3)a43a3 {a_4} = {{{x^{\left( 4 \right)}} - {x^{\left( 3 \right)}}} \over {{a_{43}} - {a_3}}} /
a51 = z15 a52=x(5)x(1)a51a1 {a_{52}} = {{{x^{\left( 5 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{51}} - {a_1}}} a53=x(5)x(2)a52a2 {a_{53}} = {{{x^{\left( 5 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{52}} - {a_2}}} a54=x(5)x(3)a53a3 {a_{54}} = {{{x^{\left( 5 \right)}} - {x^{\left( 3 \right)}}} \over {{a_{53}} - {a_3}}} a5=x(5)x(4)a54a4 {a_5} = {{{x^{\left( 5 \right)}} - {x^{\left( 4 \right)}}} \over {{a_{54}} - {a_4}}}

Then from equations (16) to (19), we can get: zx1=ϕ1(x1)=ϕ2(x1)(x1x1(1))ϕ2(x1)[ϕ2(x1)]2 {{\partial z} \over {\partial {x_1}}} = {\phi _1}\left( {{x_1}} \right) = {{{\phi _2}\left( {{x_1}} \right) - \left( {{x_1} - x_1^{\left( 1 \right)}} \right){\phi _2}\left( {{x_1}} \right)} \over {{{\left[ {{\phi _2}\left( {{x_1}} \right)} \right]}^2}}} ϕi(x1)=ϕi+1(x1)(xx1(i))ϕi+1(x1)[ϕi+1(x1)]2(i=1,2,,4) \matrix{ {\phi _i^\prime\left( {{x_1}} \right) = {{{\phi _{i + 1}}\left( {{x_1}} \right) - \left( {x - x_1^{\left( i \right)}} \right)\phi _{i + 1}^\prime\left( {{x_1}} \right)} \over {{{\left[ {{\phi _{i + 1}}\left( {{x_1}} \right)} \right]}^2}}}} \hfill \cr {\left( {i = 1,2, \cdots ,4} \right)} \hfill \cr } Where ϕi(x1) \phi _i^\prime\left( {{x_1}} \right) is the first derivative of the function ϕi(x1) with respect to the independent variable x1. ϕ5(x1)=0 \phi _5^\prime\left( {{x_1}} \right) = 0 can be obtained from equation (19). Then ϕ4(x1)=1/a5 \phi _4^\prime\left( {{x_1}} \right) = 1/{a_5} can be obtained from ϕ5(x1)=0 \phi _5^\prime\left( {{x_1}} \right) = 0 and equation (21). And so on to get the value of ϕ3(x1) \phi _3^\prime\left( {{x_1}} \right) , ϕ2(x1) \phi _2^\prime\left( {{x_1}} \right) .

Example analysis

The limit state equation g(x)=x122x2,x1 g\left( x \right) = x_1^2 - 2{x_2},{x_1} follows a lognormal distribution. Its mean and coefficient of variation are μx1 = 10, δx1 = 0.2, x2 respectively. It follows an extreme value type I distribution [1516]. Its mean and coefficient of variation are μx2 = 20, δx2 = 0.25, x1 respectively. The correlation coefficient ρ12 = 0.5 with x12. The empirical coefficient between the correlation coefficient ρ0,12 and ρ12 after the normalization of the quantity is expressed as follows F=1.029+0.001ρ12+0.014δx2+0.004ρ122+0.233δx220.197ρ12δx2 F = 1.029 + 0.001{\rho _{12}} + 0.014{\delta _{x2}} + 0.004\rho _{12}^2 + 0.233\delta _{x2}^2 - 0.197{\rho _{12}}{\delta _{x2}}

We can get F = 1.0239 by substituting ρ12 and 2 δx2 into equation (22). So we can get ρ0,12 = F ρ12 = 0.512. The lower triangular decomposition matrix L0 and its inverse matrix L-10 obtained by the Cholesky decomposition of the correlation coefficient matrix ρ0 are respectively expressed as follows

L0=[1.0000.5120.859] {L_0} = \left[ {\matrix{ {1.00} & 0 \cr {0.512} & {0.859} \cr } } \right]

L01=[1.0000.59601.1641] L_0^{ - 1} = \left[ {\matrix{ {1.00} & 0 \cr { - 0.5960} & {1.1641} \cr } } \right]

We transform x1 and x2 into standard standardized variables with equal probability: y1=[lnx1ln(μx1/1+δx12)]/ln(1+δx12) {y_1} = \left[ {\ln \,{x_1} - \ln \left( {{\mu _{x1}}/\sqrt {1 + \delta _{x1}^2} } \right)} \right]/\sqrt {\ln \left( {1 + \delta _{x1}^2} \right)} y2=Φ1{exp[exp(α(x2v))]} {y_2} = {\Phi ^{ - 1}}\left\{ {\exp \left[ { - \exp \left( { - \alpha \left( {{x_2} - v} \right)} \right)} \right]} \right\}

In the formula, the distribution parameter α=π/(6σx2)=0.2565 \alpha = \pi /\left( {\sqrt 6 {\sigma _{x2}}} \right) = 0.2565 , v = μx2 − 0.5770 / a = 17.7498 can be obtained from a procedure (11): {u1u2}=L01{y1y2}={y10.5960y1+1.1641y2} \left\{ {\matrix{ {{u_1}} \cr {{u_2}} \cr } } \right\} = L_0^{ - 1}\left\{ {\matrix{ {{y_1}} \cr {{y_2}} \cr } } \right\} = \left\{ {\matrix{ {{y_1}} \hfill \cr { - 0.5960{y_1} + 1.1641{y_2}} \hfill \cr } } \right\}

Ψx1 (x1, x2) and Ψx2 (x1, x2) are solved using the rational polynomial 5-point method. The general method solves the structure-function function partial derivatives [17]. The final calculation results are shown in Table 2.

Comparison of calculation results between this method and other methods

Calculation method Checkpoint Function function Reliable indicator
x1 x2 g(x) β
Nataf transform 6.4348 20.7033 −3.6335E-05 2.6822
Linear Rackwitz-Fiessler Transform 6.4366 20.7148 −2.0729E-06 2.6822
Extended Algorithm
Normal distribution transformation 6.4622 20.8801 −1.3240E-06 2.6604
The method of this paper 6.4314 20.6815 1.2120E-07 2.6865

It can be seen from Table 2 that:

The verification points and reliability indexes obtained by the method in this paper are consistent with the calculation results of other methods.

The absolute value of the function obtained by the method in this paper and the absolute value of the function calculated by other methods are relatively small. This shows that this paper's calculation method has higher accuracy than other methods.

Conclusion

The Rackwitz-Fiessler transform and the Lagrange multiplier method are straightforward to calculate the structural reliability. Its calculation result is reasonable. Engineers may not program after obtaining the correlation coefficient ρ0,ij. In this paper, the use of spreadsheets can be achieved using Solver. Rackwitz-Fiessler transforms, and the Lagrange multiplier method does not need to calculate the equivalent average mean and standard deviation of variables when calculating structural reliability. There is no need for trial calculation and iterative calculation in the computation of this model. Engineering examples show the effectiveness of the method and the accuracy of the results. Therefore, the calculation method proposed in this paper has specific engineering practical significance. This model is a more straightforward calculation method worth promoting.

Comparison of calculation results between this method and other methods

Calculation method Checkpoint Function function Reliable indicator
x1 x2 g(x) β
Nataf transform 6.4348 20.7033 −3.6335E-05 2.6822
Linear Rackwitz-Fiessler Transform 6.4366 20.7148 −2.0729E-06 2.6822
Extended Algorithm
Normal distribution transformation 6.4622 20.8801 −1.3240E-06 2.6604
The method of this paper 6.4314 20.6815 1.2120E-07 2.6865

Rational polynomial coefficients for variable x

a1 a2 a3 a4 a5
a1 = z11 / / / /
a21 = z12 a2=x(2)x(1)a21a1 {a_2} = {{{x^{\left( 2 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{21}} - {a_1}}} / / /
a31 = z13 a32=x(3)x(1)a31a1 {a_{32}} = {{{x^{\left( 3 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{31}} - {a_1}}} a3=x(3)x(2)a31a2 {a_3} = {{{x^{\left( 3 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{31}} - {a_2}}} / /
a41 = z14 a42=x(4)x(1)a41a1 {a_{42}} = {{{x^{\left( 4 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{41}} - {a_1}}} a43=x(4)x(2)a42a2 {a_{43}} = {{{x^{\left( 4 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{42}} - {a_2}}} a4=x(4)x(3)a43a3 {a_4} = {{{x^{\left( 4 \right)}} - {x^{\left( 3 \right)}}} \over {{a_{43}} - {a_3}}} /
a51 = z15 a52=x(5)x(1)a51a1 {a_{52}} = {{{x^{\left( 5 \right)}} - {x^{\left( 1 \right)}}} \over {{a_{51}} - {a_1}}} a53=x(5)x(2)a52a2 {a_{53}} = {{{x^{\left( 5 \right)}} - {x^{\left( 2 \right)}}} \over {{a_{52}} - {a_2}}} a54=x(5)x(3)a53a3 {a_{54}} = {{{x^{\left( 5 \right)}} - {x^{\left( 3 \right)}}} \over {{a_{53}} - {a_3}}} a5=x(5)x(4)a54a4 {a_5} = {{{x^{\left( 5 \right)}} - {x^{\left( 4 \right)}}} \over {{a_{54}} - {a_4}}}

Zwick, D., & Balachandar, S. A scalable Euler–Lagrange approach for multiphase flow simulation on spectral elements. The International Journal of High Performance Computing Applications., 2020; 34(3): 316–339 ZwickD. BalachandarS. A scalable Euler–Lagrange approach for multiphase flow simulation on spectral elements The International Journal of High Performance Computing Applications 2020 34 3 316 339 10.1177/1094342019867756 Search in Google Scholar

Sacco, R., Mauri, A. G., & Guidoboni, G. A Stabilized Dual Mixed Hybrid Finite Element Method with Lagrange Multipliers for Three-Dimensional Elliptic Problems with Internal Interfaces. Journal of Scientific Computing., 2020; 82(3): 1–31 SaccoR. MauriA. G. GuidoboniG. A Stabilized Dual Mixed Hybrid Finite Element Method with Lagrange Multipliers for Three-Dimensional Elliptic Problems with Internal Interfaces Journal of Scientific Computing 2020 82 3 1 31 10.1007/s10915-020-01163-7 Search in Google Scholar

Eskitaşçıoğlu, E. İ., Aktaş, M. B., & Baskonus, H. M. New complex and hyperbolic forms for Ablowitz–Kaup–Newell–Segur wave equation with fourth order. Applied Mathematics and Nonlinear Sciences., 2019; 4(1): 93–100 EskitaşçıoğluE. İ. AktaşM. B. BaskonusH. M. New complex and hyperbolic forms for Ablowitz–Kaup–Newell–Segur wave equation with fourth order Applied Mathematics and Nonlinear Sciences 2019 4 1 93 100 10.2478/AMNS.2019.1.00010 Search in Google Scholar

Durur, H., Kurt, A., & Tasbozan, O. New travelling wave solutions for KdV6 equation using sub equation method. Applied Mathematics and Nonlinear Sciences., 2020; 5(1): 455–460 DururH. KurtA. TasbozanO. New travelling wave solutions for KdV6 equation using sub equation method Applied Mathematics and Nonlinear Sciences 2020 5 1 455 460 10.2478/amns.2020.1.00043 Search in Google Scholar

Vadlamani, S. K., Xiao, T. P., & Yablonovitch, E. Physics successfully implements Lagrange multiplier optimization. Proceedings of the National Academy of Sciences., 2020; 117(43): 26639–26650 VadlamaniS. K. XiaoT. P. YablonovitchE. Physics successfully implements Lagrange multiplier optimization Proceedings of the National Academy of Sciences 2020 117 43 26639 26650 10.1073/pnas.2015192117760441633046659 Search in Google Scholar

Leitao, A., Margotti, F., & Svaiter, B. F. Range-relaxed criteria for choosing the Lagrange multipliers in the Levenberg–Marquardt method. IMA Journal of Numerical Analysis., 2021; 41(4): 2962–2989 LeitaoA. MargottiF. SvaiterB. F. Range-relaxed criteria for choosing the Lagrange multipliers in the Levenberg–Marquardt method IMA Journal of Numerical Analysis 2021 41 4 2962 2989 10.1093/imanum/draa050 Search in Google Scholar

Basu Roy, S., Bhasin, S., & Kar, I. N. Composite adaptive control of uncertain Euler-Lagrange systems with parameter convergence without PE condition. Asian Journal of Control., 2020; 22(1): 1–10 Basu RoyS. BhasinS. KarI. N. Composite adaptive control of uncertain Euler-Lagrange systems with parameter convergence without PE condition Asian Journal of Control 2020 22 1 1 10 10.1002/asjc.1877 Search in Google Scholar

Boiger, R., Leitao, A., & Svaiter, B. F. Range-relaxed criteria for choosing the Lagrange multipliers in nonstationary iterated Tikhonov method. IMA Journal of Numerical Analysis., 2020; 40(1): 606–627 BoigerR. LeitaoA. SvaiterB. F. Range-relaxed criteria for choosing the Lagrange multipliers in nonstationary iterated Tikhonov method IMA Journal of Numerical Analysis 2020 40 1 606 627 10.1093/imanum/dry066 Search in Google Scholar

May, J., Alobaid, F., Stroh, A., Daikeler, A., Ströhle, J., & Epple, B. Euler-Lagrange Model for the Simulation of Carbonate Looping Process. Chemie Ingenieur Technik., 2020; 92(5): 648–658 MayJ. AlobaidF. StrohA. DaikelerA. StröhleJ. EppleB. Euler-Lagrange Model for the Simulation of Carbonate Looping Process Chemie Ingenieur Technik 2020 92 5 648 658 10.1002/cite.201900159 Search in Google Scholar

Wei, R., Cao, J., & Huang, C. Lagrange exponential stability of quaternion-valued memristive neural networks with time delays. Mathematical Methods in the Applied Sciences., 2020; 43(12): 7269–7291 WeiR. CaoJ. HuangC. Lagrange exponential stability of quaternion-valued memristive neural networks with time delays Mathematical Methods in the Applied Sciences 2020 43 12 7269 7291 10.1002/mma.6463 Search in Google Scholar

Kuchta, M., Laurino, F., Mardal, K. A., & Zunino, P. Analysis and approximation of mixed-dimensional PDEs on 3D-1D domains coupled with Lagrange multipliers. SIAM Journal on Numerical Analysis., 2021; 59(1): 558–582 KuchtaM. LaurinoF. MardalK. A. ZuninoP. Analysis and approximation of mixed-dimensional PDEs on 3D-1D domains coupled with Lagrange multipliers SIAM Journal on Numerical Analysis 2021 59 1 558 582 10.1137/20M1329664 Search in Google Scholar

Kruchinin, D. V., & Perminova, M. Y. About Solving Some Functional Equations related to the Lagrange Inversion Theorem. Montes Taurus Journal of Pure and Applied Mathematics., 2021; 3(1): 62–69 KruchininD. V. PerminovaM. Y. About Solving Some Functional Equations related to the Lagrange Inversion Theorem Montes Taurus Journal of Pure and Applied Mathematics 2021 3 1 62 69 Search in Google Scholar

Lemos, N. A., & Moriconi, M. On the consistency of the Lagrange multiplier method in classical mechanics. American Journal of Physics., 2021; 89(8): 776–782 LemosN. A. MoriconiM. On the consistency of the Lagrange multiplier method in classical mechanics American Journal of Physics 2021 89 8 776 782 10.1119/10.0004135 Search in Google Scholar

Burman, E., Hansbo, P., & Larson, M. G. Dirichlet boundary value correction using Lagrange multipliers. BIT Numerical Mathematics., 2020; 60(1): 235–260 BurmanE. HansboP. LarsonM. G. Dirichlet boundary value correction using Lagrange multipliers BIT Numerical Mathematics 2020 60 1 235 260 10.1007/s10543-019-00773-4 Search in Google Scholar

Amahroq, T., Oussarhan, A., & Syam, A. On Lagrange multiplier rules for set-valued optimization problems in the sense of set criterion. Numerical Functional Analysis and Optimization., 2020; 41(6): 710–729 AmahroqT. OussarhanA. SyamA. On Lagrange multiplier rules for set-valued optimization problems in the sense of set criterion Numerical Functional Analysis and Optimization 2020 41 6 710 729 10.1080/01630563.2019.1667828 Search in Google Scholar

Liu, H., Gao, Z., Cao, L., Jiang, Z., Zhang, J., & Song, Y. Tracking control of uncertain Euler–Lagrange systems with fading and saturating actuations: A low-cost neuroadaptive proportional-integral-derivative approach. International Journal of Robust and Nonlinear Control., 2022; 32(5): 2705–2721 LiuH. GaoZ. CaoL. JiangZ. ZhangJ. SongY. Tracking control of uncertain Euler–Lagrange systems with fading and saturating actuations: A low-cost neuroadaptive proportional-integral-derivative approach International Journal of Robust and Nonlinear Control 2022 32 5 2705 2721 10.1002/rnc.5556 Search in Google Scholar

Liu, G., & Zhao, L. Adaptive fuzzy finite-time consensus tracking for multiple Euler-Lagrange systems with unknown control directions. International Journal of Adaptive Control and Signal Processing., 2020; 34(10): 1519–1536 LiuG. ZhaoL. Adaptive fuzzy finite-time consensus tracking for multiple Euler-Lagrange systems with unknown control directions International Journal of Adaptive Control and Signal Processing 2020 34 10 1519 1536 10.1002/acs.3165 Search in Google Scholar

Artículos recomendados de Trend MD

Planifique su conferencia remota con Sciendo