基于近似模型的非线性区间数优化方法及其应用

在不确定优化中,非线性区间数优化方法由于需要嵌套优化,造成计算效率低下而阻碍其应用于工程实际.本文提出了一种基于径向基函数近似模型的求解方法,以提高非线性区间数优化方法的计算效率.该方法利用拉丁超立方实验设计方法采样,建立目标函数和各约束的径向基函数近似模型.利用近似模型代替嵌套优化中的真实模型,再用非线性区间数优化方法进行求解,从而提高了非线性区间数优化方法的计算效率,使得该算法在工程应用方面成为可能.用一个测试函数验证了该方法的可行性,最后将方法应用于车身薄壁梁的耐撞性优化.     

A multi-objective optimization method for uncertain structures based on nonlinear interval number programming method

A multi-objective optimization method for uncertain structures is developed based on nonlinear interval number programming (NINP) method. The NINP method is employed to transform each uncertain objective function into a deterministic single-objective optimization problem. Using the constraint penalty function method, a deterministic multi-objective and non-constraint optimization problem is formulated in terms of penalty functions. Then the micro multi-objective genetic algorithm and the intergeneration projection genetic algorithm are adopted as outer layer and inner optimization operator to solve the nesting optimization problem, respectively. Finally, four numerical examples are provided to demonstrate the effectiveness of the present method.     

基于算法复杂度理论的拟力法计算效率评价

算法复杂度理论是一种算法效率定量评价方法,该方法通过度量算法的复杂度来客观反映其执行效率,避免了计算机系统性能等因素对效率评价的影响。拟力法是一种高效的结构非线性分析方法,以往研究工作只是在非线性分析过程和运行时间上对计算效率进行了探索,并没有从理论上量化分析。本文采用算法复杂度理论对拟力法和传统变刚度非线性求解方法进行分析,给出了两种方法的时间复杂度函数,并对比了其计算效率进行定量,从数学角度解释了拟力法计算效率高的根本原因。算例对两种方法的时间复杂度和计算时间进行对比分析,直观地说明了拟力法在计算效率方面的优越性。     

基于遗传算法的设计地震反应谱标定方法

基于自然进化原理的遗传算法是一种解决复杂非线性规划问题十分有前景的方法。对遗传算法中一些有代表性的算法作了简要分析和评价。介绍了用地面峰值加速度(PGA)、峰值速度(PGV)和峰值位移(PGD)三个参数标定设计地震反应谱的模型。针对此分段非线性函数规划问题,运用MATLAB编写遗传算法程序进行了计算。计算表明遗传算法在全局优化、非线性优化、多参数优化等方面表现出了传统算法无法比拟的优势。     

Nonlinear model predictive control based on piecewise linear Hammerstein models

This paper develops a nonlinear model predictive control (MPC) algorithm for dynamic systems represented by piecewise linear (PWL) Hammerstein models. At each sampling instant, the predicted output trajectory is linearized online at an assumed input trajectory such that the control actions can be easily calculated by solving a quadratic programming optimization problem, and such linearization and optimization may be repeated a few times for good linear approximation accuracy. A three-step procedure is developed to linearize a PWL function, where the derivatives of a PWL function are obtained by a computationally efficient look-up table approach. Unlike many existing MPC algorithms for Hammerstein systems, it does not require the inversion of static nonlinearity and can directly cope with input constraints even in multivariable systems. Two benchmark chemical reactors are studied to illustrate the effectiveness of the proposed algorithm.     

Self-adaptive one-dimensional nonlinear finite element method based on element energy projection method

The element energy projection （EEP） method for computation of super- convergent resulting in a one-dimensional finite element method （FEM） is successfully used to self-adaptive FEM analysis of various linear problems, based on which this paper presents a substantial extension of the whole set of technology to nonlinear problems. The main idea behind the technology transfer from linear analysis to nonlinear analysis is to use Newton＇s method to linearize nonlinear problems into a series of linear problems so that the EEP formulation and the corresponding adaptive strategy can be directly used without the need for specific super-convergence formulation for nonlinear FEM. As a re- sult, a unified and general self-adaptive algorithm for nonlinear FEM analysis is formed. The proposed algorithm is found to be able to produce satisfactory finite element results with accuracy satisfying the user-preset error tolerances by maximum norm anywhere on the mesh. Taking the nonlinear ordinary differential equation （ODE） of second-order as the model problem, this paper describes the related fundamental idea, the imple- mentation strategy, and the computational algorithm. Representative numerical exam- ples are given to show the efficiency, stability, versatility, and reliability of the proposed approach.     

一种土的非线性弹性本构模型参数的反演方法

旨在提出一种土的非线性弹性本构模型参数反演的方法。以现今普遍实行的地基载荷试验为基础,依据遗传算法的组合优化理论,采用正演计算和遗传算法优化相结合的方式,建立了土层非线性弹性本构模型参数反演的方法;并依据某黄土场地地基载荷试验数据,实施了黄土土层非线性弹性本构模型参数反演的全过程。计算结果表明,所建立的方法可以实现土层非线性弹性本构模型中相互关联的多个参数的组合优化,并在对初始值要求较低的情况下,可以获得良好的参数反演精度。从而为土的变形特性分析和土与其中及相邻结构的共同作用分析,提供了较好的土体本构模型参数的确定方法。     

基于区间的不确定多目标优化方法研究

基于非线性区间优化,提出了一种不确定多目标优化方法.基于区间序关系和区间可能度,把不确定多目标的目标函数和约束转化为确定性的目标函数和确定性的约束.对于复杂的工程优化问题,为了提高效率,采用拉丁方试验设计方法,构建响应面近似模型,并基于近似模型进行不确定多目标优化,从而形成了非线性区间优化方法与近似模型相结合的高效不确定多目标优化方法.数值算例表明了该方法的有效性和工程实用性.     

Hybrid approach for dynamic model identification of an electro-hydraulic parallel platform

In this paper, a hybrid optimization algorithm is proposed to identify the dynamic parameters of a 6-DOF electro-hydraulic parallel platform. The dynamic model of a parallel platform with arbitrary geometry, inertia distribution and frictions is obtained based on a structured Boltzmann–Hamel–d’Alembert formulation, and then the estimation equations are explicitly expressed in terms of a linear form with respect to the identified inertial and the friction coefficients in accordance with a linear friction model. However, when nonlinear friction models are considered, the parameter identification of the electro-hydraulic parallel platform is considered as an optimization process with an objective function minimizing the errors between the measurement and identification, and then an effective combination of the particle swarm optimization (PSO) method and the local quasi-Newton method is proposed to solve the identification problem. Experimental identification processes are carried out for the identified parameters, and the identified models are compared by the predicted forces between the LS method and the optimization technique as well as between the linear and nonlinear friction models.     

求解一类可分离凸规划的对偶显式模型DP-EM方法

推导对偶目标函数的精确显式表达式,可选用更多成熟高效的求解方法,从而进一步提高了非线性规划对偶理论求解结构拓扑优化问题的效率.研究工作来源于非线性凸规划同其对偶规划的间隙为零,可以等价转化为对偶问题求解,通常可以大大地缩小问题的规模,可是二者不具有显式关系却影响了对偶解法的应用.所幸的是,结构优化当中一大类问题包括连续体结构拓扑优化问题,不仅具有凸性,而且具有变量可分离性,于是原变量和对偶变量之间有了显式关系,因此,对偶解法成了38年来被应用的有效方法之一.然而长期以来,对偶问题的目标函数并不是显式,这缘于含参数的极小化问题导致目标函数为隐式表达,常见的显式化方法是进行二阶近似.本文突破了对偶问题难以显式化只能采用近似显式的定势,将我们提出的"对偶规划-显式模型"(DP-EM)方法应用于连续体结构拓扑优化,并与对偶序列二次规划(DSQP)算法及移动渐近线(MMA)算法为求解器的方法进行计算效率对比,结果显示:(1)MMA算法比DP-EM算法和DSQP算法的外部迭代次数均多;(2)DP-EM算法与DSQP算法外循环次数相同,而内循环数显著减少.说明了DP-EM算法具有显式对偶函数的优势.     

An optimization method based on the generalized polynomials for nonlinear variable-order time fractional diffusion-wave equation

In this paper, an optimization method based on a new class of basis functions, namely generalized polynomials (GPs), is proposed for nonlinear variable-order time fractional diffusion-wave equation. Variable-order time fractional derivative is expressed in the Caputo sense. In the proposed method, solution of the problem under consideration is expanded in terms of GPs with unknown free coefficients and control parameters. In this way, some new operational matrices of the ordinary and fractional derivatives are derived for these basis functions. The residual function and its 2-norm are employed for converting the problem under study to an optimization one and then choosing the unknown free coefficients and control parameters optimally. As a useful result, the necessary conditions of optimality are derived as a system of nonlinear algebraic equations with unknown free coefficients and control parameters. The validity and effectiveness of the method are demonstrated by solving some numerical examples. The results demonstrate that the proposed method is a powerful algorithm with good accuracy for solving such kind of problems.     

Shakedown analysis of engineering structures with limited kinematical hardening

We present a numerical method for the computation of shakedown loads of engineering structures with limited kinematical hardening under thermo-mechanical loading. The method is based on Melan’s statical shakedown theorem, which results in a nonlinear convex optimization problem. This is solved by an interior-point algorithm recently developed by the authors, specially designed for lower bound shakedown analysis of large-scale problems. Limited kinematical hardening is taken into account by use of a two-surface model, such that both alternating plasticity and incremental collapse can be captured. For the yield surface as well as for the bounding surface the von Mises criterion is used. The proposed method is validated by two examples, where numerical results are compared to those of literature where available.     

Modelling sediment transport over partially non‐erodible bottoms

In‐depth‐averaged and cross‐section‐averaged morphodynamic models, based on explicit time integration, it may happen that the computed bed level becomes lower than the top level of a non‐erodible layer (e.g. concrete, bedrock or armoured layer). This is a standard pitfall, which has been addressed in different ways. In this paper, we present an original approach for avoiding computation of non‐physical bed levels, using an iterative procedure to correct the outward sediment fluxes. The procedure is shown to be computationally efficient while it achieves a high accuracy in terms of mass conservation. We compare our original approach with the existing Struiksma's method and with a reformulation of the problem in terms of mathematical optimization of a linear or nonlinear objective function under linear constraints. The new procedure has been incorporated into an existing finite volume morphodynamic model. It has been validated with several 1D benchmarks leading to configurations with sediment transport over non‐erodible bottom. The computation time has been verified not to increase by more than 15% compared with runs without non‐erodible bottom. Copyright © 2011 John Wiley & Sons, Ltd.     

非线性函数的混沌优化方法比较研究

已有的混沌优化方法几乎都是利用Logistic映射作为混沌序列发生器，而Logistic映射产生的混沌序列的概率密度函数服从两头多、中间少的切比雪夫型分布，不利于搜索的效率和能力。为此，首先根据Logistie映射混沌轨道点密度函数的特点，建立改进的混沌-BFGS混合优化算法。之后，考虑到Kent映射混沌轨道点密度为均匀分布，建立了基于Kent映射的混沌-BFGS混合优化算法。然后对五种混合优化方法——不加改进的和改进的基于Logistic映射的混沌-BFGS法，基于Kent映射的混沌-BFGS法，Monte Carlo试验-BFGS法，网格-BFGS法进行了研究，分别对3个低维和2个高维非线性复杂测试函数进行优化计算，对它们的全局优化计算效率和寻优能力做了比较，并探讨了混合优化方法全局优化性能差异的原因。结果表明，混沌优化方法是与Monte Carlo方法类似的一种随机性试验优化方法。而且，这类优化方法的计算性能至少与以下因素有关：混沌／随机序列的统计性质，优化问题全局最优点位置。     

Kinematic limit analysis of frictional materials using nonlinear programming

In this paper, a nonlinear numerical technique is developed to calculate the plastic limit loads and failure modes of frictional materials by means of mathematical programming, limit analysis and the conventional displacement-based finite element method. The analysis is based on a general yield function which can take the form of the Mohr–Coulomb or Drucker–Prager criterion. By using an associated flow rule, a general nonlinear yield criterion can be directly introduced into the kinematic theorem of limit analysis without linearization. The plastic dissipation power can then be expressed in terms of kinematically admissible velocity fields and a nonlinear optimization formulation is obtained. The nonlinear formulation only has one constraint and requires considerably less computational effort than a linear programming formulation. The calculation is based entirely on kinematically admissible velocities without calculation of the stress field. The finite element formulation of kinematic limit analysis is developed and solved as a nonlinear mathematical programming problem subject to a single equality constraint. The objective function corresponds to the plastic dissipation power which is then minimized to give an upper bound to the true limit load. An effective, direct iterative algorithm for kinematic limit analysis is proposed in this paper to solve the resulting nonlinear mathematical programming problem. The effectiveness and efficiency of the proposed method have been illustrated through a number of numerical examples.     

A novel APSO-aided maximum likelihood identification method for Hammerstein systems

Identification of Hammerstein nonlinear models has received much attention due to its ability to describe a wide variety of nonlinear systems. In this paper the maximum likelihood estimator which was originally derived for linear systems is extended to work for Hammerstein nonlinear systems in colored-noise environment. The maximum likelihood estimate is known to be statistically efficient, but can lead to complex nonlinear multidimensional optimization problem; traditional methods solve this problem at the computational cost of evaluating second derivatives. To overcome these shortcomings, a particle swarm optimization (PSO) aided maximum likelihood identification algorithm (Maximum Likelihood-Particle Swarm Optimization, ML-PSO) is first proposed to integrate PSO’s simplicity in implementation and computation, and its ability to quickly converge to a reasonably good solution. Furthermore, a novel adaptive strategy using the evolution state estimation technique is proposed to improve PSO’s performance (maximum likelihood-adaptive particle swarm optimization, ML-APSO). A simulation example shows that ML-APSO method outperforms ML-PSO and traditional recursive least square method in various noise conditions, and thus proves the effectiveness of the proposed identification scheme.     

Two novel nonlinear multivariate grey models with kernel learning for small-sample time series prediction

For many applications, small-sample time series prediction based on grey forecasting models has become indispensable. Many algorithms have been developed recently to make them effective. Each of these methods has a specialized application depending on the properties of the time series that need to be inferred. In order to develop a generalized nonlinear multivariable grey model with higher compatibility and generalization performance, we realize the nonlinearization of traditional GM(1,N), and we call it NGM(1,N). The unidentified nonlinear function that maps the data into a better representational space is present in both the NGM(1,N) and its response function. The original optimization problem with linear equality constraints is established in terms of parameter estimation for the NGM(1,N), and two different approaches are taken to solve it. The former is the Lagrange multiplier method, which converts the optimization problem into a linear system to be solved; and the latter is the standard dualization method utilizing Lagrange multipliers, that uses a flexible estimation equation for the development coefficient. As the size of the training data increases, the estimation results of the potential development coefficient get richer and the final estimation results using the average value are more reliable. The kernel function expresses the dot product of two unidentified nonlinear functions during the solving process, greatly lowering the computational complexity of nonlinear functions. Three numerical examples show that the LDNGM(1,N) outperforms the other multivariate grey models compared in terms of generalization performance. The duality theory and framework with kernel learning are instructive for further research around multivariate grey models to follow.

     

A nonlinear programming approach to kinematic shakedown analysis of frictional materials

This paper develops a novel nonlinear numerical method to perform shakedown analysis of structures subjected to variable loads by means of nonlinear programming techniques and the displacement-based finite element method. The analysis is based on a general yield function which can take the form of most soil yield criteria (e.g. the Mohr–Coulomb or Drucker–Prager criterion). Using an associated flow rule, a general yield criterion can be directly introduced into the kinematic theorem of shakedown analysis without linearization. The plastic dissipation power can then be expressed in terms of the kinematically admissible velocity and a nonlinear formulation is obtained. By means of nonlinear mathematical programming techniques and the finite element method, a numerical model for kinematic shakedown analysis is developed as a nonlinear mathematical programming problem subject to only a small number of equality constraints. The objective function corresponds to the plastic dissipation power which is to be minimized and an upper bound to the shakedown load can be calculated. An effective, direct iterative algorithm is then proposed to solve the resulting nonlinear programming problem. The calculation is based on the kinematically admissible velocity with one-step calculation of the elastic stress field. Only a small number of equality constraints are introduced and the computational effort is very modest. The effectiveness and efficiency of the proposed numerical method have been validated by several numerical examples.     

随机结构非线性动力响应的概率密度演化分析

提出了随机结构非线性动力响应分析的概率密度演化方法．根据结构动力响应的随机状态方程，利用概率守恒原理，建立了随机结构非线性动力响应的概率密度演化方程．结合Newmark-Beta时程积分方法与Lax-Wendroff差分格式，提出了概率密度演化方程的数值分析方法．通过与Monte Carlo分析方法对比，表明所给出的概率密度演化方法具有良好的计算精度和较小的计算工作量．研究表明：随机结构非线性动力响应概率密度具有典型的演化特征，随着时间增长，概率密度曲线分布趋于复杂．     

优化迭代步长的两种改进增量谐波平衡法

增量谐波平衡法(IHB法)是一个半解析半数值的方法, 其最大优点是适合于强非线性系统振动的高精度求解. 然而, IHB法与其他数值方法一样, 也存在如何选择初值的问题, 如初值选择不当, 会存在不收敛的情况. 针对这一问题, 本文提出了两种基于优化算法的IHB法: 一是结合回溯线搜索优化算法(BLS)的改进IHB法(GIHB1), 用来调节IHB法的迭代步长, 使得步长逐渐减小满足收敛条件; 二是引入狗腿算法的思想并结合BLS算法的改进IHB法(GIHB2), 在牛顿-拉弗森(Newton-Raphson)迭代中引入负梯度方向, 并在狗腿算法中引入2个参数来调节BSL搜索方式用于调节迭代的方式, 使迭代方向沿着较快的下降方向, 从而减少迭代的步数, 提升收敛的速度. 最后, 给出的两个算例表明两种改进IHB法在解决初值问题上的有效性.