基于有效抽头和进化规划算法的自适应水声信道估计

为了实现稀疏水声信道的快速高效估计,提出一种新的基于有效抽头的信道参数模型降低运算复杂性,并采用进化算法以最小二乘误差为适应度函数对信道响应中有效抽头的阶数、位置和权重同时进行进化操作寻优,避免了有色输入造成的相邻抽头耦合导致的性能下降。仿真实验结果表明:与传统LMS算法及基于有效抽头检测的稀疏信道估计方法相比,该算法在有色输入信号下具有更优越的估计性能和收敛速度,为稀疏水声信道处理提供了一类新方法。     

基于进化规划的自适应高斯神经网络及其在被动声纳目标识别中的应用

自适应高斯神经网络能够对目标信号的功率谱有效识别特征进行自动提取和分类,但此网络使用BP算法,其误差能量函数是一个不规则的超曲面,容易陷入局部极小值.因此,提出了一种使用进化规则来设计和训练自适应高斯神经网络的新方法,该方法能够自动的确定网络的最优结构和联结权值,同时避免网络的局部优化.将该方法用于被动声呐目标的分类识别,实验结果表明基于进化规则的自适应高斯神经网络能够有效的克服局部最小问题,具有更好的识别率.     

量子进化算法

进化算法是解决优化问题的一种有效方法,但在实际应用中也存在收敛速度慢、早熟等问题,大大影响了其应用效果。本文将进化算法和量子理论结合,提出一种新的理论框架-量子进化理论及其学习算法。算法借鉴量子理论,采用量子染色体的表示形式,能使一个染色体同时表示多个状态;模拟量子坍塌的随机观察能带来丰富的种群;同时构造具有量子特点的交叉变异算子,能在防止算法早熟的同时使算法更快收敛。本文不仅从理论上证明了这一理论框架的全局收敛性, 仿真计算也表明了此算法的优越性。     

神经网络的进化学习方法研究

提出了两种用于前向神经网络的进化学习算法,实验证明它们能有效地网络权值空间中寻找全局最优解。在比较实验的基础上,得出了在神经网络的进化学习过程中变异是起主导作用的遗传算子的结论,并以此为指导配置算法的各个关键参数。通过对ＸＯＲ问题和ＩＲＩＳ模式分类问题的学习证明,这两种算法均能获得远高于传统的ＢＰ算法的性能。     

基于进化规划的自适应高斯神经网络

自适应高斯神经网络能够对目标信号的功率谱有效识别特征进行自动提取和分类，但此网络使用BP算法，其误差能量函数是一个不规则的超曲面，容易陷入局部极小值．因此，提出了一种使用进化规则来设计和训练自适应高斯神经网络的新方法．该方法能够自动地确定网络的最优结构和联结权值，同时避免网络的局部优化．实验结果表明，将该方法用于被动声纳目标的分类识别，能够有效地克服局部最小问题，具有更好的识别率．     

高斯变异差分进化算法及其应用研究

为平衡算法收敛速度和全局搜索能力,克服差分进化算法易“早熟”的缺陷,在分析引起种群多样性下降及个体进化停滞原因的基础上,通过引入高斯变异操作,提出了基于高斯变异改进的差分进化算法(modified differential evolution base on Gauss mutation, GMDE). 数值仿真及2个工程优化问题的求解结果表明本文算法能有效避免“早熟”收敛,且在算法收敛速度和全局搜索能力上取得了较好的平衡.     

非线性混合整数规划问题的改进差分进化算法

针对非线性混合整数规划问题,本文采用非固定多段映射罚函数法处理约束条件、用混合整数编码技术处理连续变量和整数变量,并在基本差分进化算法中加入一种新型的凸组合变异算子和一种指数递增交叉算子,由此构造出了一种求解非线性混合整数规划问题的改进差分进化算法。实验表明,所提出的算法全局收敛速度快,精度高,鲁棒性强。     

进化计算在生产调度问题中的应用

本文首先简要介绍了进化计算方法的基本思想和特点,然后对遗传算法求解度问题的某些策略和基本步骤作了简要的归纳和总结,并且初步提出了应用进化策略和进化规划求解调度问题的一般方法。     

弯矩剪力等比数列收敛法

本文所介绍的弯矩剪力等比数列收敛法是以弯矩剪力分配法为基础,通过采用适当的弯矩、剪力的传递方式.用一组等比数列对结构中各杆端的弯矩、剪力进行收敛计算求得精确解的方法;还可解决在弯矩剪力分配法进行结构分析计算时,结构杆端的弯矩,剪力收敛速度慢的问题。     

天然材料，动力与进化并存

很明显，除去对人造产品的渴望之外，对自然产品的创新也可以让我们利用这些产品在越来越多的地方满足我们的需要。[编者按]     

Optimum setting of PID controllers based on using evolutionary programming algorithm

Abstract

This paper is concerned with the optimum setting problem for deriving gains in proportional‐integral‐derivative (PID) controllers. The evolutionary programming (EP) algorithm has been considered as a useful technique for finding global optimization solutions for certain complicated functions in recent years. Therefore, in this paper, we attempt to use the EP algorithm in PID control systems design for deriving optimal or near optimal PID control gains such that a performance index of integratedabsolute error (IAE) is minimized. It can be easily seen from the simulation results that the proposed method will have better performance than those presented in other studies by testing a fourth‐order system with time delay.     

Sectoral dynamics and technological convergence: an evolutionary analysis of eco-innovation in the automotive sector

We know from evolutionary theory that sectoral characteristics are important to innovation. This paper investigates if sectoral characteristics also are important to eco-innovation, a hitherto under-researched theme. We argue that research into possible sectoral patterns in eco-innovation is key to understanding green industrial dynamics and the greening of the economy. This paper investigates to what degree the economy is greening horizontally (sector-wise). Starting with a sectoral case study, we undertake a longitudinal analysis of the breath and strength of the greening of the automotive sector from 1965 to 2012, focusing on powertrain technologies. The empirical analysis is based on patent data amongst big car producers and focuses on identifying changes in two main aspects: (1) the convergence/divergence of firms’ green strategies and technologies within the automotive sector; and (2) the contribution of alternative key green technological trajectories relative to the dominant design. Our findings indicate that the evolution of relative green patenting has followed a positive, linear growth over the last decades with increasing participation of alternative propulsion technologies and increasing convergence of automakers’ strategies towards a diversified portfolio.     

Tensor objects in finite element programming

This paper describes a novel programming tool, nDarray, which is designed using an Object Oriented Paradigm (OOP) and implemented in the C++ programming language. Finite element equations, represented in terms of multidimensional tensors are easily manipulated and programmed. The usual matrix form of the finite element equations are traditionally coded in FORTRAN, which makes it difficult to build and maintain complex program systems. Multidimensional data systems and their implementation details are seldom transparent and thus not easily dealt with and usually avoided. On the other hand, OOP together with efficient programming in C++ allows building new concrete data types, namely tensors of any order, thus hiding the lower level implementation details. These concrete data types prove to be quite useful in implementing complicated tensorial formulae associated with the numerical solution of various elastic and elastoplastic problems in solid mechanics. They permit implementing complex nonlinear continuum mechanics theories in an orderly manner. Ease of use and the immediacy of the nDarray programming tool in constitutive driver programming and in building finite element classes will be shown. © 1998 John Wiley & Sons, Ltd.     

Influence of the number and location of design parameters in the aerodynamic shape optimization of a transonic aerofoil and a wing through evolutionary algorithms and support vector machines

Surrogate-based optimization (SBO) has recently found widespread use in aerodynamic shape design owing to its promising potential to speed up the whole process by the use of a low-cost objective function evaluation, to reduce the required number of expensive computational fluid dynamics simulations. However, the application of these SBO methods for industrial configurations still faces several challenges. The most crucial challenge nowadays is the ‘curse of dimensionality’, the ability of surrogates to handle a high number of design parameters. This article presents an application study on how the number and location of design variables may affect the surrogate-based design process and aims to draw conclusions on their ability to provide optimal shapes in an efficient manner. To do so, an optimization framework based on the combined use of a surrogate modelling technique (support vector machines for regression), an evolutionary algorithm and a volumetric non-uniform rational B-splines parameterization are applied to the shape optimization of a two-dimensional aerofoil (RAE 2822) and a three-dimensional wing (DPW) in transonic flow conditions.     

Selecting key performance indicators for production with a linear programming approach

Modern production systems are prone to disruptions due to shorter product life cycles, growing variant diversity and progressively distributed production. At the same time, reduced time and capacity buffers diminish mitigation opportunities, requiring better tools for production control. Performance measurement with key performance indicators (KPIs) is a widely used instrument to detect changes in production system performance in order to coordinate appropriate countermeasures. The main challenge in planning KPI systems consists in determining relevant KPIs. On the one hand, enough KPIs must be selected for a sufficiently high information content. On the other hand, the cognitive abilities of users are not to be overstrained by selecting too many KPIs. This tradeoff is addressed in a proposed selection process using an integer linear programme for objective KPI selection. In order to achieve this goal, crucial facets of the information content requirement are formalised mathematically. The developed method is validated using a practical application example, showing the influence of model parameter selection on optimisation results. The formalisation of the information content is shown to be a novel and promising approach.     

Investigating replenishment policies for centralised and decentralised supply chains using stochastic programming approach

In this paper, a multiple period replenishment problem based on (s, S) policy is investigated for a supply chain (SC) comprising one retailer and one manufacturer with uncertain demand. Novel mixed-integer linear programming (MILP) models are developed for centralised and decentralised decision-making modes using two-stage stochastic programming. To compare these decision-making modes, a Monte Carlo simulation is applied to the optimization models’ policies. To deal with demand uncertainty, scenarios are generated using Latin Hypercube Sampling method and their number is reduced by a scenario reduction technique. In large test problems, where CPLEX solver is not able to reach an optimal solution in the centralised model, evolutionary strategies (ES) and imperialist competitive algorithm (ICA) are applied to find near optimal solutions. Sensitivity analysis is conducted to show the performance of the proposed mathematical models. Moreover, it is demonstrated that both ES and ICA provide acceptable solutions compared to the exact solutions of the MILP model. Finally, the main parameters affecting difference between profits of centralised and decentralised SCs are investigated using the simulation method.     

Second-order cone programming formulation of discontinuous deformation analysis

In classic discontinuous deformation analysis (DDA), artificial springs must be employed to enforce the contact condition through the open-close iteration. However, improper stiffness parameters might cause numerical problems. The main goal of this paper is to propose a new framework of DDA using second-order cone programming. The complementarity relationship at contacts can be formulated directly; thus, artificial springs are avoided. Stemming from the equations of momentum conservation of each block, the governing equations of DDA can be cast as convex optimization problems. The basic variables in the formulations can be either block displacements or contact forces. The derived optimization problems can be reformulated into a standard second-order cone programming program, which can be solved using standard efficient optimization solvers. The proposed approach is validated by a series of numerical examples.     

A non‐dominance‐based online stopping criterion for multi‐objective evolutionary algorithms

A non‐dominance criterion‐based metric that tracks the growth of an archive of non‐dominated solutions over a few generations is proposed to generate a convergence curve for multi‐objective evolutionary algorithms (MOEAs). It was observed that, similar to single‐objective optimization problems, there were significant advances toward the Pareto optimal front in the early phase of evolution while relatively smaller improvements were obtained as the population matured. This convergence curve was used to terminate the MOEA search to obtain a good trade‐off between the computational cost and the quality of the solutions. Two analytical and two crashworthiness optimization problems were used to demonstrate the practical utility of the proposed metric. Copyright © 2010 John Wiley & Sons, Ltd.