Genetic algorithm‐based neural‐network modeling approach applied to AlGaN/GaN devices

An accurate equivalent circuit large‐signal model (ECLSM) for AlGaN‐GaN high electron mobility transistor (HEMT) is presented. The model is derived from a distributed small‐signal model that efficiently describes the physics of the device. A genetic neural‐network‐based model for the gate and drain currents and charges is presented along with its parameters extraction procedure. This model is embedded in the ECLSM, which is then implemented in CAD software and validated by pulsed and continuous large‐signal measurements of on‐wafer 8 × 125‐μm GaN on SiC substrate HEMT. Pulsed IV simulations show that the model can efficiently describe the bias dependency of trapping and self‐heating effects. Single‐ and two‐tone simulation results show that the model can accurately predict the output power and its harmonics and the associated intermodulation distortion (IMD) under different input‐power and bias conditions. © 2012 Wiley Periodicals, Inc. Int J RF and Microwave CAE, 2013.     

Using anatomical knowledge expressed as fuzzy constraints to segment the heart in CT images

Segmenting the heart in medical images is a challenging and important task for many applications. In particular, segmenting the heart in CT images is very useful for cardiology and oncological applications such as radiotherapy. Although the majority of methods in the literature are designed for ventricle segmentation, there is a real interest in segmenting the heart as a whole in this modality. In this paper, we address this problem and propose an automatic and robust method, based on anatomical knowledge about the heart, in particular its position with respect to the lungs. This knowledge is represented in a fuzzy formalism and it is used both to define a region of interest and to drive the evolution of a deformable model in order to segment the heart inside this region. The proposed method has been applied on non-contrast CT images and the obtained results have been compared to manual segmentations of the heart, showing the good accuracy and high robustness of our approach.     

改进的猴王遗传算法求解组合拍卖最优竞胜标

应用遗传算法求解第一价格封闭组合拍卖问题,编码及约束条件复杂,在交叉和变异的过程中容易产生大量不可行解,并且在选择和精英保留中,一些局部最优点(早熟点)对应的个体容易被大量复制,使群体多样性下降。该文就以上的问题,改进了原有遗传算法求解组合拍卖最优竞胜标的方法,并采用了猴王遗传算法的精英保留策略。实际计算证明本算法编码方式自然并且直观表达了个体和标的的关系。在交叉和变异算子中,产生不可行解的数量下降,更积极的保存了上一代进化的部分成果。而改进猴王遗传算法的精英保留策略,使算法能够以高概率跳出早熟,达到算法速度与准确的平衡。     

遗传算法与文本特征选择方法相结合的研究

对文本特征选择的传统方法进行了深入分析并对各自的优缺点进行了详细的对比;其次,通过利用遗传算法自身的寻优特点,把它与传统特征选择方法相结合,并得出了算法。最后,在对各自算法进行了分析,表明了在性能上有了明显的提高。     

基于决策树及遗传算法的人工免疫入侵检测算法

针对传统的否定选择算法中生成大量无效抗体,并且抗体之间缺乏多样性的问题,设计了基于决策树及遗传算法的人工免疫入侵检测算法。将决策树和遗传算法引入传统的否定选择算法中,利用决策树计算抗原和抗体之间的亲和力,提出了新fitness的计算公式,并利用抗体浓度衡量抗体集的多样性,将低浓度抗体代替高浓度抗体,实现了抗体的多样性,确保了当抗体集的数量一定时尽可能覆盖最大的非自体集空间,以提高抗体集的性能。     

全自动酶免工作站计算机控制系统设计

介绍了酶免工作站软件控制部分的分析和设计。按软件工程的设计思想,归纳出系统的对象,给出系统实现单文档多视和控制类的详细设计。并将一种改进的自适应遗传算法应用于系统执行效率的优化。此开发系统已投入实际试验,从实际使用的效果看,试验执行效率提高了69.9%,血样检测的准确率也达到了99.6%以上,因此,每台酶免仪每天可以多检测120个左右的血样,如果按照现在医院血液检测的指导价,每台酶免仪每年可为医院多创造150万人民币左右的利润。     

Matlab遗传算法工具箱在非线性优化中的应用

投影寻踪是一种降维处理技术,通过它可以将多维分析问题通过投影方向转化为一维问题分析。应用该法的关键在于寻求最佳投影方向,这可以转化为一个复杂的非线性优化问题来进行解决。选取某地区大气环境质量评价的投影寻踪评价模型,编制相关的目标函数和约束函数,应用基于MATLAB的遗传算法和直接搜索工具箱进行优化求解。结果表明：该工具箱在求解此类非线性优化问题上的有效性和方便性,从而为各领域应用投影寻踪模型提供了强有力的优化工具。     

Optimal imperfect preventive maintenance policy for equipment leased during successive periods

In this paper, we consider randomly failing equipment leased several times during their life cycle with a given warranty period. A mathematical model is developed to determine the optimal efficiency levels of preventive maintenance (PM) to be performed on the equipment between successive lease periods, maximising the expected total profit of the lessor over the equipment life cycle. The model considers the expected leasing revenue as well as the equipment acquisition cost and the average PM and repair costs. PM actions allow reducing the age of the equipment to a certain extent with a corresponding cost depending on the PM level adopted. The efficiency of the PM is determinant of the expected revenue during the next lease period. Given a set of K possible PM levels and the number of lease periods n over the equipment life cycle, K^{n?1 PM} strategies are possible. A genetic algorithm is proposed in order to obtain nearly optimal policies in situations where the number of possibilities K^n?1 is very high. Obtained numerical results are discussed. Small- and big-size instances of the problem are considered in the case of a service company in the oil and gas industry specialised in leasing specific equipment such as separators, to oil companies for production activities with a limited duration of several months like well testing or short production tests.     

An experimental study on the applicability of evolutionary algorithms to craniofacial superimposition in forensic identification

Photographic supra-projection is a forensic process that aims to identify a missing person from a photograph and a skull found. One of the crucial tasks throughout all this process is the craniofacial superimposition which tries to find a good fit between a 3D model of the skull and the 2D photo of the face. This photographic supra-projection stage is usually carried out manually by forensic anthropologists. It is thus very time consuming and presents several difficulties. In this paper, we aim to demonstrate that real-coded evolutionary algorithms are suitable approaches to tackle craniofacial superimposition. To do so, we first formulate this complex task in forensic identification as a numerical optimization problem. Then, we adapt three different evolutionary algorithms to solve it: two variants of a real-coded genetic algorithm and the state of the art evolution strategy CMA-ES. We also consider an existing binary-coded genetic algorithm as a baseline. Results on several superimposition problems of real-world identification cases solved by the Physical Anthropology lab at the University of Granada (Spain) are considered to test our proposals.     

Evolving Teams of Predictors with Linear Genetic Programming

This paper applies the evolution of GP teams to different classification and regression problems and compares different methods for combining the outputs of the team programs. These include hybrid approaches where (1) a neural network is used to optimize the weights of programs in a team for a common decision and (2) a realnumbered vector (the representation of evolution strategies) of weights is evolved with each term in parallel. The cooperative team approach results in an improved training and generalization performance compared to the standard GP method. The higher computational overhead of team evolution is counteracted by using a fast variant of linear GP.