Optimization of color design for military camouflage in CIELAB color space

The assessment of military camouflage is a key consideration in the modern military field. Traditionally, the assessment relies on traditional human visual detection tests because a large scale multi‐level and multi‐factor experiments are time‐ and resource‐consuming. One aspect of camouflage assessment, to which this current study pertains, entails improving upon or “enhancing” an existing or “selected” design. The current study presents a new and practical approach for enhancing the selected military camouflage by utilizing response surface methodology (RSM) of %L*, %a*, and %b* in CIELAB color space. Ten participants were recruited to evaluate 35 variations of %L*, %a*, and %b* on camouflage similarity index (CSI) and reaction time (RT). Based on RSM, the optimum combination occurs at L*: 61.4966, a*: ?5.6505, and b*: 10.5114. In addition, a predictive algorithm to calculate the optimum shift of %L*, %a*, and %b* from the original camouflage to the improved camouflage derived from RSM is also proposed. The optimum shift occurs at ?25%L*, ?55%a*, and + 80%b*. In the end, a new design guideline is proposed for the enhancement of selected military camouflage, which adopts the present study's research findings.     

一种求解化工动态优化问题的迭代自适应粒子群方法

智能优化方法因其简单、易实现且具有良好的全局搜索能力,在动态优化中的应用越来越广泛,但传统的智能方法收敛速度相对较慢。提出了一种迭代自适应粒子群优化方法（IAPSO）来求解一般的化工动态优化问题。首先通过控制变量参数化将原动态优化问题转化为非线性规划问题,再利用所提出的迭代自适应粒子群优化方法进行求解。相比传统的粒子群优化方法,该种迭代自适应粒子群优化方法具有收敛速度更快的优点,主要原因是：该算法根据粒子种群分布特性自适应调整参数;该算法通过缩减搜索空间并迭代使用粒子群算法搜索最优解。将提出的迭代自适应粒子群方法应用到多个经典动态优化问题中,测试结果表明,该方法简单、有效,精度高,且收敛速度比传统粒子群算法有显著提升。     

基于粒子群算法的隔离壁精馏塔的综合与设计

隔离壁精馏塔（dividing-wall distillation column,DWDC）是提高两个或者多个传统精馏塔热力学效率的有效手段。由于隔离壁精馏塔内部结构复杂、相互作用强,传统的序贯优化方法计算时间长,很难达到全局最优解。标准粒子群算法应用广泛、易于实现,但易于早熟、易于陷入局部极值点。因此,本工作采用改进的元胞粒子群算法对Kaibel隔离壁精馏塔进行综合与设计研究。元胞粒子群算法通过改进粒子的学习策略,采用元胞邻域的方法可有效地将粒子分散在多个子空间。对比标准粒子群优化和元胞粒子群优化两种方法的50次优化效果,结果表明,两种粒子群算法能够对内部结构复杂、相互作用强的四组分Kaibel隔离壁精馏塔这一复杂分离系统进行优化,优化效果显著。     

多目标粒子群算法用于补料分批生化反应器动态多目标优化

多目标优化是过程系统工程的重要课题,通常以加权或约束方式将其转换为单一目标,未能反映多目标间的复杂关系,不利于随时根据需求作出有效的决策。基于群智能的粒子群算法具有全局优化性能,且易于实现。为使其适于多目标优化,应拓展功能,实施改造。以Pareto支配概念评价种群个体的优劣,设计了确定局部最优点和全局最优点的操作。又利用各粒子的局部最优点信息进行速度更新,以加强种群的多样性,避免因早熟而陷于局部最优。还设置了外部优解库,并通过分散度计算,以适当的策略进行更新,使之逐步均匀地逼近于Pareto最优解集。由此构建一种多目标粒子群优化算法（multi-objective particle swarm optimization,MOPSO）,并用于补料分批生化反应器的动态多目标优化,取得了满意的结果。可基于所搜得的Pareto最优解集,分析目标间的关系,为合理决策提供有效的支持。经与NSGA-II比较,MOPSO算法具有更为优良的性能。     

基于遗传算法和粒子群优化算法的电力系统无功优化

从数学的角度分析,电力系统无功优化是一个多变量、多约束、非连续性的混合非线性规划问题,因此,优化过程十分复杂.以减少有功网损为目标函数建立电力系统无功优化计算的数学模型,基于遗传算法和粒子群优化算法,提出一种新颖的混合策略来求解无功优化问题.IEEE 6和IEEE 14节点系统的仿真计算结果表明:与单一的遗传算法或粒子群优化算法相比,该混合策略在优化效果方面具有明显的优势.     

基于LCPSO算法的相稳定性判别

相稳定性判别为相平衡计算的基本课题,常采用Gibbs自由能曲面与切平面的距离函数（TPDF）最小化方法求解。对于强非理想体系,或在高压条件下,其TPDF表现出复杂形态,有平凡解和多极值,传统方法难以求得满足约束的全局最小值,从而导致判别失误。粒子群算法（PSO）虽有全局优化性能,但也会陷于局部极小,且缺少约束处理机制。为此,分析了PSO内在蕴含的线性特点,在种群初化、粒子运动等环节提出应对策略,构建线性约束粒子群算法（LCPSO）,确保种群在可行空间内搜索。还增设调变参数、局部加速等措施,以兼顾算法的全面探测和细化挖掘的能力,提高其全局优化效能。经多个实例的测试表明,LCPSO适用面广,既可用于超额自由能、状态方程等各类热力学模型,又能克服混合模型一阶不连续的困难,应用范围从液液相分裂拓展到汽液液相分裂。与确定性全局算法IN/GB相比,LCPSO速率高,效果好,尤对多元体系更具优势。     

Illumination correction of dyed fabrics method using rotation forest‐based ensemble particle swarm optimization and sparse least squares support vector regression

Different illuminations adversely affect color difference evaluation of textile images in dyed fabrics. To address the problem, we propose a rotation forest (RF)‐based ensemble particle swarm optimization and sparse least squares support vector regression (RF‐PSO‐SLSSVR) for building an accurate illumination correction model. In our algorithm, grey‐edge is first used to extract the statistics characteristics of the textile image. Second, as the standard LSSVR cannot yield a sparse solution, we develop sparse LSSVR (SLSSVR) by calculating the maximal independent subset in the extracted feature space. Then, SLSSVR is embedded into RF by substituting for the regression tree which is the base learner in the original RF, and the PSO technique is employed to obtain the optimal regularization parameter γ and kernel parameter σ. The final model is obtained by fusing the predictions of the different trees through a weighted average method and RF‐PSO‐SLSSVR is constructed to learn the textile illumination estimation model. To verify the effectiveness of our algorithm, we carry out the experiments on the real dyed fabric images by comparison to several related methods and the performance is measured by the different criterions, including the chromaticity error, the angle error, and the Wilcoxon signed‐rank test. Compared with the traditional SVR and ELM algorithm, the results show that the RF‐PSO‐SLSSVR method reduces ~13.6% and 10.6% over the angle RMSE.     

A discrete particle swarm optimization approach for classification of Indian coal seams with respect to their spontaneous combustion susceptibility

Mine fires due to spontaneous combustion in coal mines is a global concern. This leads to serious environmental and safety hazards and considerable economic losses. Therefore it is essential to assess and classify the coal seams with respect to their proneness to spontaneous combustion to plan the production, storage and transportation capabilities in mines. This paper presents the formulation of clustering problem into a linear assignment model and the application of a discrete particle swarm optimization approach for the classification of coal seams based on their proneness to spontaneous combustion. In this research work, twenty nine coal samples of varying ranks belonging to both high and low susceptibilities to spontaneous combustion have been collected from all the major coalfields of India. Using moisture, volatile matter, and ash content and crossing point temperature of the coal samples as the parameters, the proposed algorithm has been applied to classify the coal seams into three different categories. This classification will be useful for the planners and field engineers for taking ameliorative measures in advance for preventing the occurrence of mine fires.     

Establishment and solution of eight-lump kinetic model for FCC gasoline secondary reaction using particle swarm optimization

An eight-lump kinetic model contained 21 kinetic parameters was proposed to describe the secondary reaction process of fluid catalytic cracking (FCC) gasoline. The model was solved by hybrid particle-swarm optimization (HPSO) which incorporated evolutionary strategies and the simulated annealing method into particle swarm optimization (PSO). A series of experiments were carried out in a riser reactor over an improved Y zeolite catalyst with different temperatures, catalyst to oil ratios and vapor residence times. The product distribution was obtained to estimate the 21 kinetic parameters of model; the calculated results obtained using the HPSO algorithm agreed well with the experimental results.     

混合多目标骨干粒子群优化算法在污水处理过程优化控制中的应用

通过对污水生化处理过程的分析,选取能耗和罚款最低为优化目标,建立污水生化处理过程多目标优化控制模型。为了提高Pareto最优解集的收敛性和多样性,提出一种基于Pareto支配和分解的混合多目标骨干粒子群优化算法（HBBMOPSO）。该方法采用带自适应惩罚因子的分解方法选取个体引导者,采用Pareto支配和拥挤距离法维护外部档案和选取全局引导者。此外,采用精英学习策略增强粒子跳出局部Pareto前沿的能力。最后,将HBBMOPSO与自组织模糊神经网络预测模型和自组织控制器相结合,实现污水生化处理过程溶解氧和硝态氮设定值的动态寻优、智能决策和底层跟踪控制。利用国际基准仿真平台BSM1进行实验验证,结果表明所提HBBMOPSO方法在保证出水水质参数达标的前提下,能够有效降低污水处理过程的能耗。     

改进PSO算法在蒸汽动力系统多周期运行优化中的应用

为了满足石化企业工艺过程对蒸汽和电力不断变化的要求,实现企业降低成本、节能降耗的目的,必须保证蒸汽动力系统在最优的状态下运行。针对这一问题在以往研究的基础上提出了包括设备维护和启停费用的改进的混合整数线性规划模型,利用改进的PSO算法对其求解,并通过实例证明了利用该模型、使用改进的PSO算法能很快得到最优的方案,并节省了大量的运行成本。     

一种求解过程动态优化问题的生物地理学习粒子群算法

智能优化算法具有适用性广泛、全局搜索能力强等优点,近年来在动态优化中的应用逐渐增多。通过混合生物地理优化与粒子群优化,提出了生物地理学习粒子群（biogeography-based learning particle swarm optimization,BLPSO）算法,并用于动态优化问题的求解。BLPSO采用了新型的生物地理学习方式,该方式根据粒子“排名”,即粒子的优劣,以维度为单位构造学习粒子,提高了学习的效率。针对动态优化问题,首先通过控制向量参数化将其转化为非线性规划问题,然后采用BLPSO算法进行求解。最后,将BLPSO应用于非可微、多峰、多变量等典型动态优化问题的求解,计算结果表明BLPSO具有较好的搜索精度和收敛速度。     

群智能优化LSSVM最优聚丙烯熔融指数预报

提出了群智能优化AC_ICPSO(ant colony and immune clone particle swarm optimization)算法,融合蚁群算法与粒子群算法进行动态群体搜索,设计交叉算子和变异算子、群体多次编码、迭代选择等,来提高数据搜索的范围、精度和收敛的效率,避免早熟,降低算法的复杂度。然后利用AC_ICPSO方法对最小二乘支持向量机预报模型(LSSVM)进行参数寻优,得到最优的AC_ICPSO_LSSVM预报模型。以实际聚丙烯生产的熔融指数预报作为实例进行研究,结果表明所提出的AC_ICPSO_LSSVM方法有效,具有良好的预报精度。     

基于粒子群优化算法的球磨机制粉系统PID-ANN解耦控制器

球团厂钢球磨煤制粉系统是多变量强耦合、时滞、非线性以及生产工况变化大的复杂对象,其自动控制问题一直是控制界关注的热点。基于粒子群算法具有对整个参数空间进行高效并行搜索的特点以及PID神经网络的自调节和自适应特性,设计了具有PID结构的多变量自适应神经网络控制器。PID神经网络解耦控制方法被用来消除回路之间的耦合,神经网络连接权值由粒子群算法进行学习优化。仿真研究表明所建模型和所提控制方法具有较好的控制品质、良好的自适应解耦能力和自学习功能。该控制策略可在大范围内克服系统的非线性和强耦合问题,具有很高的工程实用价值。     

Vortex motion‐based particle swarm optimisation for energy consumption of alumina evaporation

The aim of the alumina evaporation process is to improve the concentration of sodium aluminate solution by evaporating the excess water contained in the solution. The evaporation is achieved using heat from steam. Since steam consumption is the major operating costs, in this paper, we investigate an operation optimisation problem for the evaporation process to minimise steam consumption subject to a constraint on the particular quality of the final sodium aluminate solution. This paper proposes a new particle swarm optimisation (PSO) algorithm based on vortex motion to solve this optimisation problem. We demonstrate the effectiveness of the PSO algorithm on benchmark functions. We then apply it to a real industrial evaporation process, where the optimal results show that the steam consumption is considerably reduced. © 2012 Canadian Society for Chemical Engineering     

基于新策略粒子群算法优化换热网络

换热网络综合优化是过程系统中最广泛研究的方向。尽管如此,MINLP的复杂性给粒子群算法的应用提供了广泛的空间。首先,提出两种不同机理的局部搜索策略来完善粒子群算法作为启发式算法局部搜索能力不强和精度不高的问题,使算法能更有利地接近全局最优的局部极值。其次,对含固定投资费用的算例,采用费用计算替换公式的策略,来避免迭代计算初期面积较小时因为固定投资费用权重较大而使算法陷入局部最优问题。最后用4个四股流算例分别从不同侧面说明以上两种策略的有效性,并都得到了该算例目前为止最好的局部极值。     

自适应混合粒子群约束优化算法及其在软测量模型参数估计中的应用

对于软测量模型参数估计问题, 针对传统梯度法求解非线性最小二乘模型时依赖初值、需要追加趋势分析进行验证和无法直接求解复杂问题的缺陷, 提出将参数估计化为约束优化问题, 使用混合优化算法求解的新思路。为此提出一种自适应混合粒子群约束优化算法(AHPSO-C)。在AHPSO-C算法中, 为平衡全局搜索(混沌粒子群)和局部搜索(内点法), 引入自适应内点法最大函数评价次数更新策略。对12个经典测试函数的仿真结果表明, AHPSO-C是求解约束优化问题的一种有效算法。将算法用于淤浆法高密度聚乙烯(HDPE)串级反应过程中熔融指数软测量模型参数估计, 验证了方法的可行性与优越性。     

基于全局优化改进混沌粒子群遗传算法的物料平衡数据校正

结合遗传算法（GA）和粒子群算法（PSO）的优点以及混沌运动的特性,提出了加入混沌扰动的混沌粒子群遗传算法（DCPSO-GA）,并使用5个高维非线性测试函数考察全局优化混合算法的性能。DCPSO-GA解决了在寻优搜索时出现的停滞现象,扩大了全局优化的搜索空间,丰富了粒子的多样性,且不需要函数梯度信息。测试结果证明,针对本文的5个测试函数DCPSO-GA能找到全局最优解,其收敛速度很快,大大减少了计算量。而且,经过与其他相关算法比较可知,当总的目标函数调用次数较接近或更少时,改进算法不论在计算精度还是收敛速度上,均有很大的提高。并将DCPSO-GA算法应用到重油裂解参数估计和预测中,测试结果证明,其提高了参数估计和预测的准确性,降低了误差,能有效找到全局最优解,收敛速度快,大大减少计算量。