PARETO FRONT CAPTURE USING DETERMINISTIC OPTIMIZATION METHODS IN MULTI-CRITERION AERODYNAMIC DESIGN

Deterministic optimization methods are combined with the Pareto front concept to solve multi-criterion design problems. The algorithm and the numerical implementation are applied to aerodynamic designs. Evolutionary algorithms （EAs） and the Pareto front concept are used to solve practical design problems in industry for its robustness in capturing convex, concave, discrete or discontinuous Pareto fronts of multi-objective optimization problems. However, the process is time-consuming. Therefore, deterministic optimization methods are introduced to capture the Pareto front, and the types of the captured Pareto front are explained. Numerical experiments show that the deterministic optimization method is a good alternative to EAs for capturing any convex and some concave Pareto fronts in multi-criterion aerodynamic optimization problems due to its efficiency.

Suboptimal reliable guaranteed cost control for continuous-time systems with multi-criterion constraints

The suboptimal reliable guaranteed cost control （RGCC） with multi-criterion constraints is investigated for a class of uncertain continuous-time systems with sensor faults. A fauk model in sensors, which considers outage or partial degradation of sensors, is adopted. The influence of the disturbance on the quadratic stability of the closed-loop systems is analyzed. The reliable state-feedback controller is developed by a linear matrix inequalities （LMIs） approach, to minimize the upper bound of a quadratic cost fimction under the conditions that all the closed-loop poles be placed in a specified disk, and that the prescribed level of H∞ disturbance attenuation and the upper bound constraints of control inputs＇ magnitudes be guaranteed. Thus, with the above muki-criterion constraints, the resulting closed-loop system can provide satisfactory stability, transient property, a disturbance rejection level and minimized quadratic cost performance despite possible sensor faults.

Multi-Criterion Optimal Design of Automotive Door Based on Metamodeling Technique and Genetic Algorithm

A method for optimizing automotive doors under multiple criteria involving the side impact, stiffness, natural frequency, and structure weight is presented. Metamodeling technique is employed to construct approximations to replace the high computational simulation models. The approximating functions for stiffness and natural frequency are constructed using Taylor series approximation. Three popular approximation techniques,i.e.polynomial response surface (PRS), stepwise regression (SR), and Kriging are studied on their accuracy in the construction of side impact functions. Uniform design is employed to sample the design space of the door impact analysis. The optimization problem is solved by a multi-objective genetic algorithm. It is found that SR technique is superior to PRS and Kriging techniques in terms of accuracy in this study. The numerical results demonstrate that the method successfully generates a well-spread Pareto optimal set. From this Pareto optimal set, decision makers can select the most suitable design according to the vehicle program and its application.

WiMA:Towards a Multi-Criterion Association in Software Defined Wi-Fi Networks

Despite the planned installation and operations of the traditional IEEE 802.11 networks,they still experience degraded performance due to the number of inefficiencies.One of the main reasons is the received signal strength indicator(RSSI)association problem,in which the user remains connected to the access point(AP)unless the RSSI becomes too weak.In this paper,we propose a multi-criterion association(WiMA)scheme based on software defined networking(SDN)in Wi-Fi networks.An association solution based on multi-criterion such as AP load,RSSI,and channel occupancy is proposed to satisfy the quality of service(QoS).SDNhaving an overall view of the network takes the association and reassociation decisions making the handoffs smooth in throughput performance.To implementWiMA extensive simulations runs are carried out on Mininet-NS3-Wi-Fi network simulator.The performance evaluation shows that the WiMA significantly reduces the average number of retransmissions by 5%–30%and enhances the throughput by 20%–50%,hence maintaining user fairness and accommodating more wireless devices and traffic load in the network,when compared to traditional client-driven(CD)approach and state of the art Wi-Balance approach.

On a Multi-criterion Problem of Planning Maritime Cargo Transportation

In this work, we consider a specific problem of optimal planning of maritime transportation of multiproduct cargo by ships of one （so-called ＂corporate strategy＂） or several （so-called ＂partially corporate strategy＂） companies： the core of the problem consists of the existence of the network of intermediate seaports （i.e., transitional seaports）, where for every ship arrived the cargo handling is done, and which are situated between the starting and the finishing seaports. In this work, there are mathematical models built from scratch in the form of multi-criterion optimization problem; according to the properties of the criteria and structure of the feasible solution set; are formulated different optimality conditions; are analysed different approaches for finding effective solutions （i.e., Pareto optimal solutions） and for check of the given solutions＇ effectiveness. In addition, in this work, there is considered and analysed well-known method of contraction of the Pareto boundary （goal attainment method of Gembicki）, then, it is used for reducing the built models to a one-criterion problem of linear programming.

Unsupervised construction of health indicator for rotating machinery via multi-criterion feature selection and attentive variational autoencoder

Health indicator(HI)construction is a crucial task in degradation evaluation and facilitates the prognostic and health management(PHM)of rotating machinery.Excluding interference from artificial labeling,the HI construction approaches in an unsupervised manner have attracted substantial attention.Nevertheless,current unsupervised methods generally struggle with two problems:(1)ignorance of both redundancy between features and global variability of features during the feature selection process;(2)inadequate utilization of information from different sampling moments.To tackle these problems,this work develops a novel unsupervised approach for HI construction that integrates multi-criterion feature selection and the Attentive Variational Autoencoder(Attentive VAE).Explicitly,a multi-criterion feature selection(Mc FS)algorithm together with an elaborately designed metric is proposed to determine a superior feature subset,considering the relevance,the redundancy,and the global variability of features simultaneously.Then,for the adequate utilization of the information from distinct sampling moments,a deep learning model named Attentive VAE is established.The Attentive VAE is solely fed with the selected features in the health state for model training and the HI is derived through the reconstruction error to reveal the degradation degree of machinery.Two case studies based on genuine experimental datasets are involved to quantitatively evaluate the superiority of the developed approach,demonstrating its superiority over other unsupervised methods for characterizing degradation processes.The effectiveness of both the Mc FS algorithm and the Attentive VAE is verified by ablation experiments,respectively.

Couplings in Multi-criterion Aerodynamic Optimization Problems Using Adjoint Methods and Game Strategies

The tighten couplings of game strategies with adjoint methods for multi-criterion aerodynamic design optimization are ad-dressed. Its numerical implementation is also described in details. In cooperative game,adjoint methods are coupled in parallel to compute Pareto front collaboratively. Conversely in a Nash game,adjoint methods are coupled in each player s decision making to achieve Nash equilibrium competitively. In Stackelberg game,adjoint methods used by players are nested hierarchically through incomp...

Multi-criterion system optimization model for urban multimodal traffic network

In this paper,computational models of environmental pollution and energy consumption of urban multimodal traffic network are proposed according to pertinent research and a multi-objective programming model is then developed to formulate optimization problem for such a system.Simultaneously,the main factors,such as travel time,pricing and convenience which influence travelers' choice behaviors are all considered and a combined assignment model is proposed to simulate travelers' mode and route choices.A bi-level programming model,in which the multi-objective optimization model is treated as the upper-level problem and the combined assignment model is processed as the lower-level problem,is then presented to solve multi-criterion system optimization problem for urban multimodal traffic network.The solution algorithms of the proposed models are also presented.Finally,the model and its algorithms are illustrated through a simple numerical example.

Distributed optimization using virtual and real game strategies for multi-criterion aerodynamic design

This paper introduces the virtual and real game concepts to investigate multi-criterion optimization for optimum shape design in aerodynamics. The constrained adjoint meth- odology is used as the basic optimizer. Furthermore, the above is combined with the vir- tual and real game strategies to treat single-point/multi-point airfoil optimization. In a symmetric Nash Game, each optimizer attempts to optimize one’s own target with ex- change of symmetric information with others. A Nash equilibrium is just the compromised solution among the multiple criteria. Several kinds of airfoil splitting and design cases are shown for the utility of virtual and real game strategies in aerodynamic design. Successful design results confirm the validity and efficiency of the present design method.

SSA-MLP模型在岩质边坡稳定性预测中的应用

岩质边坡的力学参数量化及稳定性分析对岩质边坡灾害的防治具有重要意义。Hoek-Brown(H B)准则是一种用于确定岩体力学参数的经典方法,能反映出边坡岩体变形和位移的非线性破坏特征。在此基础上,首先,提出一种麻雀搜索算法(Sparrow Search Algorithm,SSA)改进多层感知器(Multi-Layer Perceptron,MLP)的神经网络模型,并用于边坡稳定性预测、指标敏感性分析及参数反演。其次,将收集的1085组岩质边坡的几何参数和H B准则参数等作为输入变量,极限平衡理论Bishop法求解的安全系数作为输出变量,对SSA MLP模型进行训练学习和性能评估。最后,将该模型运用于25个边坡实例,验证模型的有效性。结果显示,该模型收敛速度快、精度高,为边坡稳定性分析和参数量化提供了一种新思路。

基于电压跌落安全需求的直流微电网多端协同支撑技术

量化直流电压稳定判据,在多端换流器协同下完成暂态电压支撑是直流微电网稳定运行的关键。根据直流微电网运行特性,分析功率波动、直流故障下系统运行点变化轨迹,获得系统具备暂态恢复能力下的直流电压跌落安全区间。基于直流微电网须遵循的等电量原则,根据电压跌落安全区间,计算系统允许放电电量,从而得到直流微电网量化稳定判据。在此基础上,利用交流主网、蓄电池、风机侧换流器的控制潜力,提出直流微电网的多端协同支撑技术,通过释放附加电量满足直流微电网在不同扰动下差异化的电压跌落安全需求,提升系统的暂态稳定裕度。搭建5端直流微电网仿真系统,验证所提控制策略对于直流电压的暂态支撑效果。

基于决策者偏好构造的概率语言型Z-number多属性决策方法

由于决策环境的日益复杂,专家评价信息往往会包含专家偏好,因此有必要对专家偏好进行识别与处理。针对这一情况,提出了基于决策者偏好构造的概率语言型Z-number并应用于多属性决策问题中。首先,针对专家偏好未知的情况,提出了一种基于修正的平均得分距离,由专家评价信息确定专家偏好。然后,依据得到的专家偏好,利用GOWPA算子与最大熵生成概率语言型Z-number。同时,获得专家评价信息的可信度,由专家可信度聚合专家信息并采用TODIM方法对方案进行排序。最后,在国家大力提倡节能减排的背景下,将所提方法应用于国家绿色矿山试点企业的选取上,验证了所提方法的有效性和可行性。

某空间机构内啮合复合行星齿轮传动系统设计与优化研究

齿轮传动系统是空间站等空间机构开展设备安装与维修、舱段转位与对接等任务的核心机械装置。空间机构齿轮传动系统需满足大扭矩、高效率、长寿命、轻量化等需求,对系统设计和优化方法提出了更高要求。针对某空间机构传动系统的设计需求,提出了“两级内啮合齿轮+三级2K-H行星齿轮”的复合行星齿轮传动系统构型方案,完成了传动系统各级齿轮的结构设计,并提出了基于多准则决策的空间机构齿轮传动系统多目标优化模型,实现多达49个结构参数的快速求解。相比优化前的传动设计方案,优化后的齿轮传动系统减重5.02%,安全系数提升8.01%,为新一代空间机械臂关节等空间机构装备研发提供了支撑。

基于故障距离的微电网系统线路保护方法分析

配置快速、有效的线路保护方案是电源分散就地技术发展的关键。针对微电网线路保护问题,提出基于故障距离的微电网系统线路保护方法。通过构建包含故障信号的等效电气量方程,并推导出故障距离的解析表达式,确立1种多端线路保护的判据,再将此方法扩展到含有多个电源接入的电力网络中。算例分析表明,所提判据方法可靠性较高,可有效避免因保护装置整定值设定不合理而导致的误动或拒动问题。

多模态多分类器融合模型预测放射性口腔黏膜炎的性能

目的评估不同放射性口腔黏膜炎(RIOM)预测模型的性能,对比分析分层多模态多分类器融合(H-MCF)模型的有效性。方法回顾性收集2022年9月~2023年2月在广州医科大学附属肿瘤医院接受观察和治疗的198例放射性口腔黏膜炎局部晚期鼻咽癌患者的数据。基于口腔放射剂量-体积参数与鼻咽癌相关的临床特征,针对不同特征选择算法和分类器两两组合得到基础分类模型。我们使用基于多准则决策的多分类器融合模型(MCF)和它的变体——H-MCF模型对基础分类模型进行融合。通过对各个模态的基础分类模型与MCF模型的性能、多模态的基础模型和MCF模型以及H-MCF模型的性能、单模态与多模态模型的性能、H-MCF与MCF以及其他集成分类器的性能进行分析比较,并通过ROC曲线下面积(AUC)、准确率(ACC)、灵敏度(SEN)和特异度(SPE)4种评价指标来评估模型的泛化性能,分析RIOM预测模型有效性。结果结合多模态特征后,H-MCF模型在预测严重RIOM方面达到了最高的准确性(AUC=0.883,ACC=0.850,SEN=0.933,SPE=0.800)。结论相较于单个分类器的预测结果,结合临床和剂量两种模态的多分类器融合算法在预测严重RIOM发病率方面表现更优。

低雷诺数翼型多目标优化设计研究

为消除低雷诺数下对称翼型小迎角范围内的气动力非线性现象,基于高斯过程回归模型及NSGA‑Ⅱ多目标遗传算法并耦合RANS的数值方法建立了低雷诺数翼型优化设计程序,以多个状态下小迎角范围内的平均巡航因子为目标函数开展翼型优化设计。结果表明:优化翼型具有明显的几何凸起,有效改变了分离泡的位置和类型,进而显著改善翼型气动特性,Re=4×104时小迎角范围内的升力系数非线性现象基本消失。进一步分析发现优化翼型的凸起增强了壁面的对流影响和来流抵抗逆压梯度的能力,进而改善了翼型的气动特性。

基于多对数螺旋曲线破坏机构的富水软岩隧道开挖面支护应力分析

为了避免对胡克-布朗准则进行任何线性化处理,将广义胡克-布朗(generalized Hoek-Brown, GHB)强度准则直接引入到开挖面稳定性分析中,基于极限分析理论构建了富水地层浅埋隧道开挖面多对数螺旋线破坏机构。在考虑孔隙水压力的作用下,推导出能量平衡方程,得到开挖面极限支护应力表达式,并且通过程序优化得到支护应力最优上限解。在此基础上,分析了GHB强度参数、孔隙水压力系数、岩土体重度以及隧道埋深比对开挖面稳定性的影响。最后通过归一化处理得到不同情况下的开挖面极限支护应力稳定性图表,通过该图表能够快速便捷地获得富水地层隧道开挖面维持稳定性所需的极限支护应力,为盾构隧道初步设计提供理论依据。

面向组网雷达干扰任务的多机伴随式编队航迹预规划方法

针对多目标突防组网雷达系统场景,为有效提高干扰效果以及突防成功率,编队航迹规划尤为重要。因此,首先构建航迹规划模型,从飞行器自身约束、航迹安全性、机间协调以及任务完成效果4个方面出发,结合多机伴随式编队及其所处环境特点,提出较为完备的航迹规划准则,形成一个新的整体目标函数;其次,为有效描述每架飞机的机动特性以及伴飞干扰机与目标飞机间的联系,提高算法搜索能力,提出基于多球面矢量(multi-spherical vector-based,MS)方法;为进一步提高算法的探索和开发能力,提出多面球矢量逐航迹点学习混合粒子群优化(multi-spherical vector-based hybrid particle swarm optimization with track point by track point learning,TLHPSO)算法,并将两者相结合,形成基于多面球矢量的逐航迹点学习混合粒子群优化(MS-based hybrid particle swarm optimization with track point by track point learning,MS-TLHPSO)航迹规划方法;最后,构建相应仿真场景进行验证。对比结果表明,MS方法以及TLHPSO优化算法在寻优能力上具有明显优势;同时,所提算法在不同初始场景下最优解的平均值均优于其他算法,充分说明所提算法能够在保证稳定性的前提下规划具有更高可信度的编队航迹。

基于多模型推断估计池塘混养条件下凡纳滨对虾的生长参数

对混养草鱼(Ctenopharyngodon idella)池塘中的凡纳滨对虾(Litopenaeus vannamei)进行了5个月的周期性采样,应用Brody、Gompertz、Logistic、Richards、von Bertalanffy和广义von Bertalanffy等6种备选生长模型对其生长过程进行了拟合研究,利用最大似然法估算了以上6种模型的生长参数,并综合应用赤池信息准则(AIC)和贝叶斯信息准则(BIC)对模型拟合度进行检验。研究结果表明凡纳滨对虾体长-体质量回归关系为W=0.0106L^(3.066)(R^(2)=0.987),经t-检验b值显著大于3,凡纳滨对虾表现为正异速生长(b>3)。AIC和BIC检验结果表明,von Bertalanffy、广义von Bertalanffy和Brody生长模型因不受数据支持而被舍弃,Richards、Logistic和Gompertz模型均受到了数据的较强支持,但没有任何一个模型的Akaike权重(ω_(i))大于95%,可被单独用来描述凡纳滨对虾体长与日龄的生长关系。因此应用多模型推断对这3种模型的生长参数进行加权平均处理,据此求出混养条件下凡纳滨对虾的多模型平均体长渐进值为13.741 cm(渐进体质量为32.694 g),作为描述凡纳滨对虾生长参数的稳健估计值。

基于LMD与AO-PNN的中介轴承故障诊断方法

针对航空发动机中介轴承受噪声干扰大、传递路径复杂导致采用传统方法难以进行故障诊断的问题,提出了一种基于局部均值分解(LMD)与相关系数-能量比-峭度准则、结合天鹰座优化算法(AO)优化概率神经网络(PNN)的中介轴承故障诊断方法。使用LMD对传感器采集的振动信号进行分解;利用相关系数-能量比-峭度准则判决筛选分解得到的PF分量,重构筛选后的信号;计算重构信号的多尺度排列熵(MPE),以构建特征向量;通过AO优化的PNN的平滑因子,将优化后的神经网络用于中介轴承的故障诊断。基于中介轴承故障试验数据对诊断结果进行了分析,结果表明:提出的方法可以有效诊断高背景噪声、复杂路径干扰下的航空发动机中介轴承的典型故障,与粒子群优化的概率神经网络方法(PSO-PNN)和传统的PNN方法相比,其诊断准确率分别提高了3.875%和8.125%,具有较好的全局收敛性和计算鲁棒性。