CEE-Gr:A Global Router with Performance Optimization Under Multi-Constraints

A global routing algorithm with performance optimization under multi constraints is proposed,which studies RLC coupling noise,timing performance,and routability simultaneously at global routing level.The algorithm is implemented and the global router is called CEE Gr.The CEE Gr is tested on MCNC benchmarks and the experimental results are promising.

A LEVEL SET METHOD FOR STRUCTURAL TOPOLOGY OPTIMIZATION WITH MULTI-CONSTRAINTS AND MULTI-MATERIALS

Combining the vector level set model,the shape sensitivity analysis theory with the gradient projection technique,a level set method for topology optimization with multi-constraints and multi-materials is presented in this paper.The method implicitly describes structural material in- terfaces by the vector level set and achieves the optimal shape and topology through the continuous evolution of the material interfaces in the structure.In order to increase computational efficiency for a fast convergence,an appropriate nonlinear speed mapping is established in the tangential space of the active constraints.Meanwhile,in order to overcome the numerical instability of general topology opti- mization problems,the regularization with the mean curvature flow is utilized to maintain the interface smoothness during the optimization process.The numerical examples demonstrate that the approach possesses a good flexibility in handling topological changes and gives an interface representation in a high fidelity,compared with other methods based on explicit boundary variations in the literature.

Critical chain construction with multi-resource constraints based on portfolio technology in South-to-North Water Diversion Project

Recently, the critical chain study has become a hot issue in the project management research field. The construction of the critical chain with multi-resource constraints is a new research subject. According to the system analysis theory and project portfolio theory, this paper discusses the creation of project portfolios based on the similarity principle and gives the definition of priority in multi-resource allocation based on quantitative analysis. A model with multi-resource constraints, which can be applied to the critical chain construction of the A-bid section in the South-to-North Water Diversion Project, was proposed. Contrast analysis with the comprehensive treatment construction method and aggressive treatment construction method was carried out. This paper also makes suggestions for further research directions and subjects, which will be useful in improving the theories in relevant research fields.

Explicit Correlation Model of Multi-source Constraints for Re-design Parts with Complex Curved Surface

In precision machining of complex curved surface parts with high performance, geometry accuracy is not the only constraint, but the performance should also be met. Performance of this kind of parts is closely related to the geometrical and physical parameters, so the final actual size and shape are affected by multiple source constraints, such as geometry, physics, and performance. These parts are rather difficult to be manufactured and new manufacturing method according to performance requirement is urgently needed. Based on performance and manufacturing requirements for complex curved surface parts, a new classification method is proposed, which divided the complex curved surface parts into two categories： surface re-design complex curved surface parts with multi-source constraints（PRCS） and surface unique complex curved surface parts with pure geometric constraints（PUCS）. A correlation model is constructed between the performance and multi-source constraints for PRCS, which reveals the correlation between the performance and multi-source constraints. A re-design method is also developed. Through solving the correlation model of the typical paws performance-associated surface, the mapping relation between the performance-associated surface and the related removal amount is obtained. The explicit correlation model and the method for the corresponding related removal amount of the performance-associated surface are built based on the classification of surface re-design complex curved surface parts with multi-source constraints. Research results have been used in the actual processing of the typical parts such as radome, common bottom components, nozzle, et al., which shows improved efficiency and accuracy of the precision machining for the surface re-design parts with complex curved surface.

Differential Evolution Algorithm Based on Ensemble of Constraint Handling Techniques and Multi-Population Framework

Aimed at improving the insufficient search ability of constraint differential evolution with single constraint handling technique when solving complex optimization problem, this paper proposes a constraint differential evolution algorithm?based on ensemble of constraint handling techniques and multi-population?framework, called ECMPDE. First, handling three improved variants of differential evolution algorithms are dynamically matched with two constraint handling techniques through the constraint allocation mechanism. Each combination includes three variants with corresponding constraint handling technique?and these combinations are in the set. Second, the population is divided into three smaller subpopulations and one larger reward subpopulation. Then a combination with three constraint algorithms is randomly selected from the set, and the three constraint algorithms are run in three sub-populations respectively. According to the improvement of fitness value, the optimal constraint?algorithm is selected to run on the reward sub-population, which can share?information and close cooperation among populations. In order to verify the effectiveness of the proposed algorithm, 12 standard constraint optimization problems?and 10 engineering constraint optimization problems are tested. The experimental results show that ECMPDE is an effective algorithm for solving constraint optimization problems.

Multi-model predictive control with local constraints based on model switching

Because model switching system is a typical form of Takagi-Sugeno（T-S） model which is an universal approximator of continuous nonlinear systems, we describe the model switching system as mixed logical dynamical （MLD） system and use it in model predictive control （MPC） in this paper. Considering that each local model is only valid in each local region,we add local constraints to local models. The stability of proposed multi-model predictive control （MMPC） algorithm is analyzed, and the performance of MMPC is also demonstrated on an inulti-multi-output（MIMO） simulated pH neutralization process.

Generation of optimal persistent formations for heterogeneous multi-agent systems with a leader constraint

In this study, we consider the generation of optimal persistent formations for heterogeneous multi-agent systems, with the leader constraint that only specific agents can act as leaders. We analyze three modes to control the optimal persistent formations in two-dimensional space, thereby establishing a model for their constrained generation. Then, we propose an algorithm for generating the optimal persistent formation for heterogeneous multi-agent systems with a leader constraint （LC-HMAS-OPFGA）, which is the exact solution algorithm of the model, and we theoretically prove its validity. This algorithm includes two kernel sub-algorithms, which are optimal persistent graph generating algorithm based on a minimum cost arborescence and the shortest path （MCA-SP-OPGGA）, and the optimal persistent graph adjusting algorithm based on the shortest path （SP-OPGAA）. Under a given agent formation shape and leader constraint, LC-HMAS-OPFGA first generates the network topology and its optimal rigid graph corresponding to this formation shape. Then, LC-HMAS- OPFGA uses MCA-SP-OPGGA to direct the optimal rigid graph to generate the optimal persistent graph. Finally, LC- HMAS-OPFGA uses SP-OPGAA to adjust the optimal persistent graph until it satisfies the leader constraint. We also demonstrate the algorithm, LC-HMAS-OPFGA, with an example and verify its effectiveness.

Surface Deformation with Multi-constraints Based on the Minimum Energy

A method for deformation of 3D point clouds models was proposed with multi-constraints including arc-length constraints and multi-points position constraints. The energy function was built for the polyline which had been converted from the curve. Based on the minimum energy curve method, the curve on the mesh was deformed. The test results show that the proposed method has good performance. Compared with the other method,shape preserving of the curve is better. Finally,this method is used for the deformation of the 3D mannequin model. Circumference changes of the mannequin model can be reflected by the arc-length change in the size of the cross section.

Cooperative Multi-Agent Reinforcement Learning with Constraint-Reduced DCOP

Cooperative multi-agent reinforcement learning（ MARL） is an important topic in the field of artificial intelligence,in which distributed constraint optimization（ DCOP） algorithms have been widely used to coordinate the actions of multiple agents. However,dense communication among agents affects the practicability of DCOP algorithms. In this paper,we propose a novel DCOP algorithm dealing with the previous DCOP algorithms＇ communication problem by reducing constraints.The contributions of this paper are primarily threefold：（1） It is proved that removing constraints can effectively reduce the communication burden of DCOP algorithms.（2） An criterion is provided to identify insignificant constraints whose elimination doesn＇t have a great impact on the performance of the whole system.（3） A constraint-reduced DCOP algorithm is proposed by adopting a variant of spectral clustering algorithm to detect and eliminate the insignificant constraints. Our algorithm reduces the communication burdern of the benchmark DCOP algorithm while keeping its overall performance unaffected. The performance of constraint-reduced DCOP algorithm is evaluated on four configurations of cooperative sensor networks. The effectiveness of communication reduction is also verified by comparisons between the constraint-reduced DCOP and the benchmark DCOP.

计及多目标约束的成品卷烟销售订单配送调度优化

针对制造业供应链中的跨地域多运输中心大数据订单物流调度问题,对涉及多约束条件和多目标的异构运输车辆调度问题展开研究。根据烟草制造工业的实际运输任务需求,考虑制造订单安排、车辆准运证限制、运输车辆物理限制、运输时间窗、排产出库时间等实际约束,搭建了以包括车辆等待卷烟出库的最短运输时间,最低运输费用及最高车辆利用率为优化目标的成品烟物流车辆跨地域调度优化模型。设计了基于差分进化的混合多元宇宙算法(DE-MVO)用以解决该模型。最后基于某烟草工业企业的实际生产订单数据进行实验,将所设计的算法分别与粒子群(PSO)算法、差分进化(DE)算法、鲸鱼优化(WOA)算法、遗传优化(GA)算法、乌鸦搜索(CSA)算法、天鹰座优化器(AO)、多元宇宙(MVO)等前沿算法进行横向对比,对比结果表明所设计的算法在全局搜索能力、寻优精度等方面更优于其他算法,进一步表明所建立的模型在求解制造业供应链运输车辆调度问题中具有可行性。

基于分层约束强化学习的综合能源多微网系统优化调度

构建多微网系统是消纳可再生能源、提升电网稳定性的有效方式。通过各微网的协调调度,可有效提升微网的运行效益以及可再生能源的消纳水平。现有多微网优化问题场景多元,变量众多,再加上源荷不确定性及多微网主体的数据隐私保护等问题,为模型的高效求解带来了巨大挑战。为此,该文提出了一种分层约束强化学习优化方法。首先,构建了多微网分层强化学习优化框架,上层由智能体给出各微网储能优化策略和微网间功率交互策略;下层各微网以上层策略为约束,基于自身状态信息采用数学规划法对各微网内部的分布式电源出力进行自治优化。通过分层架构,减小通信压力,保护微网内部数据隐私,充分发挥强化学习对源荷不确定性的自适应能力,大幅提升了模型求解速度,并有效兼顾了数学规划法的求解精度。此外,将拉格朗日乘子法与传统强化学习方法相结合,提出一种约束强化学习求解方法,有效地解决了传统强化学习方法难以处理的约束越限问题。最后通过算例验证了该方法的有效性和优势。

Diagnosis and Resolution of Infeasibility in the Constraint Method for Solving Multi Objective Linear Programming Problems

In this paper we discuss about infeasibility diagnosis and infeasibility resolution, when the constraint method is used for solving multi objective linear programming problems. We propose an algorithm for resolution of infeasibility, which is a combination of interactive, weighting and constraint methods.Numerical examples are provided to illustrate the techniques developed.

融合设备状态和产品要求的电解铜箔制造工艺参数优化

针对当前电解铜箔制造工艺参数的优化仅考虑产品要求的问题,提出了一种融合设备状态和产品要求的电解铜箔制造工艺参数优化方法。首先,为辨识铜箔制造中工艺、设备与产品参数之间的非线性关系,建立了基于神经网络的“工艺-设备”与产品参数的映射模型,并通过构建当前生产模式与历史生产模式的相似性度量方法,确定当前工艺参数的优化初始点;然后,基于所建立的映射模型、设备状态和产品要求构建多目标约束函数,并利用遗传算法优化工艺参数,获得同时满足设备状态和产品要求的最佳工艺参数。企业的生产验证表明:相较于仅考虑产品要求的工艺优化方法,所提方法优化后的工艺参数能在相同的设备状态下获得质量更好的产品,在常温/高温抗拉强度、常温/高温延伸率和树脂剥离强度这5项关键指标上,分别提升了6.1 MPa、6.1 MPa、1.9%、0.7%和0.22 N/mm,生箔制备和表面处理的生产效率也提高了2.65%和7.5%,能够有效保证并提高铜箔的质量和生产效益。

基于多策略改进鲸鱼算法的船舶推力分配方法

船舶动力定位推力分配求解是一种复杂的多约束多目标优化问题,鲸鱼优化算法(WOA)处理该类问题较传统推力分配算法计算更简便、精度更高,但以牺牲时间为代价,需解决实时可靠收敛问题。针对上述问题,提出一种改进鲸鱼推力分配算法,通过采用等式约束项的特解系数为个体变量,降低搜索空间维度和范围,为算法获取可靠解和快速收敛奠定基础;通过空间区块化和“精英”筛选法初始化种群,增加初始种群分散性和“精英”个体数量,提高寻优效率和收敛稳定性;引入动态种群边界获取合理的搜索空间,提高算法收敛速度;最后运用类梯度法进行种群更新,模拟区域遍历搜寻,以较少迭代次数获取可靠解。以CybershipⅢ船模为对象进行仿真验证,结果表明该算法可行,具有较好的实时性和稳定收敛性,提高了推力分配性能。

基于CutMix数据增强与多约束损失函数的YOLOv7盾构隧道渗漏水检测

盾构隧道强度投影图像中渗漏水尺寸不一致且像素占比偏小,现有目标检测模型的关键特征学习能力较弱,存在渗漏水病害目标检测精度偏低的问题。本文提出了基于CutMix数据增强与多约束损失函数的改进YOLOv7盾构隧道渗漏水检测方法。首先采用镶嵌CutMix方法对隧道图像进行数据增强,将多张不同的训练样本进行随机裁剪,拼接融合成具有综合特征的新样本;然后以YOLOv7网络为骨架结构,引入高效通道注意力模块,提高渗漏水关键特征的自主学习与表达能力;最后引入多约束几何条件的损失函数,提高预测框几何形状的精度,从而提升模型预测精度。在光线良好、光线不佳和存在遮挡等复杂环境情况下,选取Fast R-CNN、SSD、YOLOv5、YOLOv7这4种算法进行对比,试验表明,本文算法渗漏水检测精度达85.90%,平均精度比同类算法分别提高5.55%、8.89%、3.93%、2.75%,具有较高的稳健性和泛化能力。

基于平均一致协议的分布式自适应多智能体聚集控制

分布式聚集控制问题是多智能体协同控制中的一个重要问题。由于智能体的可移动性和感知能力有限,传统的分布式聚集算法难以保证连通性,从而聚集成多个簇群。此外,去中心化的大规模聚集控制给获取全局聚集点带来了巨大的挑战。针对连通性保护问题,基于平均一致协议与约束集,提出了一个带有连通性约束的多智能体聚集协议(Multi-Agent Rendezvous Protocol with Connectivity constraints,MARP-CC)。然后针对聚集点无法预测的问题,提出了位置合成(Location Synthesis Strategy,LSS)和位置重定向(Location Redirection Strategy,LRS)两种控制策略。智能体根据当前连通情况,自适应选择最优的控制策略进行迭代。结合这两种控制策略,提出了带连通性约束的分布式自适应多智能体聚集算法(Distributed Adaptive Multi-Agent Rendezvous algorithm with Connectivity Constraints,DAMAR-CC)。对算法的收敛性和连通性进行了分析,并通过大量的仿真说明了DAMAR-CC可以引导智能体稳定地聚集在初始拓扑的几何中心。

面向水土生态系统服务协同的土地利用模拟

针对水土生态系统服务功能的相关成果难以应用到实际土地利用规划中的问题。本研究以黄土高原为研究区,使用K-means聚类算法在小流域尺度上划分土地利用结构,结合InVEST模型探究土地利用结构对水土生态系统服务功能的影响。以阳泉市为典型区,设置多情景,结合多目标优化算法和PLUS模型进行未来土地利用模拟。结果表明:在小流域尺度上,黄土高原的土地利用结构可以划分为6类。农地结构、林地结构、农地-草地结构、草地结构、农地-林地-草地结构、裸地结构,最优土地利用占比分别是0.6—0.7,0.5,0.45/0.4(农地/草地),0.75—0.85,0.15/0.4/0.25—0.35(农地/林地/草地)。各土地利用结构中生态系统服务之间的约束曲线基本上是驼峰型、凹波、半凹波和凸波。从未来的土地利用数量上看,3种情景下水域和灌木面积均处于较低水平。其中,对照情景中城建用地面积最大,基准情景次之,协同情景最小。3种情景下,阳泉市均以城市建设为中心,向外依次为耕地、草地和林地的层次分布特征。土地利用的发展过程中各种土地类型之间存在替代规律,即城市建设倾向于取代耕地和草地,草地则可能取代林地和耕地,而林地则可能与草地发生替代关系。研究成果对黄土高原土地利用规划有重要参考价值。

基于特征约束和自适应损失平衡的机器遗忘方法

随着数字化进程的加速推进,数据要素已成为现代社会运转的核心驱动力.由于深度学习模型训练需要大量数据作为输入,其数据隐私保护问题也愈发重要.机器遗忘技术使模型能够删除特定数据的影响,同时保持对剩余数据的泛化性能,为深度学习模型中的数据要素安全保护提供了有效的解决方案.现有的机器遗忘方法主要分为精确遗忘和近似遗忘2类,但前者需要干预模型原始训练流程,后者则在遗忘效果和模型泛化能力之间难以找到平衡点.为此,提出了一种基于特征约束和自适应损失平衡的近似遗忘框架.首先,对于“遗忘”过程,使用同样未经过遗忘样本训练的随机模型作为教师来引导遗忘模型的特征输出,实现模型对数据要素在特征层面的遗忘.然后,采用少量剩余数据进行微调训练,来“恢复”模型对于其他数据的泛化性能.将上述机器遗忘框架视为一个多任务优化问题,在“遗忘”和“恢复”2个任务中引入自适应损失平衡,实现任务的稳步训练.以卷积神经网络模型为例,在3个公开数据集上对比了UNSIR等多种基线方法,实验结果表明,该方法构建的遗忘模型不仅保障了机器遗忘效果,在剩余数据的准确率、时间开销、预测结果分布等指标上优于同类方法,更加有效地保护了模型的泛化性能.

基于多源约束自适应视觉SLAM关键帧选取研究

该文针对现有关键帧选择方法在复杂场景下的稳定性和适应性方面不足问题,提出一种多源约束的自适应视觉SLAM关键帧选取方法。该算法基于相机几何测量原理,设计自适应阈值进行关键帧选取策略;针对复杂环境下的剧烈运动情况,设计基于IMU的实时状态检测机制和熵函数约束标准,进一步提高关键帧选取的稳定性和适应性。在EuRoC数据集和TUM数据集上对该方法进行定性和定量评估。在单目惯性和立体惯性模式下,将估计轨迹与参考轨迹进行对比,以绝对轨迹误差(absolute trajectory error,ATE)、关键帧数量和运行时间作为评判指标,并与ORB-SLAM3方法进行比较。结果显示,提出的方法可显著提高视觉SLAM在复杂环境下的定位精度和稳定性。

A Penalty Function Algorithm with Objective Parameters and Constraint Penalty Parameter for Multi-Objective Programming

In this paper, we present an algorithm to solve the inequality constrained multi-objective programming (MP) by using a penalty function with objective parameters and constraint penalty parameter. First, the penalty function with objective parameters and constraint penalty parameter for MP and the corresponding unconstraint penalty optimization problem (UPOP) is defined. Under some conditions, a Pareto efficient solution (or a weakly-efficient solution) to UPOP is proved to be a Pareto efficient solution (or a weakly-efficient solution) to MP. The penalty function is proved to be exact under a stable condition. Then, we design an algorithm to solve MP and prove its convergence. Finally, numerical examples show that the algorithm may help decision makers to find a satisfactory solution to MP.