最优化凸分组在目标检测中的应用研究

针对不同场景下静态图像中单目标的检测问题,结合自然界各个目标特有的凸属性特点,提出了一种基于最优化凸分组的目标检测方法。比较系统地论述了最优化凸分组的基本原理,介绍了详细的实现过程,主要包括Canny边缘检测参数的设置、基于边缘点的线段拟合、凸分组中凸多边形的构造以及最优化凸多边形的判定。实验结果表明,该方法对任意场景下的单目标检出率和检测准确性良好,结合目标凸属性的最优化判定方式具有检出速度快,且不受机器学习中的样本数据影响的特点,具有很好的普遍适应性。     

再制造成本随机分布的生产优化问题研究

研究单一厂商制造/再制造集成系统的两期生产优化问题.首先,在回收率一定的条件下,建立回收产品的再制造成本与再制造率之间的函数关系;然后,建立以追求利润最大化为目标的模型,验证了该模型为凸规划,给出了K-T条件表达式,并分析了伽马分布条件下解的特征及其临界条件;最后,通过算例对该模型的性质和规律作进一步分析.     

基于改进粒子群算法的无人机路径规划

针对传统粒子群算法PSO求解无人机路径规划问题时存在极易陷入局部最优的问题,在PSO算法中引入细菌觅食算法BFO的趋化操作、迁徙操作,以提高其寻优能力。首先根据无人机飞行环境建立三维高程环境模型,并使用路径长度代价、障碍危险代价和航迹高程代价来构造适应度函数;然后在分析了粒子群算法和细菌觅食算法原理及特点的基础上,给出了算法的改进方法及其具体流程。最后,通过Matlab仿真验证表明：混合算法有效改善了粒子群算法的缺陷,在进行无人机路径规划时,相比于传统PSO算法,混合算法寻优精度和稳定性有明显改善。     

从局部极小到全局最优

所有控制决策问题本质上均可归结为优化问题,但大部分存在多极小,因此如何摆脱局部极小以实现全局最优一直是理论界和工程界关注的热点课题。文章总结了若干全局优化技术的机制和特点,包括模拟退火、进化计算、禁忌搜索、变邻域搜索、噪声方法、巢分区、混沌搜索、隧道方法、平滑技术、混合算法等,力求为优化研究人员了解全局优化技术和开发高效算法提供指导。     

An interval algorithm for multi-objective optimization

This paper presents an interval algorithm for solving multi-objective optimization problems. Similar to other interval optimization techniques, [see Hansen and Walster (2004)], the interval algorithm presented here is guaranteed to capture all solutions, namely all points on the Pareto front. This algorithm is a hybrid method consisting of local gradient-based and global direct comparison components. A series of example problems covering convex, nonconvex, and multimodal Pareto fronts is used to demonstrate the method.     

A Fully Distributed Hybrid Control Framework For Non-Differentiable Multi-Agent Optimization

This paper develops a fully distributed hybrid control framework for distributed constrained optimization problems. The individual cost functions are non-differentiable and convex. Based on hybrid dynamical systems, we present a distributed state-dependent hybrid design to improve the transient performance of distributed primal-dual first-order optimization methods. The proposed framework consists of a distributed constrained continuous-time mapping in the form of a differential inclusion and a distributed discrete-time mapping triggered by the satisfaction of local jump set. With the semistability theory of hybrid dynamical systems, the paper proves that the hybrid control algorithm converges to one optimal solution instead of oscillating among different solutions. Numerical simulations illustrate better transient performance of the proposed hybrid algorithm compared with the results of the existing continuous-time algorithms.      

LSF参数量化方法及其优化

文章提出了一种多级量化LSF参数的方法,并进一步提出了优化量化的方法以及性能分析。优化算法减少了矢量量化算法的运算量和运算时间。加快了矢量量化的编码速度,且对矢量量化的失真情况没有影响。     

Learning Convex Optimization Models

A convex optimization model predicts an output from an input by solving a convex optimization problem. The class of convex optimization models is large, and includes as special cases many well-known models like linear and logistic regression. We propose a heuristic for learning the parameters in a convex optimization model given a dataset of input-output pairs, using recently developed methods for differentiating the solution of a convex optimization problem with respect to its parameters. We describe three general classes of convex optimization models, maximum a posteriori (MAP) models, utility maximization models, and agent models, and present a numerical experiment for each.      

有向网络分布式优化的Barzilai-Borwein梯度跟踪方法

本文研究有向网络上的分布式优化问题, 其全局目标函数是网络上所有光滑强凸局部目标函数的平均值.受Barzilai-Borwein步长改善梯度方法表现的启发, 本文提出了一种分布式Barzilai-Borwein梯度跟踪方法. 与文献中使用固定步长的分布式梯度算法不同, 所提出的方法中每个智能体利用其局部梯度信息自动地计算其步长. 通过同时使用行随机和列随机权重矩阵, 该方法避免了由特征向量估计引起的计算和通信. 当目标函数是光滑和强凸函数时, 本文证明了该算法产生的迭代序列可以线性地收敛到最优解. 对分布式逻辑回归问题的仿真结果验证了所提出的算法比使用固定步长的分布式梯度算法表现更好     

A hybrid biogeography-based optimization algorithm for job shop scheduling problem

In this paper, a hybrid biogeography-based optimization (HBBO) algorithm has been proposed for the job-shop scheduling problem (JSP). Biogeography-based optimization (BBO) is a new bio-inpired computation method that is based on the science of biogeography. The BBO algorithm searches for the global optimum mainly through two main steps: migration and mutation. As JSP is one of the most difficult combinational optimization problems, the original BBO algorithm cannot handle it very well, especially for instances with larger size. The proposed HBBO algorithm combines the chaos theory and “searching around the optimum” strategy with the basic BBO, which makes it converge to global optimum solution faster and more stably. Series of comparative experiments with particle swarm optimization (PSO), basic BBO, the CPLEX and 14 other competitive algorithms are conducted, and the results show that our proposed HBBO algorithm outperforms the other state-of-the-art algorithms, such as genetic algorithm (GA), simulated annealing (SA), the PSO and the basic BBO.     

面向期限感知分布式矩阵相乘的高效存储方案

分布式矩阵相乘是众多分布式机器学习、科学计算等应用中的关键操作,但其性能会受到系统中常见的落后节点的严重影响。最近研究者提出了基于喷泉码的编码矩阵相乘方法,能够充分利用落后节点的部分计算结果,从而大幅度减轻落后节点问题,但忽略了工作节点的存储开销。在考虑存储开销与计算完成时间之间的权衡关系的基础上,首先提出了面向异构工作节点的计算期限感知的存储优化问题;然后进一步通过理论分析,提出了基于期望近似的解决思路,并通过松弛将问题转化为凸优化问题以方便高效求解。仿真实验表明,在保证较大的任务成功率的情况下,所提方案的存储开销会随着任务期限的放宽迅速下降,并且该方案能够更大幅度降低编码带来的存储开销。也就是说,所提方案能够在保障整体计算在期限内大概率完成的前提下,大幅度降低总体的额外存储负载。     

An ant colony optimization approach to multi-objective optimal design of symmetric hybrid laminates for maximum fundamental frequency and minimum cost

An ant colony optimization algorithm for optimum design of symmetric hybrid laminates is described. The objective is simultaneous maximization of fundamental frequency and minimization of cost. Number of surface and core layers made of high-stiffness and low-stiffness materials, respectively, and fiber orientations are the design variables. Optimal stacking sequences are given for hybrid graphite/epoxy-glass/epoxy laminated plates with different aspect ratios and number of plies. The results obtained by ant colony optimization are compared to results obtained by a genetic algorithm and simulated annealing. The effectiveness of the hybridization concept for reducing the weight and keeping the fundamental frequency at a reasonable level is demonstrated. Furthermore, it is shown that the proposed ant colony algorithm outperforms the two other heuristics.     

时滞广义区间系统的弹性保性能控制

研究了一类时滞广义区间系统的弹性保性能控制问题.利用线性矩阵不等式处理方法,导出了系统弹性保性能控制器存在的条件,证明了该条件等价于一个线性矩阵不等式的可行性问题,并用该线性矩阵不等式的可行解给出了弹性保性能控制器的一个参数化表示.进一步,通过求解一个凸优化问题,给出了系统的最优弹性保性能控制器的设计方法.最后的数值例子说明了所给方法的有效性.     

混合自适应粒子群算法在电力经济调度中的应用

以电力系统中发电成本最低为目标,结合实际发电运行中系统平衡约束和机组操作约束条件,建立电力经济调度(ED)模型。由于标准粒子群算法存在易陷入局部最优的问题,用这种方法求解ED模型得到的最终结果会不太理想。为此,本文提出一种非线性自适应权重调整策略来增强算法全局搜索和局部搜索能力,首先引入小生境优化种群策略使算法跳出局部最优,然后将这种改进后的混合自适应粒子群算法（HAPSO）应用于求解ED模型。最后,算例分析结果表明本文所改进算法的有效性,提高了求解精度。     

一种基于冷链低碳物流路径的混合优化算法

冷链物流运输路径优化问题是国内外学者近几年研究的一个热点问题.首先分析了冷链物流运输的特点和需求,并且提出了考虑综合代价、带硬时间窗和容量约束的单物流中心车辆路径优化模型,其中综合代价包括固定代价、车辆运输代价、货损代价、制冷代价、时间惩罚代价和碳排放代价.车辆路径规划问题是一个NP问题,需要使用智能优化算法对解空间进...     

基于交替方向乘子法的非光滑损失坐标优化算法

交替方向乘子法(ADMM)在机器学习问题中已有一些实际应用。针对大规模数据的处理和非光滑损失凸优化问题,将镜面下降方法引入原ADMM批处理算法,得到了一种新的改进算法,并在此基础上提出了一种求解非光滑损失凸优化问题的坐标优化算法。该算法具有操作简单、计算高效的特点。通过详尽的理论分析,证明了新算法的收敛性,在一般凸条件下其具有目前最优的收敛速度。最后与相关算法进行了对比,实验结果表明该算法在保证解稀疏性的同时拥有更快的收敛速度。     

Multi-material structural topology optimization with decision making of stiffness design criteria

A new methodology for making design decisions of structures using multi-material optimum topology information is presented. Multi-material analysis contributes significant applications to enhance the bearing capacity and performance of structures. A method that chooses an appropriate material combination satisfying design stiffness requirement economically is currently needed. An alternative method of making design-decision is to utilize a multi-material topology optimization (MMTO) approach. This study provides a new computational design optimization procedure as a guideline to find the optimal multi-material design by considering structure strain energy and material cost. The MMTO problem is analyzed using an alternative active-phase approach. The procedure consists of three design steps. First, steel grid configurations and composite with material properties are defined as a given structure for automatic design decision-making (DDM). And then design criteria of the steel composites structure is given to be limited strain energy by designers and engineers. Second, topology changes in the automatic distribution of multi-steel materials combination and volume control of each material during optimization procedures are achieved and at the same time, their converged minimal strain energy is produced for each material combination. And third, the strain energy and material cost which is computed based on the material ratio in the combinations are used as design decision parameters. A study in constructional steel composites to produce optimal and economical multi-material designs demonstrates the efficiency of the present DDM methodology.     

基于弦内角映射参数的多边形三角剖分快速优化

为提高三角剖分质量及其优化的速度，提出了基于弦内角映射参数的三角剖分优化算法，三角剖分优化问题实质是凸四边形的对角线选择问题，在两个三角形组成的凸四边形中，将弦内角映射成两三角形公共边中垂线上的映射参数值，经过证明映射参数与弦内角具有等价的三角剖分优化判别特性，因此三角剖分局部优化转化为映射参数的判别问题。理论分析与实践表明，该方法实现非常容易，且映射参数计算简洁方便、快捷可靠，弦内角映射参数判别方法避免了传统方法所必需的角度、半径距离计算，明显地提高了计算效率。     

非平衡有向网络的完全分布式凸优化

分布式凸优化问题的目的是如何以分布式方法最小化局部智能体成本函数和,而现有分布式算法的控制步长选取依赖于系统智能体个数、伴随矩阵等全局性信息,有悖于分布式算法的初衷.针对此问题,提出一种基于非平衡有向网络的完全分布式凸优化算法(FDCOA).基于多智能体一致性理论和梯度跟踪技术,设计了一种非负余量迭代策略,使得FDCOA的控制步长收敛范围仅与智能体局部信息相关,进而实现控制步长的分布式设置.进一步分析了FDCOA在固定强连通和时变强连通网络情形下的收敛性.仿真结果表明本文构建的分布式控制步长选取方法对FDCOA在有向非平衡下的分布式凸优化问题是有效的.