变异PSO算法协同神经元网络在轧制力预报中的应用

 为了避免BP神经元网络易陷入局部极值和基本粒子群(PSO) 神经元网络早熟收敛问题，采用一种自适应变异的粒子群优化算法训练神经元网络，根据轧制力的实测值和神经元网络的预报值确定粒子群算法的适应度函数，按照权重梯度方向进行变异操作，并首次将该方法应用到热连轧机组轧制力预报中。通过攀钢热轧板厂现场数据运算表明，该方法的预报误差平均值比传统数学模型低165%，比BP神经元网络低055%，收敛速度比BP神经元网络提高了约1/4，为进一步提高精轧机组轧制力预报精度提供了一种新的有效方法。     

基于粒子群优化算法的钢铁企业铁路车辆调度

给出了求解铁路车辆调度问题的粒子群算法流程;分析了求解不同调度问题的3种粒子表示法,即基于粒子位置次序(Particle Position Sequence,PPS)的粒子表示法、基于粒子位置取整操作(Particle Position Rounding off,PPR)的粒子表示法和基于PPS PPR的混合粒子表示法;讨论了PPS PPR混合粒子表示法与调度解空间的映射关系和解码方法。将第3种方法应用于实际车辆调度系统中,求解出机车送货作业行驶的最短路径,建立了基于粒子群优化算法的企业铁路优化调度模型。     

基于粒子群优化算法的钢铁企业铁路车辆调度

给出了求解铁路车辆调度问题的粒子群算法流程;分析了求解不同调度问题的3种粒子表示法,即基于粒子位置次序(Particle Position Sequence,PPS)的粒子表示法、基于粒子位置取整操作(Particle Position Rounding off,PPR)的粒子表示法和基于PPS PPR的混合粒子表示法;讨论了PPS PPR混合粒子表示法与调度解空间的映射关系和解码方法。将第3种方法应用于实际车辆调度系统中,求解出机车送货作业行驶的最短路径,建立了基于粒子群优化算法的企业铁路优化调度模型。     

多目标粒子群优化算法研究综述

针对多目标粒子群优化算法的研究进展进行综述。首先,回顾了多目标优化和粒子群算法等基本理论;其次,分析了多目标优化所涉及的难点问题;再次,从最优粒子选择策略,多样性保持机制,收敛性提高手段,多样性与收敛性平衡方法,迭代公式、参数、拓扑结构的改进方案5个方面综述了近年来的最新成果;最后,指出多目标粒子群算法有待进一步解决的问题及未来的研究方向。      

基于改进差分进化算法的加热炉调度方法

提出一种以燃料消耗量最小为优化目标的加热炉生产调度新方法。首先基于热力学第一定律分析了流入及流出加热炉的各项能量,并对燃料消耗量的计算式进行了理论推导。进而根据加热炉区实际生产调度特点归纳各约束条件,以多台加热炉总燃料消耗量最小为优化目标,构建调度优化数学模型。采用自适应差分进化算法搭配禁忌搜索算法进行综合求解,并通过9组实际钢坯生产案例模拟验证了该算法的可行性和有效性。同时,为了探究加热炉燃料消耗量的影响因素,提出了分别衡量加热炉区缓冲等待、炉内加热两部分时间同理想生产时间匹配程度的评价参数μ₁和μ₂,并分析了燃料消耗量对二者的敏感性,结果表明:当连铸坯到达加热炉节奏与热轧工序出坯节奏之比由0.5增至2时,燃料消耗量对两评价参数的敏感性逐渐减弱。      

改进粒子群算法及其在热连轧负荷分配中的应用

对一种已有的自适应算法进行了改进,并将该算法思想引入到粒子群算法的改进中,在种群进化到一定代数时按照改进自适应算法改变搜索范围的大小,实现了自动调整搜索范围、提高收敛速度和精度并可有效防止粒子群算法早熟收敛的目的,同时通过实验仿真进行了验证.将该改进粒子群算法应用到热连轧机精轧机组的负荷分配优化计算中,程序运行时间小于5s,满足实时性的要求,为其提供了一种更为有效的优化手段.     

基于粒子群优化算法和博弈论的网络学习控制系统带宽调度

研究了双层网络学习控制系统的带宽调度优化问题.为了合理分配子系统的带宽,引入了网络定价体系和动态带宽调度方法,建立了非合作博弈模型,从而将网络控制系统的网络资源分配问题转换为非合作博弈竞争模型下的Nash均衡点求解问题.在此基础上,采用粒子群优化算法得到此框架下的纳什均衡解,并进一步给出了网络控制系统的时间片调度方法.仿真结果表明了所提方法的有效性.     

改进粒子群优化神经网络及其在产品质量建模中的应用

针对传统神经网络优化算法易陷入局部最优值的问题,在标准粒子群算法的基础上,对粒子速度与位置更新策略进行改进,提出一种基于改进粒子群优化算法的BP神经网络建模方法.使用sinc函数、波士顿住房数据及某钢厂带钢热镀锌生产的实际数据进行验证.结果表明,与标准的反向传播神经网络和支持向量机相比,基于改进粒子群优化的神经网络模型可以有效提高预测精度.     

基于改进差分进化算法的钢铁企业氧气系统优化调度

针对制氧系统在平衡调度过程中由于氧气生产的连续性与用氧设备消耗的间歇性之间的矛盾所造成供需不平衡、氧气放散率高的问题,研究小时级的氧气系统供需协调优化方案。考虑氧气系统中的设备产能、管网平衡和机组变负荷等约束,建立了氧气系统优化调度模型,优化氧气系统的综合经济效益,设计了基于自适应变异、中间重组交叉的差分进化(differential evolution, DE)算法进行求解。试验结果表明,DE算法在有效时间内能够求解模型。对于小规模算例,在有限的时间内能够找到近优解;对于大规模算例,DE算法从解的质量和求解速度上都要优于求解器的解,能够满足实际生产的需求。     

基于PSO的Live Wire交互式图像分割算法

传统Live Wire算法易受伪轮廓干扰,并且算法执行速度较慢.针对这些问题,提出一种基于PSO的Live Wire交互式图像分割算法.算法首先构造新的代价函数,引入相邻节点间梯度幅值变化函数来减轻伪轮廓的干扰,提高了算法的分割精度;其次,为了提高算法的执行效率,应用粒子群算法求取图像中任意两点间最短路径来定位目标边界,并与经典的基于Dijkstra动态规划图搜索的Live Wire算法进行比较.实验结果表明,与传统方法相比,所提算法在分割精度和执行效率上都有很大提高.     

基于改进粒子群优化算法的机器人路径规划

提出了一种改进粒子群优化算法的移动机器人路径规划方法。该方法首先将粒子群分成两组,对其中一组加入变异算子,能提高种群的多样性和避免粒子群优化算法的早熟。该方法模型简单,算法复杂度低,收敛速度快。仿真实验结果获得了从起点到终点的无碰撞路径,证实了该方法的有效性和可行性。     

An Improved Particle Swarm Optimization for Feature Selection

Particle Swarm Optimization (PSO) is a popular and bionic algorithm based on the social behavior associated with bird flocking for optimization problems. To maintain the diversity of swarms, a few studies of multi-swarm strategy have been reported. However, the competition among swarms, reservation or destruction of a swarm, has not been considered further. In this paper, we formulate four rules by introducing the mechanism for survival of the fittest, which simulates the competition among the swarms. Based on the mechanism, we design a modified Multi-Swarm PSO (MSPSO) to solve discrete problems,which consists of a number of sub-swarms and a multi-swarm scheduler that can monitor and control each sub-swarm using the rules. To further settle the feature selection problems, we propose an Improved Feature Selection (IFS) method by integrating MSPSO, Support Vector Machines (SVM) with F-score method. The IFS method aims to achieve higher generalization capability through performing kernel parameter optimization and feature selection simultaneously. The performance of the proposed method is compared with that of the standard PSO based, Genetic Algorithm (GA) based and the grid search based methods on 10 benchmark datasets, taken from UCI machine learning and StatLog databases. The numerical results and statistical analysis show that the proposed IFS method performs significantly better than the other three methods in terms of prediction accuracy with smaller subset of features.     

基于改进微粒群优化的泊松曲线沉降预测模型

在现有文献研究的基础上,对泊松曲线沉降预测模型作了进一步的研究,给出了泊松曲线沉降预测模型的一个新方法-改进的微粒群最优化方法.该算法不需要计算梯度,容易应用于实际问题中.通过对微粒群算法的修正,使改进算法具有更加精确和快速的收敛性.经实例计算表明,这种方法具有较高的精度.     

微粒群优化的Canny边缘检测方法

通过对Canny算法进行改进,提出了一种基于量子行为的微粒群优化算法的 边缘提取算法.改进算法对噪声抑制效果明显,能够删除伪边缘,得到精确的边缘.实验结果表明,该算法在保证实时性的同时,具有很好的检测精度和准确度.     

Particle Swarm Optimization-Supported Simulation for Construction Operations

This study proposes an integration of particle swarm optimization (PSO) and a construction simulation so as to determine efficiently the optimal resource combination for a construction operation. The particle-flying mechanism is utilized to guide the search process for the PSO-supported simulation optimization. A statistics method, i.e., multiple-comparison procedure, is adopted to compare the random output performances resulting from the stochastic simulation model so as to rank the alternatives (i.e., particle-represented resource combinations) during the search process. The indifference zone and confidence interval facilitate consideration of the secondary performance measure (e.g., productivity) when the main performance measures (e.g., cost) of the competing alternatives are close. The experimental analyses demonstrate the effectiveness and efficiency of the proposed simulation optimization. The study aims to providing an alternative combination of optimization methodology and general construction simulation by utilizing PSO and a statistics method so as to improve the efficiency of simulation in planning construction operations.     

Permutation-Based Particle Swarm Optimization for Resource-Constrained Project Scheduling

In the light of particle swarm optimization (PSO) which utilizes both local and global experiences during search process, a permutation-based scheme for the resource-constrained project scheduling problem (RCPSP) is presented. In order to handle the permutation-feasibility and precedence-constraint problems when updating the particle-represented sequence or solution for the RCPSP, a hybrid particle-updating mechanism incorporated with a partially mapped crossover of a genetic algorithm and a definition of an activity-move-range is developed. The particle-represented sequence should be transformed to a schedule (including start times and resource assignments for all activities) through a serial method and accordingly evaluated against the objective of minimizing project duration. Experimental analyses are presented to investigate the performances of the permutation-based PSO. The study aims at providing an alternative for solving the RCPSP in the construction field by utilizing the advantages of PSO.     

Particle Swarm Optimization for Construction Site Unequal-Area Layout

Layout of temporary facilities on a construction site is essential to enhancing productivity and safety, and is a complex issue due to the unique nature of construction. This paper proposes a particle swarm optimization (PSO)-based methodology to solve the construction site unequal-area facility layout problem. A priority-based particle representation of the candidate solutions to the layout problem is proposed. The particle-represented solution in terms of priorities should be transformed to the specific layout plan with consideration of nonoverlap and geometric constraints. In addition, a modified solution space boundary handling approach is proposed for controlling particle updating with regard to the priority value range. Computational experiments are carried out to justify the efficiency of the proposed method and investigate its underlying performances. This study aims at providing an alternative and effective means for solving the construction site unequal-area layout problem by utilizing the PSO algorithm.     

改进粒子群算法对1420mm五机架冷连轧机轧制规程的优化

某厂1 420mm五机架冷连轧机自试生产以来,出现板形质量不稳定的现象。分析认为,这种现象与现有轧制规程的设定有关。现有的轧制规程是一种经验值,对带钢板形质量未作深入的考虑。为此,对现有轧制规程进行了研究,运用Matlab建立了一种设定给定产品轧制规程的新方法,并基于最优板形质量的目标函数,采用改进的粒子群优化算法,制定出兼顾板形和设备负荷能力的优化的轧制规程。投入实际生产后的实测数据对比表明,优化后的轧制规程生产带钢平直度小于6I的分布由66%上升到88%,带钢板形明显改善。     

Particle Swarm Optimization for Preemptive Scheduling under Break and Resource-Constraints

This paper introduces a particle swarm optimization (PSO)-based methodology to implement preemptive scheduling under break and resource-constraints (PSBRC) for construction projects. The PSBRC under study allows the preemptive activities to be interrupted in off-working time and not to resume immediately in the next working period because all the limited resources are to be reallocated during a break. The potential solution to the PSBRC, i.e., a set of priorities deciding the order to start the activities or restart the interrupted activities, is represented by the multidimensional particle position. Hence PSO is applied to search for the optimal schedule for the PSBRC, in which a parallel scheme is adopted to transform the particle-represented priorities to a schedule. Computational analyses are presented to verify the effectiveness of the proposed methodology. This paper provides an attempt to make use of preemption and break for the resource-constrained construction project with the objective of minimizing project duration.     

基于粒子群优化的过程神经网络学习算法

基于粒子群优化为过程神经元网络提出了一种新的学习算法.新算法在对网络输入函数和连接权函数进行正交基函数展开后,将网络中的结构参数和其他参数整合成一个粒子,再用粒子群优化算法进行全局优化.新算法不依赖于函数梯度信息,不需要手动调节网络结构.粒子群优化具有良好的全局优化性能和收敛性能,保证了过程神经元网络的全局学习能力和新学习算法的收敛能力,更好地发挥过程神经网络的逼近性能.两个实际预测问题的实验结果表明,基于粒子群优化的学习算法比现有的基于梯度的基函数展开方法以及误差反传神经网络模型具有更好的预测精度.