Edge Crossing Minimization Algorithm for Hierarchical Graphs Based on Genetic Algorithms

We present an edge crossing minimization algorithm for hierarchical graphs based on genetic algorithms, and comparing it with some heuristic algorithms. The proposed algorithm is more efficient and has the following advantages: the frame of the algorithms is unified, the method is simple, and its implementation and revision are easy.

Discrete logistics network design model under interval hierarchical OD demand based on interval genetic algorithm

Aimed at the uncertain characteristics of discrete logistics network design,an interval hierarchical triangular uncertain OD demand model based on interval demand and network flow is presented.Under consideration of the system profit,the uncertain demand of logistics network is measured by interval variables and interval parameters,and an interval planning model of discrete logistics network is established.The risk coefficient and maximum constrained deviation are defined to realize the certain transformation of the model.By integrating interval algorithm and genetic algorithm,an interval hierarchical optimal genetic algorithm is proposed to solve the model.It is shown by a tested example that in the same scenario condition an interval solution[3275.3,3 603.7]can be obtained by the model and algorithm which is obviously better than the single precise optimal solution by stochastic or fuzzy algorithm,so it can be reflected that the model and algorithm have more stronger operability and the solution result has superiority to scenario decision.

An Approach to Assembly Sequence Plannning Based on Hierarchical Strategy and Genetic Algorithm

Using group and subassembly cluster methods, the hierarchical structure of a product is ?generated automatically, which largely reduces the complexity of planning. Based on genetic algorithm, the optimal of assembly sequence of each structure level can be obtained by sequence-by-sequence search. As a result, a better assembly sequence of the product can be generated by combining the assembly sequences of all hierarchical structures, which provides more parallelism and flexibility for assembly operations. An industrial example is solved by this new approach.

Wind Turbine Optimal Preventive Maintenance Scheduling Using Fibonacci Search and Genetic Algorithm

Maintenance scheduling is essential and crucial for wind turbines (WTs) to avoid breakdowns andreduce maintenance costs. Many maintenance models have been developed for WTs’ maintenance planning, suchas corrective, preventive, and predictive maintenance. Due to communities’ dependence on WTs for electricityneeds, preventive maintenance is the most widely used method for maintenance scheduling. The downside tousing this approach is that preventive maintenance (PM) is often done in fixed intervals, which is inefficient. In thispaper, a more detailed maintenance plan for a 2 MW WT has been developed. The paper’s focus is to minimize aWT’s maintenance cost based on a WT’s reliability model. This study uses a two-layer optimization framework:Fibonacci and genetic algorithm. The first layer in the optimization method (Fibonacci) finds the optimal numberof PM required for the system. In the second layer, the optimal times for preventative maintenance and optimalcomponents to maintain have been determined to minimize maintenance costs. The Monte Carlo simulationestimates WT component failure times using their lifetime distributions from the reliability model. The estimatedfailure times are then used to determine the overall corrective and PM costs during the system’s lifetime. Finally,an optimal PM schedule is proposed for a 2 MW WT using the presented method. The method used in this papercan be expanded to a wind farm or similar engineering systems.

Development and Comparison of Hybrid Genetic Algorithms for Network Design Problem in Closed Loop Supply Chain

This paper presents four different hybrid genetic algorithms for network design problem in closed loop supply chain. They are compared using a complete factorial experiment with two factors, viz. problem size and algorithm. Based on the significance of the factor “algorithm”, the best algorithm is identified using Duncan’s multiple range test. Then it is compared with a mathematical model in terms of total cost. It is found that the best hybrid genetic algorithm identified gives results on par with the mathematical model in statistical terms. So, the best algorithm out of four algorithm proposed in this paper is proved to be superior to all other algorithms for all sizes of problems and its performance is equal to that of the mathematical model for small size and medium size problems.

Multicast Routing Based on Hybrid Genetic Algorithm

A new multicast routing algorithm based on the hybrid genetic algorithm (HGA) is proposed. The coding pattern based on the number of routing paths is used. A fitness function that is computed easily and makes algorithm quickly convergent is proposed. A new approach that defines the HGA's parameters is provided. The simulation shows that the approach can increase largely the convergent ratio, and the fitting values of the parameters of this algorithm are different from that of the original algorithms. The optimal mutation probability of HGA equals 0.50 in HGA in the experiment, but that equals 0.07 in SGA. It has been concluded that the population size has a significant influence on the HGA's convergent ratio when it's mutation probability is bigger. The algorithm with a small population size has a high average convergent rate. The population size has little influence on HGA with the lower mutation probability.

Cluster Analysis Assisted Float-Encoded Genetic Algorithm for a More Automated Characterization of Hydrocarbon Reservoirs

A genetic algorithm-based joint inversion method is presented for evaluating hydrocarbon-bearing geological formations. Conventional inversion procedures routinely used in the oil industry perform the inversion processing of borehole geophysical data locally. As having barely more types of data than unknowns in a depth, a set of marginally over-determined inverse problems has to be solved along a borehole, which is a rather noise sensitive procedure. For the reduction of noise effect, the amount of overdetermination must be increased. To fulfill this requirement, we suggest the use of our interval inversion method, which inverts simultaneously all data from a greater depth interval to estimate petrophysical parameters of reservoirs to the same interval. A series expansion based discretization scheme ensures much more data against unknowns that significantly reduces the estimation error of model parameters. The knowledge of reservoir boundaries is also required for reserve calculation. Well logs contain information about layer-thicknesses, but they cannot be extracted by the local inversion approach. We showed earlier that the depth coordinates of layerboundaries can be determined within the interval inversion procedure. The weakness of method is that the output of inversion is highly influenced by arbitrary assumptions made for layer-thicknesses when creating a starting model (i.e. number of layers, search domain of thicknesses). In this study, we apply an automated procedure for the determination of rock interfaces. We perform multidimensional hierarchical cluster analysis on well-logging data before inversion that separates the measuring points of different layers on a lithological basis. As a result, the vertical distribution of clusters furnishes the coordinates of layer-boundaries, which are then used as initial model parameters for the interval inversion procedure. The improved inversion method gives a fast, automatic and objective estimation to layer-boundaries and petrophysical parameters, which is demonstrated by a hydrocarbon field example.

Implementation of Hybrid Genetic Algorithm for CLSC Network Design Problem—A Case Study on Fashion Leather Goods Industry

The implementation of closed loop supply chain system is becoming essential for fashion leather products industry to ensure an economically sustainable business model and eco-friendly industrial practice as demanded by the environmental regulations, consumer awareness and the prevailing social consciousness. In this context, this research work addresses a closed loop supply chain network problem of fashion leather goods industry, with an objective of minimizing the total cost of the entire supply chain and also reducing the total waste from the end of life product returns. The research work commenced with a literature review on the reverse and closed loop supply chain network design problems of fashion and leather goods industry dealt in the past. Then, the identified CLSCND problem is solved using a mathematical model based on Mixed Integer Non-Linear Programme (MINLP) and then a suitable Hybrid Genetic Algorithm (HGA) developed for the CLSCND is implemented for obtaining optimum solution. Both the MINLP model and HGA are customized as per the CLSCND problem chosen and implemented for the industrial case of an Indian Fashion Leather Goods Industry. Finally, the solutions obtained for MINLP model in LINGO 15 and for HGA in VB.NET platform are compared and presented. The optimum solution obtained from the suitable HGA is illustrated as an optimum shipment pattern for the closed loop supply chain network design problem of the fashion leather goods industry case.

Hierarchical resource allocation for integrated modular avionics systems

Recently the integrated modular avionics （IMA） architecture which introduces the concept of resource partitioning becomes popular as an alternative to the traditional federated architecture. A novel hierarchical approach is proposed to solve the resource allocation problem for IMA systems in distributed environments. Firstly, the worst case response time of tasks with arbitrary deadlines is analyzed for the two-level scheduler. Then, the hierarchical resource allocation approach is presented in two levels. At the platform level, a task assignment algorithm based on genetic simulated annealing （GSA） is proposed to assign a set of pre-defined tasks to different processing nodes in the form of task groups, so that resources can be allocated as partitions and mapped to task groups. While yielding to all the resource con- straints, the algorithm tries to find an optimal task assignment with minimized communication costs and balanced work load. At the node level, partition parameters are optimized, so that the computational resource can be allocated further. An example is shown to illustrate the hierarchal resource allocation approach and manifest the validity. Simulation results comparing the performance of the proposed GSA with that of traditional genetic algorithms are presented in the context of task assignment in IMA systems.

GA与分层优化结合的掺水集油工艺参数优化

为降低进入特高含水阶段油田的掺水集油能耗,对掺水集油工艺参数进行优化。以掺水温度、掺水量、掺水压力为外层决策变量,加热炉和掺水泵运行方案为内层决策变量,以掺水集油工艺综合能耗最低为目标函数,建立掺水集油工艺参数优化模型。提出了一种遗传算法(Genetic Algorithm,GA)与分层优化结合的求解策略,将掺水温度作为染色体上的基因,先优化个体对应加热炉和掺水泵运行方案,得到个体对应的最小集油能耗,再通过种群的不断迭代进化,得到掺水集油工艺的最优参数。研究结果表明,与优化前相比,掺水温度降低8.3℃,掺水量减少131.6 m^(3)/d,掺水压力降低0.15 MPa,集输吨液综合能耗降低13.64%,优化效果良好。研究成果可为进入特高含水阶段油田降低掺水集油能耗提供借鉴。

基于多信息融合的层次聚类测井曲线自动分层方法

随着人工智能的快速发展,机器学习的应用范围越来越广泛,将机器学习的方法用于测井曲线分层可以提高分层效率和精度。在利用测井资料进行岩性识别、沉积相分析等研究时,先要对测井曲线进行分层。文中提出一种基于多信息融合的层次聚类分层方法,实现了对测井曲线的自动分层。首先,采用滤波的方式滤除曲线上的噪点,对数据进行归一化处理,消除量纲的影响;其次,通过特征优选,选择包含较多地层信息的特征曲线,构造一个滤波器,将其中相似性较高的曲线融合,曲线融合的权值通过遗传算法求得;最后,使用层次聚类方法对多信息融合后的测井数据进行划分,将分层结果与人工分层结果进行对比验证。该方法能够提高分层效率,为地质勘探工作提供可靠的分层依据。

等待时间受限Flowshop调度的HGA算法

针对等待时间受限的Flowshop调度问题,提出嵌入约束满足和变邻域搜索技术的混合遗传算法。该算法基于约束满足思想,通过递归回溯和约束传播修复工件的开工时间,以解消工件在相邻阶段的等待时间受限冲突,根据回溯工件的位置信息设计相应的交叉算子和变异算子,利用变邻域搜索技术增强算法的收敛性。仿真实验表明该混合遗传算法的有效性,并分析等待时间上限对目标值的影响。

间歇式信息传输条件下无人机搜索覆盖规划

在基站通信范围受限条件下,若无人机(unmanned aerial vehicle,UAV)执行覆盖搜索任务时经常返回至基站通信范围内实现间歇式信息传输,能够扩展其覆盖区域和提高执行任务的灵活性。为最小化所有环境位点信息传回基站的时间之和,需解决覆盖规划和间歇式通信时机选择的耦合问题。在覆盖的目标点较少且分散时,采用改进的层次聚类方法求解每次往返需要覆盖的路径点集合。在需要进行区域全覆盖时,则在求解完区域的覆盖路径后,以最小化时间之和为目标,对目标函数进行分析,确定最优返回次数的搜索范围,压缩解空间。对该搜索范围进行遍历搜索得到最优往返次数,然后利用遗传算法优化UAV返回位点。与前沿算法对比,所提算法在目标函数和覆盖路径质量上具有一定的提升。

基于加速度与HGA-BP神经网络的人体行为识别

在基于加速度传感器的人体行为识别中,分类器复杂度较高,易产生过拟合现象。为此,通过递阶遗传算法(HGA)训练BP神经网络作为分类器,采用三级染色体递阶结构表示神经网络的结构和参数。设计新的适应度函数,采用选择、交叉和变异操作联合优化BP网络的精确度和复杂度。测试结果表明,在基于加速度信号的行为识别系统中,相比基本HGA和其他常用算法,利用改进的HGA训练BP网络分类器可以有效控制网络结构,在保证隐层神经元数目较少的情况下,尽可能降低输出误差,实现两者的动态平衡,且对测试样本的识别正确率可达94.63%。

三维荧光光谱结合HGA-RBF神经网络在多环芳烃浓度检测中的应用（英文）

采用FS920荧光光谱仪分析了苯并[k]荧蒽(BkF)、苯并[b]荧蒽(BbF)和两者混合物的荧光特性.结果表明BkF的两个荧光峰分别位于306nm/405nm和306nm/430nm,BbF的两个荧光峰分别位于306nm/410nm和306nm/435nm.BkF和BbF不同浓度配比及其相互间的荧光干扰,使得混合物荧光特性差异较大,荧光强度和浓度间关系变得复杂.为准确测定混合物中BkF和BbF的浓度,采用递阶算法优化的径向基神经网络对其进行检测,结果表明BkF和BbF的平均回收率分别为98.45%和97.71%.该方法能够实现多环芳烃类污染物共存成分的识别和浓度预测.

一种基于HGA和数据挖掘的AMG模型

提出一种基于数据挖掘和混合遗传算法(HGA)的自适应模型生成(AMG)模型。采用改进的聚类算法,从网络和系统的行为记录中划分出正常/异常行为库,利用HGA从行为库中挖掘出入侵规则加入规则库中,通过混合检测模块进行检测。实验结果证明,该AMG模型能以更高的检测率、更低的误检率检测未知的网络入侵。

众包制造环境下协同产品族设计与延迟决策的主从关联优化

针对众包制造环境下产品族设计与延迟决策之间的协同优化问题,建立了以产品族设计为主、延迟决策为从的混合整数非线性双层规划模型。模型上层由制造商设计产品族架构以最大化自身的期望利润,下层由多个分销商优化延迟产品模块的类型以最大化各自的期望利润,且各分销商之间的决策具有相互独立性。考虑到延迟产品模块事先不确定,提出虚拟延迟结构将该优化问题具体化,并证明需要被延迟的产品模块。开发了一个双层嵌套遗传算法对模型进行求解。用智能冰箱产品族延迟案例验证了所提模型和算法的有效性,并对所提延迟偏好参数进行了灵敏度分析。

基于HGA的模糊神经控制器设计及其应用

将神经网络与模糊控制相结合,实现了模糊控制器的自学习和自适应。给出一种基于递阶遗传算法的模糊神经网络优化算法,通过对每个染色体采用递阶编码,可以同时优化模糊神经网络结构和权值参数。将这种模糊神经网络控制器应用于镍氢电池的充电控制中,证明了算法的有效性。

基于HGA的最小旅行时间多旅行商问题研究

为了解决最小化旅行时间的多旅行商一类问题,提出了一种递阶遗传算法和矩阵解码方法。该算法根据问题的特点,采用一种递阶编码方案,此编码与多旅行商问题一一对应。用递阶遗传算法优化多旅行商问题不需设计专门的遗传算子,操作简单,并且解码方法适于求解距离矩阵对称和距离矩阵非对称的多旅行商问题。计算结果表明,递阶遗传算法是有效的,能适用于优化最小化完成时间的多旅行商问题。

促进用户负荷特性优化的分时电价机制设计方法

在电力供应结构和消费方式变化背景下,提出一种分时电价机制设计方法,旨在改善用户负荷特性,促进电网削峰填谷。根据工业、商业和居民三类用户的负荷分布特性,通过分层聚类方法对不同季节的峰谷时段进行重新划分,并引入深谷时段,解决原有分时电价机制时段划分不准确的问题。在满足覆盖发电成本前提下,通过减少用户用电成本与扩大峰谷价差,进一步激励用户调整用电行为,并采用量子遗传算法(quantum genetic algorithm,QGA)对时段电价制定的优化问题进行求解。通过实际算例,计算用户响应改进后分时电价机制前后的削峰量和填谷量,验证所设计的分时电价机制可以降低用户用电成本,并有效转移电网高峰时段负荷,缓解时段性、季节性的供电压力问题。