System optimization-oriented spare parts dynamic configuration model for multi-echelon multi-indenture system

In order to optimize the spares configuration project at different stages during the life cycle, the factor of time is considered to relax the assumption of the spares steady demand in multi-echelon technique for recoverable item control (METRIC) theory. According to the method of systems analysis, the dynamic palm theorem is introduced to establish the prediction model of the spares demand rate, and its main influence factors are analyzed, based on which, the spares support effectiveness evaluation index system is studied, and the system optimization-oriented spares dynamic configuration method for multi-echelon multi-indenture system is proposed. Through the analysis of the optimization algorithm, the layered marginal algorithm is designed to improve the model calculation efficiency. In a given example, the multi-stage spares configuration project during its life cycle is gotten, the research result conforms to the actual status, and it can provide a new way for the spares dynamic optimization.

Location and allocation problem for spare parts depots on integrated logistics support

In equipment integrated logistics support(ILS), the supply capability of spare parts is a significant factor. There are lots of depots in the traditional support system, which makes too many redundant spare parts and causes high cost of support. Meanwhile,the inconsistency among depots makes it difficult to manage spare parts. With the development of information technology and transportation, the supply network has become more efficient. In order to further improve the efficiency of supply-support work and the availability of the equipment system, building a system of one centralized depot with multiple depots becomes an appropriate way.In this case, location selection of the depots including centralized depots and multiple depots becomes a top priority in the support system. This paper will focus on the location selection problem of centralized depots considering ILS factors. Unlike the common location selection problem, depots in ILS require a higher service level. Therefore, it becomes desperately necessary to take the high requirement of the mission into account while determining location of depots. Based on this, we raise an optimal depot location model. First, the expected transportation cost is calculated.Next, factors in ILS such as response time, availability and fill rate are analyzed for evaluating positions of open depots. Then, an optimization model of depot location is developed with the minimum expected cost of transportation as objective and ILS factors as constraints. Finally, a numerical case is studied to prove the validity of the model by using the genetic algorithm. Results show that depot location obtained by this model can guarantee the effectiveness and capability of ILS well.

Research on Three-Echelon Inventory Model and Algorithm for Valuable Spare Parts in Weapon Equipment

According to the current situation of repair and supply for valuable spare parts in the weapon equipment of the PLA,using the research results of the repairable spare parts inventory theory,an inventory model of the valuable spare parts in three-echelon repair and supply system is established by expanding the classical METRIC model and theory,and the genetic algorithm is utilized to solve the model.In the algorithm,the chromosome representation and initial population production,the crossover and mutation operators are designed.By using an example,a simulation analysis is carried out to verify the model's correctness.

基于人工鱼群-遗传算法的多品种小批量零件数控加工工艺优化研究

基于多品种小批量零件加工成本高的问题,基于人工鱼群-遗传算法(AFSA-GA)构建了数控机床能耗模型,以实现零件加工能耗下降。首先,将数控机床功率划分为各工序功率模型,基于功率模型与工作时间关系得出机床运转能耗模型,结合产品表面粗糙度模型,对各工序能耗模型及整体粗糙度进行归一化处理,形成整体能耗模型;其次,以能耗及粗糙度为目标函数,建立AFSA-GA算法,通过对各工序能耗求解得出最适当的机床功率及其所对应的能耗和表面粗糙度;最后,针对所获得的最优功率,进行优化结果的验证,为五轴机床的实际加工提供解决方案。

数据与知识双驱动的备件需求模糊预测模型

针对知识驱动型需求预测模型所需的专家知识稀缺、数据驱动型需求预测模型可解释性不足的问题,提出了数据与知识双驱动的备件需求模糊预测模型。该模型基于模糊聚类算法将数值型数据聚类为结构简单、可解释性强的规则库,运用模糊逻辑将领域专家知识表示为Mamdani型规则库。在此基础上,引入了一种新型智能计算理论——模糊网络理论对两类规则库进行合并运算,形成初始预测模型。采用遗传算法优化模型规则库的模糊集参数来提高模型预测准确性。通过与模糊聚类算法进行对比,提出的模型在可解释性以及准确性指标上均具有优势。

基于小样本数据的风电备品备件需求预测方法研究

全球风电装机容量不断攀升,我国风电产业也出现跨越式发展,风电行业的蓬勃发展给运维管理工作带来巨大挑战。风电运维管理中的备品备件需求预测是决定运维成本的重要因素,尤其是复杂工况下的备品备件预测是风电运维管理面临的亟待解决的问题。本文围绕风电行业运维数据呈现的小样本特征,对风电行业的小样本备品备件需求预测方法开展研究。根据风电行业备品备件运维需求特征,选择小波变换提取备品备件的历史需求数据的趋势与细节特征,选择合适的时间窗将时间序列转化为监督学习形式的时间序列对,提出混合遗传算法和支持向量回归的预测算法GA-dwtSVR,最后开展了验证实验,实验结果表明:相较于传统预测方法,GA-dwtSVR在风电备品备件需求预测中有较好的性能,可以满足风电行业备品备件基于小样本数据的预测需求。

基于遗传算法的核电备件库存模型仿真优化

备品备件是核电厂设备检修的物质基础,然而持有过量的备件库存将增加核电厂运营成本,通过设置合理的备件库存参数,在保障供应和控制库存之间寻找到一个合理的平衡点。针对核电备件库存参数设置问题,将遗传算法引入备件库存参数优化领域。通过建立核电备件库存参数评价方法,将核电备件库存参数设置转换为多变量优化问题,使用遗传算法计算最优库存参数。选用大亚湾核电厂的备件进行仿真测试,当给备件增加多种领用波动时,使用遗传算法计算的备件库存参数均优于当前数据库中的库存参数,结果表明遗传算法应用到备件库存参数优化领域具有重要的实用价值。

基于遗传算法的不规则天然气管道支承件位置避振优化

以多支承的天然气管道为研究对象,在有限元建模过程中通过引入弹簧单元来模拟管道支承件的力学特性。在避振优化部分,以管道系统1阶固有频率最大为优化目标,将支承件位置转换为节点编号作为设计变量,支承件的可移动范围作为约束条件。通过遗传算法不断修正管道系统刚度矩阵,从而得到合理的管道支承件位置。

LOCATORS OPTIMIZATION FOR MEASURING FIXTURE DESIGN

'N-2-1' principle is widely recognized in the fixture design for deformablesheet metal workpieces, where N, the locators on primary datum, is the key to sheet metal fixturedesign. However, little research is done on how to determine the positions and the number of Nlocators. In practice, the N locators are frequently designed from experience, which is oftenunsatisfactory for achieving the precision requirement in fixture design. A new method to lay outthe N locators for measuring fixture of deformable sheet metal workpiece is presented, given thefixed number of A'. Finite-element method is used to model and analysis the deformation of differentlocator layouts. A knowledge based genetic algorithm (KBGA) is applied to identify the optimumlocator layout for measuring fixture design. An example of a door outer is used to verify theoptimization approach.

Repairable Configuration Optimization Model under Two-Echelon Inventory

Spares inventory configuration optimization is an effective way to improve readiness and reduce life cycle cost of equipment.Through analyzing two-echelon spares support system,the METRIC model basic theory was used.An inventory configuration optimization model of two-echelon spares support system was proposed which took the spares expected shortfall as the object and made the minimum repairable parts expected shortfall instead of the maximum spares supportability as the objective function.Marginal efficiency analysis algorithm was applied to optimizing the spares configuration and generating a rational spares inventory configuration.Finally,several examples are given to verify the model.

Optimizing combination of aircraft maintenance tasks by adaptive genetic algorithm based on cluster search

It is significant to combine multiple tasks into an optimal work package in decision-making of aircraft maintenance to reduce cost,so a cost rate model of combinatorial maintenance is an urgent need.However,the optimal combination under various constraints not only involves numerical calculations but also is an NP-hard combinatorial problem.To solve the problem,an adaptive genetic algorithm based on cluster search,which is divided into two phases,is put forward.In the first phase,according to the density,all individuals can be homogeneously scattered over the whole solution space through crossover and mutation and better individuals are collected as candidate cluster centres.In the second phase,the search is confined to the neighbourhood of some selected possible solutions to accurately solve with cluster radius decreasing slowly,meanwhile all clusters continuously move to better regions until all the peaks in the question space is searched.This algorithm can efficiently solve the combination problem.Taking the optimization on decision-making of aircraft maintenance by the algorithm for an example,maintenance which combines multiple parts or tasks can significantly enhance economic benefit when the halt cost is rather high.

Part Orientation Optimization in Three Dimensional Printing

Part orientation is one of the important factors in manufacturing with three dimensional printing (3DP) of rapid prototyping system. The minimum building time, being achieved by the part’s minimum building height and minimum width in 3DP’s y direction, is set as the objective of optimizing the part orientation. Genetic algorithm is adopted to find the minimum building height of parts and then ergodic searching is used to get the minimum width of part in y direction. A corresponding program for optimizing part orientation in 3DP was developed. Tests show the proposed approach is effective and fast.

悬臂式掘进机截齿系统维修与备件库存联合决策

为延长悬臂式掘进机截齿系统的使用寿命,降低运维费用,根据截齿系统的维修需求以及随之产生的备件需求,提出了一种基于定期检查的(m,τ,S,s)预防性维修与备件库存联合策略。以费用率最低为目标,考虑可用度约束,建立了截齿系统的维修与备件库存联合决策模型,以决策最优的预防性更换阈值m、检查周期τ、目标库存水平S和备件订购阈值s。基于工程数据,通过遗传算法获得了最优目标以及对应的决策变量,并对模型参数进行敏感性分析,验证了模型的有效性。结果表明,当预防性更换阈值为8把,检查周期为12 d,目标库存水平为75把,备件订购阈值为45把时,可用度为0.87,最低费用率为162.45元/h,截齿系统实现了全局最优的联合决策。因此,将维修与备件库存进行合理的整合优化,可进一步提高截齿系统运行的安全性和经济性。

基于改进粒子群算法的多约束随船备件配置优化

为合理配置船舶随船备件,构建了多约束条件下的备件配置优化模型,提出了一种改进的粒子群算法对模型求解。首先,以备件费用、存储空间、重量为约束条件,以最大化系统备件保障概率为优化目标,通过构造含罚函数的适应度函数,将多约束最优化问题转化为无约束最优化问题;其次,设计了一种基于模拟退火改进的粒子群算法求解优化模型;最后,通过开展仿真实验,证明了该方法在求解随船备件配置优化问题的有效性和优越性。

基于GA-灰色BP神经网络船舶备件需求预测模型

船舶维修备件的及时保障是船舶设备正常生产作业的保证,备件需求量的准确预测对于航运公司降低运营成本、提高管理效率起着至关重要的作用。文章从某船设备维修备件的历史数据进行分析,构建一种基于灰色模型的多层前馈(BP)神经网络模型,并采用遗传算法(GA)进行优化。首先对船舶设备备件需求影响因素进行分析,通过灰色模型确定备件需求的关键指标,并将结果作为BP神经网络的输入层,从而输出需求预测值。与灰色模型、GA-BP模型预测值对比发现,基于GA-灰色BP神经网络组合模型对于船舶备件的需求预测精度误差仅为0.147%。结果表明,使用GA优化可以提高灰色BP神经网络的预测精度,为船舶维修备件需求预测提供了一种新思路。

一种频率-相位混合编码的MIMO雷达波形设计

针对MIMO雷达中常规单载频子脉冲相位编码信号的频带较窄、距离分辨率不高,以及线性调频-相位编码信号模糊函数有“栅瓣”的问题,提出了一种基于置换矩阵时频分布的线性调频-相位编码(perm-LFM-PC,PLP)混合波形的正交波形集设计方法。由于置换矩阵时频分布调制方法引入置换序列调频编码,使得PLP波形的峰值旁瓣(ASP)和互相关峰值(CP)有所升高。为此,以ASP和CP的总和为代价函数,建立PLP波形集优化模型,并提出一种多种群优化遗传算法(MSO-GA),对调频编码序列和相位编码序列进行联合优化。仿真结果表明,PLP波形具有易于实现宽频带的优点和图钉状模糊函数,经过算法优化后的PLP波形集的ASP和CP较低,适用于MIMO雷达系统。

针对任务区域保障的雷达组网优化部署研究

针对任务区域保障的雷达组网优化部署原则,建立雷达组网模型,在此基础上采用传统遗传算法对该模型进行仿真验证。为弥补不足并真正可以支撑实际区域保障任务应用,针对传统遗传算法进行改进,引入免疫遗传算法并仿真验证,证明免疫遗传算法针对雷达组网优化部署的可行性和优越性,并针对免疫遗传算法的不足进行两次改进。实验证明二次改进免疫遗传算法远远优于传统遗传算法和免疫遗传算法,能有效提高雷达组网优化部署效率,使雷达组网雷达威力探测最大空间效能提高,算法收敛速度快,满足实际的约束条件,具有较高的实际区域保障任务应用价值。

考虑并行制造的云服务外购件供应商组合优选

云制造环境下,为解决并行结构下外购件供应商组合优选难的问题,在考虑并行制造的前提下,课题组以时间、成本、可靠性和灵活性为优化目标,构建了云服务多目标组合优选模型。提出了一种改进的非支配排序遗传算法(NSGA-Ⅲ-TS)对模型进行求解,该算法采用反向学习策略提高初始解的质量;而后使用融合邻域搜索(采用交换、反转和插入3种方式)和禁忌搜索算法的策略加强局部搜索,提高算法后期局部寻优能力;对得到的一系列优选供应商组合综合评价,在考虑实际需求的基础上为需求方选出了最优供应商组合及备选方案。通过实际案例验证了该模型的有效性和算法的可行性。

改进遗传算法的飞机结构件装配序列优化研究

为获取最佳的装配序列,提高飞机装配效率,扩展装配序列多样性,设计改进遗传算法的飞机结构件装配序列优化方法,以此作为飞机结构件装配序列优化的评价指标并设计对应的约束条件。依据评价指标及对应的约束条件,建立装配序列优化的目标函数;利用改进遗传算法求解目标函数的解。实验证明:该方法可有效优化装配序列,提高装配效率;在同种群容量时,该方法获取的可行装配序列数量较多,即装配序列多样性较佳,且收敛速度较快。

基于改进BP算法的模具零件表面抛光质量预测研究

针对模具零件表面自动化抛光的工艺参数现场调试困难、抛光后模具零件表面质量不一致等问题,提出了基于改进BP算法的模具零件表面抛光质量预测模型。通过采集模具零件表面抛光试验样本参数,构建预测模型输入参数集,将混沌理论、动态权重、动态学习因子和高斯变异策略引入粒子群优化算法(PSO),利用改进后的粒子群优化算法(IPSO)对BP算法中权值和阈值的更新策略进行优化,并构建了基于IPSO-BP算法的模具零件表面抛光质量预测模型,结合快速非支配排序遗传算法(NS-GA-II)建立多目标优化模型,实现对模具零件表面抛光质量的高精度预测以及抛光工艺参数的优化,对比5种常规预测模型,结果表明基于IPSO-BP算法的预测模型具有较高的预测精度。