The research on optimization methods for constellation launch deployment strategies focused on the consideration of mission interval time constraints at the launch site.Firstly,a dynamic modeling of the constellation ...The research on optimization methods for constellation launch deployment strategies focused on the consideration of mission interval time constraints at the launch site.Firstly,a dynamic modeling of the constellation deployment process was established,and the relationship between the deployment window and the phase difference of the orbit insertion point,as well as the cost of phase adjustment after orbit insertion,was derived.Then,the combination of the constellation deployment position sequence was treated as a parameter,together with the sequence of satellite deployment intervals,as optimization variables,simplifying a highdimensional search problem within a wide range of dates to a finite-dimensional integer programming problem.An improved genetic algorithm with local search on deployment dates was introduced to optimize the launch deployment strategy.With the new description of the optimization variables,the total number of elements in the solution space was reduced by N orders of magnitude.Numerical simulation confirms that the proposed optimization method accelerates the convergence speed from hours to minutes.展开更多
The level of genetic variation within a breeding population affects the effectiveness of selection strategies for genetic improvement.The relationship between genetic variation level within Pinus tabuliformis breeding...The level of genetic variation within a breeding population affects the effectiveness of selection strategies for genetic improvement.The relationship between genetic variation level within Pinus tabuliformis breeding populations and selection strategies or selection effectiveness is not fully investigated.Here,we compared the selection effectiveness of combined and individual direct selection strategies using half-and full-sib families produced from advanced-generation P.tabuliformis seed orchard as our test populations.Our results revealed that,within half-sib families,average diameter at breast height(DBH),tree height,and volume growth of superior individuals selected by the direct selection strategy were higher by 7.72%,18.56%,and 31.01%,respectively,than those selected by the combined selection strategy.Furthermore,significant differences(P<0.01)were observed between the two strategies in terms of the expected genetic gains for average tree height and volume.In contrast,within full-sib families,the differences in tree average DBH,height,and volume between the two selection strategies were relatively minor with increase of 0.17%,2.73%,and 2.21%,respectively,and no significant differences were found in the average expected genetic gains for the studied traits.Half-sib families exhibited greater phenotypic and genetic variation,resulting in improved selection efficiency with the direct selection strategy but also introduced a level of inbreeding risk.Based on genetic distance estimates using molecular markers,our comparative seed orchard design analysis showed that the Improved Adaptive Genetic Programming Algorithm(IAPGA)reduced the average inbreeding coefficient by 14.36% and 14.73% compared to sequential and random designs,respectively.In conclusion,the combination of the direct selection strategy with IAPGA seed orchard design aimed at minimizing inbreeding offered an efficient approach for establishing advanced-generation P.tabuliformis seed orchards.展开更多
This work addresses the cut order planning(COP)problem for multi-color garment production,which is the first step in the clothing industry.First,a multi-objective optimization model of multicolor COP(MCOP)is establish...This work addresses the cut order planning(COP)problem for multi-color garment production,which is the first step in the clothing industry.First,a multi-objective optimization model of multicolor COP(MCOP)is established with production error and production cost as optimization objectives,combined with constraints such as the number of equipment and the number of layers.Second,a decoupled multi-objective optimization algorithm(DMOA)is proposed based on the linear programming decoupling strategy and non-dominated sorting in genetic algorithmsⅡ(NSGAII).The size-combination matrix and the fabric-layer matrix are decoupled to improve the accuracy of the algorithm.Meanwhile,an improved NSGAII algorithm is designed to obtain the optimal Pareto solution to the MCOP problem,thereby constructing a practical intelligent production optimization algorithm.Finally,the effectiveness and superiority of the proposed DMOA are verified through practical cases and comparative experiments,which can effectively optimize the production process for garment enterprises.展开更多
针对智慧医疗场景中高密度无线体域网(wireless body area network,WBAN)多优先级数据传输与计算资源受限的挑战,研究提出一种融合动态优先级评估与量子优化的任务卸载策略。首先通过构建医疗物联网(healthcare Internet of Things,H-I...针对智慧医疗场景中高密度无线体域网(wireless body area network,WBAN)多优先级数据传输与计算资源受限的挑战,研究提出一种融合动态优先级评估与量子优化的任务卸载策略。首先通过构建医疗物联网(healthcare Internet of Things,H-IoT)高密度WBAN网络模型,集成任务优先级分层机制与动态信道状态感知模块,建立基于生理数据特征的通信质量评估体系。其次设计多维动态调度框架,利用生理参数偏离度、数据滞留时间及抢占事件等指标实时调整任务优先级权重,结合抢占式调度策略保障急诊数据的低时延传输。再进一步改进量子遗传算法(improved quantum genetic algorithm,IQGA),采用动态量子旋转门角度调整机制优化局部搜索性能,并引入灾变修正函数提升全局收敛效率。仿真实验表明,该策略在任务平均处理时间、系统能耗、高优先级任务时延及收敛速度方面分别实现71.51%、88.21%、89.63%和78.74%的性能优化,系统综合收益提升达114.43%。研究成果为高密度医疗物联网场景下的实时任务调度与资源分配提供了理论支撑与技术路径。展开更多
研究单转运系统分布式置换流水线调度问题,任一工厂内连续两台机器间有一台运输能力有限的转运机器人。基于此,提出一种多策略融合改进遗传算法以最小化最大完工时间。引入Logistic-tent混沌搜索、基于K-均值聚类的NEH算法和修正NEH算...研究单转运系统分布式置换流水线调度问题,任一工厂内连续两台机器间有一台运输能力有限的转运机器人。基于此,提出一种多策略融合改进遗传算法以最小化最大完工时间。引入Logistic-tent混沌搜索、基于K-均值聚类的NEH算法和修正NEH算法以改善初始工厂加工序列群的质量,运用结合均匀多点交叉和互换变异的自适应交叉变异算子或工厂内/间交叉变异算子进行解的调整,设计一种基于主工厂的邻域搜索(key-factory-based local search,KFLS)和半初始化策略进行再次优化。仿真结果表明了该算法的有效性。展开更多
文摘The research on optimization methods for constellation launch deployment strategies focused on the consideration of mission interval time constraints at the launch site.Firstly,a dynamic modeling of the constellation deployment process was established,and the relationship between the deployment window and the phase difference of the orbit insertion point,as well as the cost of phase adjustment after orbit insertion,was derived.Then,the combination of the constellation deployment position sequence was treated as a parameter,together with the sequence of satellite deployment intervals,as optimization variables,simplifying a highdimensional search problem within a wide range of dates to a finite-dimensional integer programming problem.An improved genetic algorithm with local search on deployment dates was introduced to optimize the launch deployment strategy.With the new description of the optimization variables,the total number of elements in the solution space was reduced by N orders of magnitude.Numerical simulation confirms that the proposed optimization method accelerates the convergence speed from hours to minutes.
基金financially supported by the Biological BreedingNational Science and Technology Major Project(2023ZD0405806)the National Key R&D Program for the 14th Five-Year Plan in China(2022YFD2200304).
文摘The level of genetic variation within a breeding population affects the effectiveness of selection strategies for genetic improvement.The relationship between genetic variation level within Pinus tabuliformis breeding populations and selection strategies or selection effectiveness is not fully investigated.Here,we compared the selection effectiveness of combined and individual direct selection strategies using half-and full-sib families produced from advanced-generation P.tabuliformis seed orchard as our test populations.Our results revealed that,within half-sib families,average diameter at breast height(DBH),tree height,and volume growth of superior individuals selected by the direct selection strategy were higher by 7.72%,18.56%,and 31.01%,respectively,than those selected by the combined selection strategy.Furthermore,significant differences(P<0.01)were observed between the two strategies in terms of the expected genetic gains for average tree height and volume.In contrast,within full-sib families,the differences in tree average DBH,height,and volume between the two selection strategies were relatively minor with increase of 0.17%,2.73%,and 2.21%,respectively,and no significant differences were found in the average expected genetic gains for the studied traits.Half-sib families exhibited greater phenotypic and genetic variation,resulting in improved selection efficiency with the direct selection strategy but also introduced a level of inbreeding risk.Based on genetic distance estimates using molecular markers,our comparative seed orchard design analysis showed that the Improved Adaptive Genetic Programming Algorithm(IAPGA)reduced the average inbreeding coefficient by 14.36% and 14.73% compared to sequential and random designs,respectively.In conclusion,the combination of the direct selection strategy with IAPGA seed orchard design aimed at minimizing inbreeding offered an efficient approach for establishing advanced-generation P.tabuliformis seed orchards.
基金Supported by the Natural Science Foundation of Zhejiang Province(No.LQ22F030015).
文摘This work addresses the cut order planning(COP)problem for multi-color garment production,which is the first step in the clothing industry.First,a multi-objective optimization model of multicolor COP(MCOP)is established with production error and production cost as optimization objectives,combined with constraints such as the number of equipment and the number of layers.Second,a decoupled multi-objective optimization algorithm(DMOA)is proposed based on the linear programming decoupling strategy and non-dominated sorting in genetic algorithmsⅡ(NSGAII).The size-combination matrix and the fabric-layer matrix are decoupled to improve the accuracy of the algorithm.Meanwhile,an improved NSGAII algorithm is designed to obtain the optimal Pareto solution to the MCOP problem,thereby constructing a practical intelligent production optimization algorithm.Finally,the effectiveness and superiority of the proposed DMOA are verified through practical cases and comparative experiments,which can effectively optimize the production process for garment enterprises.
文摘针对智慧医疗场景中高密度无线体域网(wireless body area network,WBAN)多优先级数据传输与计算资源受限的挑战,研究提出一种融合动态优先级评估与量子优化的任务卸载策略。首先通过构建医疗物联网(healthcare Internet of Things,H-IoT)高密度WBAN网络模型,集成任务优先级分层机制与动态信道状态感知模块,建立基于生理数据特征的通信质量评估体系。其次设计多维动态调度框架,利用生理参数偏离度、数据滞留时间及抢占事件等指标实时调整任务优先级权重,结合抢占式调度策略保障急诊数据的低时延传输。再进一步改进量子遗传算法(improved quantum genetic algorithm,IQGA),采用动态量子旋转门角度调整机制优化局部搜索性能,并引入灾变修正函数提升全局收敛效率。仿真实验表明,该策略在任务平均处理时间、系统能耗、高优先级任务时延及收敛速度方面分别实现71.51%、88.21%、89.63%和78.74%的性能优化,系统综合收益提升达114.43%。研究成果为高密度医疗物联网场景下的实时任务调度与资源分配提供了理论支撑与技术路径。
文摘研究单转运系统分布式置换流水线调度问题,任一工厂内连续两台机器间有一台运输能力有限的转运机器人。基于此,提出一种多策略融合改进遗传算法以最小化最大完工时间。引入Logistic-tent混沌搜索、基于K-均值聚类的NEH算法和修正NEH算法以改善初始工厂加工序列群的质量,运用结合均匀多点交叉和互换变异的自适应交叉变异算子或工厂内/间交叉变异算子进行解的调整,设计一种基于主工厂的邻域搜索(key-factory-based local search,KFLS)和半初始化策略进行再次优化。仿真结果表明了该算法的有效性。
