Photovoltaic(PV)systems utilize maximum power point tracking(MPPT)controllers to optimize power output amidst varying environmental conditions.However,the presence of multiple peaks resulting from partial shading pose...Photovoltaic(PV)systems utilize maximum power point tracking(MPPT)controllers to optimize power output amidst varying environmental conditions.However,the presence of multiple peaks resulting from partial shading poses a challenge to the tracking operation.Under partial shade conditions,the global maximum power point(GMPP)may be missed by most traditional maximum power point tracker.The flower pollination algorithm(FPA)and particle swarm optimization(PSO)are two examples of metaheuristic techniques that can be used to solve the issue of failing to track the GMPP.This paper discusses and resolves all issues associated with using the standard FPA method as the MPPT for PV systems.The first issue is that the initial values of pollen are determined randomly at first,which can lead to premature convergence.To minimize the convergence time and enhance the possibility of detecting the GMPP,the initial pollen values were modified so that they were near the expected peak positions.Secondly,in the modified FPA,population fitness and switch probability values both influence swapping between two-mode optimization,which may improve the flower pollination algorithm’s tracking speed.The performance of the modified flower pollination algorithm(MFPA)is assessed through a comparison with the perturb and observe(P&O)method and the standard FPA method.The simulation results reveal that under different partial shading conditions,the tracking time for MFPA is 0.24,0.24,0.22,and 0.23 s,while for FPA,it is 0.4,0.35,0.45,and 0.37 s.Additionally,the simulation results demonstrate that MFPA achieves higher MPPT efficiency in the same four partial shading conditions,with values of 99.98%,99.90%,99.93%,and 99.26%,compared to FPA with MPPT efficiencies of 99.93%,99.88%,99.91%,and 99.18%.Based on the findings from simulations,the proposed method effectively and accurately tracks the GMPP across a diverse set of environmental conditions.展开更多
为了降低花朵授粉算法(FPA)重复探索的情况,并提高算法的种群多样性和空间搜索能力,提出一种基于神经网络优化的花朵授粉算法(NNFPA)。设定自适应控制因子,从而动态地切换全局与局部搜索;利用多方信息的全局搜索策略提高算法收敛速度并...为了降低花朵授粉算法(FPA)重复探索的情况,并提高算法的种群多样性和空间搜索能力,提出一种基于神经网络优化的花朵授粉算法(NNFPA)。设定自适应控制因子,从而动态地切换全局与局部搜索;利用多方信息的全局搜索策略提高算法收敛速度并维持花粉种群的多样性,同时减少在算法迭代后期种群对社会属性的依赖;基于神经网络的局部搜索策略让算法具有记忆功能,这样算法就能具有稳定搜索策略,从而降低算法的不确定性,使它能更充分地探索解空间。选取9个常规测试函数与CEC2014测试集中的部分函数进行仿真实验,得到的结果表明:与标准FPA以及变种算法HSFPA(FPA based on Hybrid Strategy)相比,NNFPA在所选测试函数上具有较高的搜索精度和收敛速度。可见NNFPA具有更好的寻优能力。展开更多
针对局部配电网中互联微电网系统的调度问题,提出了一种优化调度模型。首先,配电网调度层通过协调各微电网与配电网之间的交互功率来改善配电网的运行状况;然后,微电网层根据配电网层的优化调度结果制定生产计划,并建立基于储能系统荷...针对局部配电网中互联微电网系统的调度问题,提出了一种优化调度模型。首先,配电网调度层通过协调各微电网与配电网之间的交互功率来改善配电网的运行状况;然后,微电网层根据配电网层的优化调度结果制定生产计划,并建立基于储能系统荷电状态(state of charge,SOC)的阶梯式功率修正策略,以保证储能系统的可调度能力。选用全局搜索能力强的改进花朵授粉算法(flower pollination algorithm,FPA)求解配电网层优化调度模型,并以IEEE-14节点系统为算例进行仿真分析。结果表明:改进FPA算法的初始种群质量与收敛速度均有所提高;配电网的网损和微电网交互功率的波动性下降;各储能系统的SOC保持在0.25~0.6区间内。研究结果证明了优化调度模型的电网友好性和改进FPA算法的有效性,且阶梯式功率修正策略可以保证各储能系统的持续可调度能力。展开更多
虽然花授粉算法对于求解优化问题十分有效,但也存在收敛性慢的问题。为了解决此问题,提出一种带有时变因子的差分进化花授粉算法(Differential Evolution Flower Pollination Algorithm with Time Variant Factor,TVDFPA)。对步长因子...虽然花授粉算法对于求解优化问题十分有效,但也存在收敛性慢的问题。为了解决此问题,提出一种带有时变因子的差分进化花授粉算法(Differential Evolution Flower Pollination Algorithm with Time Variant Factor,TVDFPA)。对步长因子进行改进,同时在迭代过程中加入差分进化的策略,通过种群杂交,提高算法的收敛速度和寻优能力。通过标准测试函数进行测试,仿真结果表明TVDFPA的收敛速度比原始花授粉算法、混沌和声的花授粉(HFPCHS)、模拟退火花授粉算法(SFPA)快,收敛精度也有较大提高。进而结合花授粉算法的特点,建立带有变参数的双适应值比较法来求解压力容器设计问题,实验结果表明改进之后的算法具有较好的求解性能。展开更多
文摘Photovoltaic(PV)systems utilize maximum power point tracking(MPPT)controllers to optimize power output amidst varying environmental conditions.However,the presence of multiple peaks resulting from partial shading poses a challenge to the tracking operation.Under partial shade conditions,the global maximum power point(GMPP)may be missed by most traditional maximum power point tracker.The flower pollination algorithm(FPA)and particle swarm optimization(PSO)are two examples of metaheuristic techniques that can be used to solve the issue of failing to track the GMPP.This paper discusses and resolves all issues associated with using the standard FPA method as the MPPT for PV systems.The first issue is that the initial values of pollen are determined randomly at first,which can lead to premature convergence.To minimize the convergence time and enhance the possibility of detecting the GMPP,the initial pollen values were modified so that they were near the expected peak positions.Secondly,in the modified FPA,population fitness and switch probability values both influence swapping between two-mode optimization,which may improve the flower pollination algorithm’s tracking speed.The performance of the modified flower pollination algorithm(MFPA)is assessed through a comparison with the perturb and observe(P&O)method and the standard FPA method.The simulation results reveal that under different partial shading conditions,the tracking time for MFPA is 0.24,0.24,0.22,and 0.23 s,while for FPA,it is 0.4,0.35,0.45,and 0.37 s.Additionally,the simulation results demonstrate that MFPA achieves higher MPPT efficiency in the same four partial shading conditions,with values of 99.98%,99.90%,99.93%,and 99.26%,compared to FPA with MPPT efficiencies of 99.93%,99.88%,99.91%,and 99.18%.Based on the findings from simulations,the proposed method effectively and accurately tracks the GMPP across a diverse set of environmental conditions.
文摘为了降低花朵授粉算法(FPA)重复探索的情况,并提高算法的种群多样性和空间搜索能力,提出一种基于神经网络优化的花朵授粉算法(NNFPA)。设定自适应控制因子,从而动态地切换全局与局部搜索;利用多方信息的全局搜索策略提高算法收敛速度并维持花粉种群的多样性,同时减少在算法迭代后期种群对社会属性的依赖;基于神经网络的局部搜索策略让算法具有记忆功能,这样算法就能具有稳定搜索策略,从而降低算法的不确定性,使它能更充分地探索解空间。选取9个常规测试函数与CEC2014测试集中的部分函数进行仿真实验,得到的结果表明:与标准FPA以及变种算法HSFPA(FPA based on Hybrid Strategy)相比,NNFPA在所选测试函数上具有较高的搜索精度和收敛速度。可见NNFPA具有更好的寻优能力。
文摘针对局部配电网中互联微电网系统的调度问题,提出了一种优化调度模型。首先,配电网调度层通过协调各微电网与配电网之间的交互功率来改善配电网的运行状况;然后,微电网层根据配电网层的优化调度结果制定生产计划,并建立基于储能系统荷电状态(state of charge,SOC)的阶梯式功率修正策略,以保证储能系统的可调度能力。选用全局搜索能力强的改进花朵授粉算法(flower pollination algorithm,FPA)求解配电网层优化调度模型,并以IEEE-14节点系统为算例进行仿真分析。结果表明:改进FPA算法的初始种群质量与收敛速度均有所提高;配电网的网损和微电网交互功率的波动性下降;各储能系统的SOC保持在0.25~0.6区间内。研究结果证明了优化调度模型的电网友好性和改进FPA算法的有效性,且阶梯式功率修正策略可以保证各储能系统的持续可调度能力。
文摘虽然花授粉算法对于求解优化问题十分有效,但也存在收敛性慢的问题。为了解决此问题,提出一种带有时变因子的差分进化花授粉算法(Differential Evolution Flower Pollination Algorithm with Time Variant Factor,TVDFPA)。对步长因子进行改进,同时在迭代过程中加入差分进化的策略,通过种群杂交,提高算法的收敛速度和寻优能力。通过标准测试函数进行测试,仿真结果表明TVDFPA的收敛速度比原始花授粉算法、混沌和声的花授粉(HFPCHS)、模拟退火花授粉算法(SFPA)快,收敛精度也有较大提高。进而结合花授粉算法的特点,建立带有变参数的双适应值比较法来求解压力容器设计问题,实验结果表明改进之后的算法具有较好的求解性能。
