In this note we construct certain sequences of finite point sets in [0, 1) s (s ≥ 1) and give the upper bounds of their discrepancy. Furthermore we prove that these sequences are uniformly distrbuted in [0, 1) s .
锅炉燃烧优化在电厂锅炉经济稳定运行中起着重要作用,NO_(x)排放预测是其中的一个基本环节,因此提出了一种基于改进蜣螂优化算法优化卷积神经网络(convolutional neural network,CNN)与双向长短期记忆神经网络(long short term memory,L...锅炉燃烧优化在电厂锅炉经济稳定运行中起着重要作用,NO_(x)排放预测是其中的一个基本环节,因此提出了一种基于改进蜣螂优化算法优化卷积神经网络(convolutional neural network,CNN)与双向长短期记忆神经网络(long short term memory,LSTM)的组合模型超参数的超超临界锅炉NO_(x)排放预测的方法。首先通过Pearson相关性判定与NO_(x)排放相关的特征参数;其次建立CNN-LSTM预测模型,利用卷积神经网络CNN提取分层数据结构,长短期记忆网络挖掘长期依赖关系,然后结合佳点集、t分布变异策略对蜣螂算法进行改进,用改进后的算法对LSTM超参数进行优化得到最终预测模型;最后与其他神经网络模型进行对比验证。以某660 MW机组锅炉深度调峰实际数据进行预测,结果得到NO_(x)排放浓度实际值与预测值的平均绝对误差为3.3516,平均相对误差为2.4667,数据结果表明该预测模型具有更准确的预测效果。展开更多
为改善分布式电源(Distributed Generation,DG)并入电网后配电网重构算法的性能,提出一种基于佳点集的蜜蜂进化型遗传算法(Bee Evolutionary Genetic Algorithm Based on Good Point Set,GBEGA)。该算法的关键有三点:1.提出一种基于佳...为改善分布式电源(Distributed Generation,DG)并入电网后配电网重构算法的性能,提出一种基于佳点集的蜜蜂进化型遗传算法(Bee Evolutionary Genetic Algorithm Based on Good Point Set,GBEGA)。该算法的关键有三点:1.提出一种基于佳点集的种群初始化方法,该方法比随机方法产生的种群在搜索空间更为均匀;2.引进佳点集交叉算子,该算子能在父代附近进行更加精细的搜索;3.采用自适应的交叉变异概率,有利于算法开采与勘探的平衡。将DG处理为PQ、PV两种模型,并将GBEGA与相关文献中的算法关于IEEE33和IEEE69节点系统进行了对比测试。仿真结果表明,GBEGA适合于含DG的配电网重构,在全局寻优能力和收敛速度上表现出色。展开更多
With the integral-level approach to global optimization, a class of discon-tinuous penalty functions is proposed to solve constrained minimization problems. Inthis paper we propose an implementable algorithm by means ...With the integral-level approach to global optimization, a class of discon-tinuous penalty functions is proposed to solve constrained minimization problems. Inthis paper we propose an implementable algorithm by means of the good point set ofuniform distribution which conquers the default of Monte-Carlo method. At last weprove the convergence of the implementable algorithm.展开更多
基金Supported by the National Natural Science Foundation of Chinathe "333 Project" Foundation of Jiangsu Province of China
文摘In this note we construct certain sequences of finite point sets in [0, 1) s (s ≥ 1) and give the upper bounds of their discrepancy. Furthermore we prove that these sequences are uniformly distrbuted in [0, 1) s .
文摘锅炉燃烧优化在电厂锅炉经济稳定运行中起着重要作用,NO_(x)排放预测是其中的一个基本环节,因此提出了一种基于改进蜣螂优化算法优化卷积神经网络(convolutional neural network,CNN)与双向长短期记忆神经网络(long short term memory,LSTM)的组合模型超参数的超超临界锅炉NO_(x)排放预测的方法。首先通过Pearson相关性判定与NO_(x)排放相关的特征参数;其次建立CNN-LSTM预测模型,利用卷积神经网络CNN提取分层数据结构,长短期记忆网络挖掘长期依赖关系,然后结合佳点集、t分布变异策略对蜣螂算法进行改进,用改进后的算法对LSTM超参数进行优化得到最终预测模型;最后与其他神经网络模型进行对比验证。以某660 MW机组锅炉深度调峰实际数据进行预测,结果得到NO_(x)排放浓度实际值与预测值的平均绝对误差为3.3516,平均相对误差为2.4667,数据结果表明该预测模型具有更准确的预测效果。
文摘为改善分布式电源(Distributed Generation,DG)并入电网后配电网重构算法的性能,提出一种基于佳点集的蜜蜂进化型遗传算法(Bee Evolutionary Genetic Algorithm Based on Good Point Set,GBEGA)。该算法的关键有三点:1.提出一种基于佳点集的种群初始化方法,该方法比随机方法产生的种群在搜索空间更为均匀;2.引进佳点集交叉算子,该算子能在父代附近进行更加精细的搜索;3.采用自适应的交叉变异概率,有利于算法开采与勘探的平衡。将DG处理为PQ、PV两种模型,并将GBEGA与相关文献中的算法关于IEEE33和IEEE69节点系统进行了对比测试。仿真结果表明,GBEGA适合于含DG的配电网重构,在全局寻优能力和收敛速度上表现出色。
基金This work is supported by the National Natural Science Foundation of China(grants 19871053)and by the Science and Technology Development Foundation of Shanghai
文摘With the integral-level approach to global optimization, a class of discon-tinuous penalty functions is proposed to solve constrained minimization problems. Inthis paper we propose an implementable algorithm by means of the good point set ofuniform distribution which conquers the default of Monte-Carlo method. At last weprove the convergence of the implementable algorithm.
