针对汽轮机变工况运行存在负荷偏差的问题,提出一种基于差分进化算法(differential evolution,DE)和多标签随机森林(multi-label random forest,MLRF)结合的汽轮机负荷偏差原因分类模型。通过斯皮尔曼(Spearman)相关性系数,分析影响汽...针对汽轮机变工况运行存在负荷偏差的问题,提出一种基于差分进化算法(differential evolution,DE)和多标签随机森林(multi-label random forest,MLRF)结合的汽轮机负荷偏差原因分类模型。通过斯皮尔曼(Spearman)相关性系数,分析影响汽轮机负荷出力的相关联变量;采用DE算法优化MLRF模型参数,建立基于DE-MLRF的汽轮机负荷偏差原因分类模型。结合某660 MW汽轮机实际运行数据进行实验验证,结果表明,与其他7种算法相比,DE算法优化的MLRF模型误报率(1.9024%)、漏报率(1.8541%)最低,可为汽轮机负荷偏差原因定位提供决策支持。展开更多
de Bruijn序列结构是一个查寻表,其核心是它的表标签。因此构造出查寻表标签对于生成de Bruijn序列十分重要。给出一种定值标签构造法,即对大多数节点设定为定值标签,对少部分节点则根据查寻表标签的必要条件指定或任意选定标签。该方...de Bruijn序列结构是一个查寻表,其核心是它的表标签。因此构造出查寻表标签对于生成de Bruijn序列十分重要。给出一种定值标签构造法,即对大多数节点设定为定值标签,对少部分节点则根据查寻表标签的必要条件指定或任意选定标签。该方法构造的查寻表标签数随着m,n增长而成指数式增长。在局部看是有效的,但与查寻表标签本身数目的惊人增长比较起来就很渺小了。该方法在目前缺乏更好的方法的情况下还是最有效的。展开更多
de Bruijn序列的结构是一个查寻表,其核心是它的表标签。因此构造出查寻表标签对于生成de Bruijn序列十分重要。给出两种k位修正构造法。方法1为k位提升构造法,即对大部分节点将其第k(k=1,2,…,n-1)位提升一个定值c(1≤c≤m),来作为该...de Bruijn序列的结构是一个查寻表,其核心是它的表标签。因此构造出查寻表标签对于生成de Bruijn序列十分重要。给出两种k位修正构造法。方法1为k位提升构造法,即对大部分节点将其第k(k=1,2,…,n-1)位提升一个定值c(1≤c≤m),来作为该节点的标签。方法2为k位收缩构造法,即对大部分节点将其第k(k=1,2,…,n-1)位向定值(r0≤r≤m)收缩,来作为该节点的标签。这些方法构造的查寻表标签数随着m,n增长而成指数式增长。与定值构造法一样,在局部看是有效的,但与查寻表标签本身数目的惊人增长比较起来就很渺小。方法2与定值标签构造法比较其速度提高了关于m,n的指数式倍。展开更多
A large number of community discovery algorithms have been proposed in the last decade. Recently, the sharp increase of network scale has become a great challenge for traditional community discovery algorithms. Label ...A large number of community discovery algorithms have been proposed in the last decade. Recently, the sharp increase of network scale has become a great challenge for traditional community discovery algorithms. Label propagation algorithm is a semi-supervised machine learning method, which has linear time complexity when coping with large scale networks. However, the output result has less stability and the quality of the output communities still remains to be improved. Therefore, we propose a novel coreleader based label propagation algorithm for community detection called CLBLPA. Firstly, we find core leaders of potential community by using a greedy method. Then we utilize the label influence potential to guide the process of label propagation. Thus we can accelerate the convergence of algorithm and improve the stability of the output. Experimental results on synthetic datasets and real networks show that CLBLPA can significantly improve the quality of the output communities.展开更多
文摘针对汽轮机变工况运行存在负荷偏差的问题,提出一种基于差分进化算法(differential evolution,DE)和多标签随机森林(multi-label random forest,MLRF)结合的汽轮机负荷偏差原因分类模型。通过斯皮尔曼(Spearman)相关性系数,分析影响汽轮机负荷出力的相关联变量;采用DE算法优化MLRF模型参数,建立基于DE-MLRF的汽轮机负荷偏差原因分类模型。结合某660 MW汽轮机实际运行数据进行实验验证,结果表明,与其他7种算法相比,DE算法优化的MLRF模型误报率(1.9024%)、漏报率(1.8541%)最低,可为汽轮机负荷偏差原因定位提供决策支持。
基金supported by the National Natural Science Foundation of China under Grant No. 61272277, 41301409, 41571390the Fundamental Research Funds for the Central Universities under Grant No. 274742
文摘A large number of community discovery algorithms have been proposed in the last decade. Recently, the sharp increase of network scale has become a great challenge for traditional community discovery algorithms. Label propagation algorithm is a semi-supervised machine learning method, which has linear time complexity when coping with large scale networks. However, the output result has less stability and the quality of the output communities still remains to be improved. Therefore, we propose a novel coreleader based label propagation algorithm for community detection called CLBLPA. Firstly, we find core leaders of potential community by using a greedy method. Then we utilize the label influence potential to guide the process of label propagation. Thus we can accelerate the convergence of algorithm and improve the stability of the output. Experimental results on synthetic datasets and real networks show that CLBLPA can significantly improve the quality of the output communities.
