科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2015年
6期
100-102
,共3页
云计算%资源调度%经验模态分解%预失真估计
雲計算%資源調度%經驗模態分解%預失真估計
운계산%자원조도%경험모태분해%예실진고계
cloud computing%resource scheduling%empirical mode decomposition%pre distortion estimation
在云计算构建下实现对旅游信息的优化调度,提高旅游资源数据的开发和利用效率。传统的旅游信息资源采用经验模态分解的调度算法,通过人工分区实现信息资源的调度模型具有随机性和无组织性,资源分配效率不高。提出一种基于逆模型参数预失真估计的云计算架构下旅游信息资源优化调度模型。设计旅游资源信息调度网络,云计算构架的逆模型网络结构采用BP网络的输出结构,训练包括前向传播信息,反向传播误差两个过程。设计循环堆栈的约束指向矢量,使旅游资源信息调度网络的路由节点的资源分配输出误差最小。采用三亚的5个旅游景点作为研究对象,以2003~2014年度的游客数据作为模型参量,仿真结果表明,模型对旅游资源的融合性能较好,有效剔除簇内非相干数据,提高数据融合效率,分层调度融合度能达到90%以上,优化调度性能优越。
在雲計算構建下實現對旅遊信息的優化調度,提高旅遊資源數據的開髮和利用效率。傳統的旅遊信息資源採用經驗模態分解的調度算法,通過人工分區實現信息資源的調度模型具有隨機性和無組織性,資源分配效率不高。提齣一種基于逆模型參數預失真估計的雲計算架構下旅遊信息資源優化調度模型。設計旅遊資源信息調度網絡,雲計算構架的逆模型網絡結構採用BP網絡的輸齣結構,訓練包括前嚮傳播信息,反嚮傳播誤差兩箇過程。設計循環堆棧的約束指嚮矢量,使旅遊資源信息調度網絡的路由節點的資源分配輸齣誤差最小。採用三亞的5箇旅遊景點作為研究對象,以2003~2014年度的遊客數據作為模型參量,倣真結果錶明,模型對旅遊資源的融閤性能較好,有效剔除簇內非相榦數據,提高數據融閤效率,分層調度融閤度能達到90%以上,優化調度性能優越。
재운계산구건하실현대여유신식적우화조도,제고여유자원수거적개발화이용효솔。전통적여유신식자원채용경험모태분해적조도산법,통과인공분구실현신식자원적조도모형구유수궤성화무조직성,자원분배효솔불고。제출일충기우역모형삼수예실진고계적운계산가구하여유신식자원우화조도모형。설계여유자원신식조도망락,운계산구가적역모형망락결구채용BP망락적수출결구,훈련포괄전향전파신식,반향전파오차량개과정。설계순배퇴잔적약속지향시량,사여유자원신식조도망락적로유절점적자원분배수출오차최소。채용삼아적5개여유경점작위연구대상,이2003~2014년도적유객수거작위모형삼량,방진결과표명,모형대여유자원적융합성능교호,유효척제족내비상간수거,제고수거융합효솔,분층조도융합도능체도90%이상,우화조도성능우월。
To realize the optimization of scheduling in cloud computing on tourism information construction, improve the de?velopment of tourism resources and the utilization efficiency of data. Scheduling algorithm of empirical mode decomposition using traditional tourism information resources, through the scheduling model of artificial partition to realize information re?source is random and has no organization, resource allocation efficiency is not high. Presented a method for calculating the tourism information resource optimization scheduling model under the framework of distortion estimation inverse model pa?rameters based on pre cloud. Design of tourist resources information scheduling network, output structure of cloud inverse network structure model computing architectures using BP network, training including prior to the dissemination of informa?tion. Using 5 tourist attractions in Sanya as the research object, to 2003~2014 annual visitors data as the model parame?ters, simulation results show that the fusion performance model of tourism resources better, effectively eliminate the intra cluster coherent data, improve the efficiency of data fusion, a hierarchical scheduling fusion degree can reach above 90%, the optimal scheduling of superior performance.