摘要
针对目前基于单服务器的测井曲线自动处理系统的缺陷,研究了高性能测井曲线自动处理集群架构与实现。该集群通过高速以太网连接,节点采用对等结构构件系统,通过分布式哈希表进行数据发布、负载均衡及系统自动扩容。该系统采用分布式处理技术提高了计算能力,采用反馈神经网络进行计算参数选优,又提高了数据处理的准确性。现场实践表明,该系统能够提高系统整体容量与处理能力。
In order to overcome the disadvantages of the logging curve automatic processing systems based on a single server, the research and implementation of the high powered logging curve automatic processing clusters are presented. The systems can employ HSE （High Speed Ethernet） to connect hosts, which construct a Peer-to-Peer system; and DHT （Distributed Hash Table） is utilized to balance load and to issue data. The distributed technologies that the system adopted extent the computing and storage capacity; feedback neural networks optimize parameter selections to improve the veracity. Applications show that this system can increase the system processing capacity.
出处
《重庆邮电学院学报（自然科学版）》
2006年第5期669-672,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
关键词
测井曲线
自动处理
分布式
集群
logging curve
automatic processing
distributed
cluster
