电网自动化分布并行操作系统中智能化任务调度的设计与实现
摘要
电力产业是国家的基础能源产业，随着科技的发展，其自动化系统也在不
断的发展。基于对现有应用环境的分析，我们发现具有高可靠性、高性价比、
高可扩展性和高性能的分布式并行计算机自动化系统具有极大的应用前景。为
了获得这样的系统，我们基于Linux系统开发了具有自主版权的分布式并行操
作系统，以便在这个系统上开发电网自动化系统。
本文首先介绍当前的分布并行系统的发展情况和电网自动化系统的发展情
况，从而提出了电网自动化系统对分布并行操作系统任务调度的需求；接着讨
论分布并行操作系统中任务调度的问题、矛盾和常见算法；然后提出了衡量任
务调度算法好坏的几个指标，并以此分析了投标算法的性能，提出了改进的意
见。
本文重点描述了在电网自动化分布并行操作系统中使用的智能化任务调
度，包括任务调度模型、任务调度算法、节点负载衡量和镜像执行的处理。然
后介绍了任务调度的实现情况。
本文最后介绍了系统测试情况。就系统的调度时延、可靠性、有效利用率
和负载平衡情况进行了测试，并给出了测试结果和说明。
The electric power industry is a national foundation industry.
Electric power automation system is developing with the development of
science and technology. The analysis of the electric power automation
system show that the distributed parallel computer automation with high
reliability, high price/performance, and high expandability has open vast
application . We are developing a electric power automation system based
on Linux distributed parallel operating system.
First this thesis presents the development situation of the
distributed parallel syste駃 and electric power automation, and give out
the requirements for the distributed parallel system in the electric power
automation. Then this thesis discusses the task scheduling algorithm in
distributed parallel system, poses several parameters to evaluate task
scheduling algorithm, analysis the performance of bidding algorithm,- and
puts forward some improvement.
This thesis places emphasis on distributing task梥cheduling model,
task scheduling algorithm, load balance and mirror image processing. The
implementation of a task梥cheduling algorithm is also introduced.
The scheduling delay, reliability, utilization efficient and load
balance are tested. The test results are given in the end of this thesis.
断的发展。基于对现有应用环境的分析，我们发现具有高可靠性、高性价比、
高可扩展性和高性能的分布式并行计算机自动化系统具有极大的应用前景。为
了获得这样的系统，我们基于Linux系统开发了具有自主版权的分布式并行操
作系统，以便在这个系统上开发电网自动化系统。
本文首先介绍当前的分布并行系统的发展情况和电网自动化系统的发展情
况，从而提出了电网自动化系统对分布并行操作系统任务调度的需求；接着讨
论分布并行操作系统中任务调度的问题、矛盾和常见算法；然后提出了衡量任
务调度算法好坏的几个指标，并以此分析了投标算法的性能，提出了改进的意
见。
本文重点描述了在电网自动化分布并行操作系统中使用的智能化任务调
度，包括任务调度模型、任务调度算法、节点负载衡量和镜像执行的处理。然
后介绍了任务调度的实现情况。
本文最后介绍了系统测试情况。就系统的调度时延、可靠性、有效利用率
和负载平衡情况进行了测试，并给出了测试结果和说明。
The electric power industry is a national foundation industry.
Electric power automation system is developing with the development of
science and technology. The analysis of the electric power automation
system show that the distributed parallel computer automation with high
reliability, high price/performance, and high expandability has open vast
application . We are developing a electric power automation system based
on Linux distributed parallel operating system.
First this thesis presents the development situation of the
distributed parallel syste駃 and electric power automation, and give out
the requirements for the distributed parallel system in the electric power
automation. Then this thesis discusses the task scheduling algorithm in
distributed parallel system, poses several parameters to evaluate task
scheduling algorithm, analysis the performance of bidding algorithm,- and
puts forward some improvement.
This thesis places emphasis on distributing task梥cheduling model,
task scheduling algorithm, load balance and mirror image processing. The
implementation of a task梥cheduling algorithm is also introduced.
The scheduling delay, reliability, utilization efficient and load
balance are tested. The test results are given in the end of this thesis.
引文
1：王鼎新、郑纬民、沈美明；并行机群的若干关键技术；清华大学学报[自 然科学版]1998年第三卷；1997-10-20。
2：周一萍、郑守淇、白英彩；基于PVM的并行Lisp机制与实现；小型微 型计算机系统，第19卷，第8期：1998年8月。
3: Amnon Barak, Orcn la'adan; The MOSIX Multicomputer Operation System for High Performance Cluster Computing: http://www.cs.huji. ac.il/mosix.
4: Chen-chingLiu, Juhwan Jung, Gerald T.Heydt, Vi jay Vital and Arun G. Phadke; The Strategic Power Infrastructure Defense(SPID) System: [J]IEEE Control Systems Magazine: August 2000, 20(4) :40-52．
5: Juan Carlos Gomez, M. M. Morcos; Effect of Distribution System Protection on Voltage Sags; [J]IEEE Power Engineering Review, May 2000, 20(5) :66-68．
6: S. Jovanovic, F.Millicevic. Triangular Distribution Load Flow. [J]IEEE Power Engineering Review, May 2000, 20(5) :60-62．
7：赵谷泉；变电所综合自动化的结构模式及通信总线选择：[J]电力自动 化设备，1999，19(5) ：55-56。
8：王学超，王玉生：分布式配电网调度与管理自动化系统的设计开发：[J] 电力系统自动化。1999，23(7) ：40-42。
9：《地区电网调度自动化设计技术规程》(DL5002-91) 。
10：《电网调度自动化系统实用化要求》(能源调自[1988] 30号)。
11: Mukesh singhal, Niranjan G-shivaratri; Advanced Concepts in Operating System, Distributed, Database, And Multiprocessor Operating Systems, McGraw-Hid, INC; 1994．
12: GramaAY, GuptaA, KumarV; Isoeff iciency: Measuring the scalability of parallel algorithms and architectures; IEEE Parallel & Distributed Technology, 1993, 1(2) : 12-2l3．
13申 俊、郑纬民；异构并行工作站机群系统的性能评价指标；计算机研 究与发展：第35卷第3期1998年3月。
14：喻占武、胡瑞敏、庞丽萍、李德仁：一种支持分布式进程迁移的动态 负载平衡征募算法的研究；小型微型计算机系统，第20卷，第5期：1999年 5月。
15: Lambda, Amnon Barak, Oren La' adan; he MOSIX Multicomputer Operating System for High Performance Cluster Computing:
16：陆桑璐、谢立；基于簇结构的负载平衡模型簇平衡：计算机研究与发 展：第35卷第9期；1998年9月。
17: Sun Microsystems; RFC 1057 RPC: Remote Procedure Call Protocol specification: Version 2; Jun-01-1988．
18: An Agreement between the Internet Society and Sun Microsystems, Inc. in the Matter of ONC RPC and XDR Protocols.
19：邱毓兰、刘智满、王卓立、彭德纯：Linux操作系统环境中的进程迁 移算法研究：武汉大学学报(自然科学版)：第45卷，第3期：1999年6月。
20: Amnon Barak, Oren La' adan, Amnon Shiloh; Scalable Cluster Computing with MOSIX for LINUX; http://www.mosix.cs.huji.ac.il;.
21：傅强、郑纬民；LBS一基于PVM的动态任务负载平衡系统；小型微型计 算机系统；第19卷，第10期；1998年10月：6-11。
22: Zhou. S. ; An Experimental Assessment of Resource Queue Lengths as Load Indices , Proceedings of the 1987 winter USENIX Conference ; Washington.D.C, Jan. 1987, PP. 73-82．
23: Andrew S. Tanenbaum; Distributed Operating System; Prentice-Hall International, Inc 1996 . 12．
24: Aspnes、 Y.Azar, A. Fiat, A.Plotkin and 0． Waarts; On-Line Machine Scheduling with Applications to Load,Balancing and virtual Circuit Routing; [C]In Proceedings of the ACM Symposium on Theory of Computing (STOC), May 1993．
25: B. Awerbuch, Y. Azar, S. Plotkin and O. Waarts; Competitive Touting
of Virtual Circuits with Unknown Duration; [C]In ACM-SIAM Symposium on Discrete Algorithms(SODA), January 1994
2：周一萍、郑守淇、白英彩；基于PVM的并行Lisp机制与实现；小型微 型计算机系统，第19卷，第8期：1998年8月。
3: Amnon Barak, Orcn la'adan; The MOSIX Multicomputer Operation System for High Performance Cluster Computing: http://www.cs.huji. ac.il/mosix.
4: Chen-chingLiu, Juhwan Jung, Gerald T.Heydt, Vi jay Vital and Arun G. Phadke; The Strategic Power Infrastructure Defense(SPID) System: [J]IEEE Control Systems Magazine: August 2000, 20(4) :40-52．
5: Juan Carlos Gomez, M. M. Morcos; Effect of Distribution System Protection on Voltage Sags; [J]IEEE Power Engineering Review, May 2000, 20(5) :66-68．
6: S. Jovanovic, F.Millicevic. Triangular Distribution Load Flow. [J]IEEE Power Engineering Review, May 2000, 20(5) :60-62．
7：赵谷泉；变电所综合自动化的结构模式及通信总线选择：[J]电力自动 化设备，1999，19(5) ：55-56。
8：王学超，王玉生：分布式配电网调度与管理自动化系统的设计开发：[J] 电力系统自动化。1999，23(7) ：40-42。
9：《地区电网调度自动化设计技术规程》(DL5002-91) 。
10：《电网调度自动化系统实用化要求》(能源调自[1988] 30号)。
11: Mukesh singhal, Niranjan G-shivaratri; Advanced Concepts in Operating System, Distributed, Database, And Multiprocessor Operating Systems, McGraw-Hid, INC; 1994．
12: GramaAY, GuptaA, KumarV; Isoeff iciency: Measuring the scalability of parallel algorithms and architectures; IEEE Parallel & Distributed Technology, 1993, 1(2) : 12-2l3．
13申 俊、郑纬民；异构并行工作站机群系统的性能评价指标；计算机研 究与发展：第35卷第3期1998年3月。
14：喻占武、胡瑞敏、庞丽萍、李德仁：一种支持分布式进程迁移的动态 负载平衡征募算法的研究；小型微型计算机系统，第20卷，第5期：1999年 5月。
15: Lambda, Amnon Barak, Oren La' adan; he MOSIX Multicomputer Operating System for High Performance Cluster Computing:
16：陆桑璐、谢立；基于簇结构的负载平衡模型簇平衡：计算机研究与发 展：第35卷第9期；1998年9月。
17: Sun Microsystems; RFC 1057 RPC: Remote Procedure Call Protocol specification: Version 2; Jun-01-1988．
18: An Agreement between the Internet Society and Sun Microsystems, Inc. in the Matter of ONC RPC and XDR Protocols.
19：邱毓兰、刘智满、王卓立、彭德纯：Linux操作系统环境中的进程迁 移算法研究：武汉大学学报(自然科学版)：第45卷，第3期：1999年6月。
20: Amnon Barak, Oren La' adan, Amnon Shiloh; Scalable Cluster Computing with MOSIX for LINUX; http://www.mosix.cs.huji.ac.il;.
21：傅强、郑纬民；LBS一基于PVM的动态任务负载平衡系统；小型微型计 算机系统；第19卷，第10期；1998年10月：6-11。
22: Zhou. S. ; An Experimental Assessment of Resource Queue Lengths as Load Indices , Proceedings of the 1987 winter USENIX Conference ; Washington.D.C, Jan. 1987, PP. 73-82．
23: Andrew S. Tanenbaum; Distributed Operating System; Prentice-Hall International, Inc 1996 . 12．
24: Aspnes、 Y.Azar, A. Fiat, A.Plotkin and 0． Waarts; On-Line Machine Scheduling with Applications to Load,Balancing and virtual Circuit Routing; [C]In Proceedings of the ACM Symposium on Theory of Computing (STOC), May 1993．
25: B. Awerbuch, Y. Azar, S. Plotkin and O. Waarts; Competitive Touting
of Virtual Circuits with Unknown Duration; [C]In ACM-SIAM Symposium on Discrete Algorithms(SODA), January 1994