多Agent地方财政风险预算编制系统设计、开发及关键技术研究
|
摘要
防范和化解财政风险，消除经济隐患，是各级政府和有关部门极为关注的问题。由于财政风险具有隐蔽性很强的显著特征，往往都是风险累积（或多种风险累加）到超出财政承受能力，无力进行有效化解时，才会显现为财政风险，进而演化为财政危机。因此，建立财政风险的事先分析和监控机制是预防财政风险的关键。以政府财政预算为基础，编制财政风险预算是一种管理财政风险的财政计划，也是防范财政风险的一种行之有效的方法。
本文以重庆市科技攻关重点项目“南岸区人民政府电子政务应用示范（7220-A-1）”子项目“财政风险分析与控制系统”为选题背景，认真研究财政风险的特征，结合南岸区政府财政实际情况，提出对引起地方财政风险的政府负债，按性质进行分类；对每一类负债设置一个Agent进行分析和量化；设置财政风险Agent和其他辅助Agent，组成多Agent系统；利用多个Agent的通讯与协作，编制地方财政风险预算并计算评估财政风险状况的一系列指标。经过充分细致的调研工作，本文完成了南岸区政府财政风险项目透明化，并通过数学分析和推导，建立评估财政风险状况的指标体系。在此基础上，提出系统的总体目标、体系结构及财政风险预算编制过程，重点研究系统的核心“财政风险预算编制子系统的设计与实现”，构建系统的多Agent分布式计算体系结构。
本文以C/S和B/S相结合的三层体系结构设计了财政预算编制子系统，它是系统的一个重要组成部分。同时，详细说明其实现。
本文提出的Agent结构及多Agent协作机制是研究工作理论与实践原型创造的初浅探索。
All levels of governments and related departments pay much attention on keeping away and eliminating financial risks. Owing to its remarkable characteristic of concealment, financial risks usually emerge when debt risks are accumulated beyond the sustaining capacity of finance and can't be eliminated by any methods. And then it may result in financial crisis. Therefore, establishing the system of beforehand analysis as well as the mechanism of monitoring is the key of preventing financial risks. Based on financial budget, making financial risks budget is a financial plan of managing financial risks. Moreover, it is available to avoid financial risks.
This paper is on the basis of the research of the demonstrative item "financial risks analysis and control system" of the Chonqqing Science and Technology tackle key item "Nanan district government electronic demonstrative application of government affair(7220-A-1)". Based on well understanding and comprehending the characteristics of financial risks and meticulous considering the actual conditions of Nanan district government, a method that constructs a Multiagent system to edit local financial risk budget and calculate a series of indexes of evaluating financial risks situations is presented in this paper. Categorizing ingredients that give rise to local financial risk, scheming agents in accordance with these ingredients, and scheming cooperative agent and some auxiliary agents to constitute the system are also proposed. Through plentiful and painstaking investigation, the transparentizing of financial risks items of Nanan district government, as well as the analysis and mathematical developing of these indexes which evaluate financial risks situations have finished. Whereafter, the paper present the system's total target, construction and the process of making financial risks budget, and then lay stress on the design of the agent-based distributing computation structure and actualization of the system's key part, that is, financial risks budget edit subsystem.
The financial budget subsystem, as one of the important parts of the system, has designed and actualized by three layers that combine C/S and B/S.
The agent structure and Multiagent cooperative mechanism in this paper are shallow explore of research on theory and practice of prototype creation.
本文以重庆市科技攻关重点项目“南岸区人民政府电子政务应用示范（7220-A-1）”子项目“财政风险分析与控制系统”为选题背景，认真研究财政风险的特征，结合南岸区政府财政实际情况，提出对引起地方财政风险的政府负债，按性质进行分类；对每一类负债设置一个Agent进行分析和量化；设置财政风险Agent和其他辅助Agent，组成多Agent系统；利用多个Agent的通讯与协作，编制地方财政风险预算并计算评估财政风险状况的一系列指标。经过充分细致的调研工作，本文完成了南岸区政府财政风险项目透明化，并通过数学分析和推导，建立评估财政风险状况的指标体系。在此基础上，提出系统的总体目标、体系结构及财政风险预算编制过程，重点研究系统的核心“财政风险预算编制子系统的设计与实现”，构建系统的多Agent分布式计算体系结构。
本文以C/S和B/S相结合的三层体系结构设计了财政预算编制子系统，它是系统的一个重要组成部分。同时，详细说明其实现。
本文提出的Agent结构及多Agent协作机制是研究工作理论与实践原型创造的初浅探索。
All levels of governments and related departments pay much attention on keeping away and eliminating financial risks. Owing to its remarkable characteristic of concealment, financial risks usually emerge when debt risks are accumulated beyond the sustaining capacity of finance and can't be eliminated by any methods. And then it may result in financial crisis. Therefore, establishing the system of beforehand analysis as well as the mechanism of monitoring is the key of preventing financial risks. Based on financial budget, making financial risks budget is a financial plan of managing financial risks. Moreover, it is available to avoid financial risks.
This paper is on the basis of the research of the demonstrative item "financial risks analysis and control system" of the Chonqqing Science and Technology tackle key item "Nanan district government electronic demonstrative application of government affair(7220-A-1)". Based on well understanding and comprehending the characteristics of financial risks and meticulous considering the actual conditions of Nanan district government, a method that constructs a Multiagent system to edit local financial risk budget and calculate a series of indexes of evaluating financial risks situations is presented in this paper. Categorizing ingredients that give rise to local financial risk, scheming agents in accordance with these ingredients, and scheming cooperative agent and some auxiliary agents to constitute the system are also proposed. Through plentiful and painstaking investigation, the transparentizing of financial risks items of Nanan district government, as well as the analysis and mathematical developing of these indexes which evaluate financial risks situations have finished. Whereafter, the paper present the system's total target, construction and the process of making financial risks budget, and then lay stress on the design of the agent-based distributing computation structure and actualization of the system's key part, that is, financial risks budget edit subsystem.
The financial budget subsystem, as one of the important parts of the system, has designed and actualized by three layers that combine C/S and B/S.
The agent structure and Multiagent cooperative mechanism in this paper are shallow explore of research on theory and practice of prototype creation.
引文
[1] 汪柏坤: 县级财政风险不容忽视. 湖北经济报, 2001-03-27(004).
[2] 卢福禄: 关注地方财政风险[J]. 了望新闻周刊, 2000, 11(49):12-15.
[3] 张焕坤, 黄振清: 财政风险——积聚与对策.
http://www.e521.com/magazine/hbck/200103/0419104703.htm
[4] 傅志华, 黄运: 防范地方财政风险的国际经验与借鉴.
http://www.3idea.net/jryw/jryw02041817.htm
[5] 李俊生: 编制财政风险预算,防范财政风险[N]. 中国财经报, 2000-07-11(002).
[6] 王怀民: 基于Agent的分布计算环境[J]. 计算机学报, 1996, 10(3):17-22.
[7] N. V. Findler: Multiagent coordination and cooperation in a distributed dynamic environment with limited resource [J]. Artificial Intelligence in Engineering, 1995, 9:229-238
[8] Y.Shoham.AGENTO: A Simple Agent Language and its Interpreter. In: AAA1-91,1991.
[9] R.Goodwin: Formalizing Properties of Agents. Technical Report CMU-CS-93-159, School of Computer Science, Carmesie Mellon university, 1993.
[10] Demaze au: Coordination Patterns in Multi-Agent World Application to Computer vision and Robotics. IEEE Colloquium on Intellience Agents. Savoy place, 1991.
[11] J.Bats. A. B. Loyall, W. S. Reilly: An Architecture for action. Emotion and Social Behaviour. Technical Report CMU-CS-92-144, School of Computer Science, Carnegie Mellon university, 1992.
[12] M. H. Coen. SodaBot: A Software Agent Environment and Construction System. Technical Report 1493. MIT, Cambrigdge, 1994.
[13] D. C Smith etc. KidSim: Programming Agents without a Programming Language. Communication of the ACM, 1994, 37.
[14] 董红斌: 多Agent技术研究[J]. 计算机学报, 1999, 10(2): 29-32.
[15] 刘金琨, 尔联洁: 多智能体技术应用综述, 2001, 16(2).
[16] 赵晓: 中国财政风险及其纾缓.
http://www.unirule.org.cn/symposium/c200.htm
[17] 段清和: 防范和化解地方财政风险的对策探讨. 湖北经济报,2000-12-22(A03).
[18] 申嫦娥: 地方财政风险与财政制度改革, 财会研究, 2000(11).
[19] Katia P Sycara: Multiagent System. AI Magazine, 1998.
[20] 胡代平: 多Agent宏观经济决策支持系统的开发[J]. 中国矿业大学学报, 2001, 30(1):17-22.
[21] Ibrahim F Imam, Yves Kodratoff: Intelligent Adaptive Agents. AI Magazine, 1997.
[22] 史殿习, 王怀民, 邹鹏等: 一种基于Agent的分布式集成框架的设计与实现[J]. 计算机科学, 1998, 25(4): 42-45.
[23] Jennings N R, Sycara K P, Wooldriage M J: Agent technology: Foundations, applications and market [M]. Heidelberg: Springer-Verlag, 1997.
[24] Wittig T. ARCHON: An architecture for multi-agent systems [M]. Chichester England: Ellis Horwood, 1992.
[25] NilsJ. Nilsson: Artifical Intelligence: A new Synthesis. 北京机械工业出版社, 2000.
[26] 周笑波, 谢立, 周晓方: 一个基于多服务员系统的Internet市场服务管理模型[J]. 计算机学报, 1999, 22(4): 424-430.
[27] 张东摩, 李红兵: 人工智能研究动态与发展趋势[J]. 计算机科学, 1998, 25(2): 5-8.
[28] 朱山风, 陈莘萌等: 一种基于三层 C/S结构的MIS设计, 武汉大学学报, 1999.2: 53-56.
[29] 黄立, 杨林林, 王海: 三层结构技术及其在 Internet上的应用. 中国计算机报, 1998, 9: 80-81.
[30] 琚春华, 王光明: 基于C/S环境的商业决策支持系统研究与设计. 控制与决策, 2000, 1: 38-42.
[31] 姚国祥, 毛涛, 沈镇林: 基于 WWW的 Browser/ Server结构的信息系统研究与设计. 计算机科学, 1998, 6: 75-77.
[32] Sarit Kraus: Negotiation and cooperation in Multiagent environments [J]. Artificial Intelligence, 1997, 94: 79-97.
[33] Randall Davis, Reid G. Smith: Negotiation as a Mataphor for Distributed Problem Solving [J]. Artificial Intelligence, 1998, 20: 63-109.
[34] V R Lesser: Arestrospective of FA/C problem aolving [J]. IEEE Trans on SMC, 1991, 21(6): 81-96.
[35] 罗诩, 石纯一: Agent协作求解中形成联盟的行为策略[J]. 计算机学报, 1997, 20(11): 961-965.
[36] Jennings N R: Controlling cooperative problem solving industrial Multiagent systems using joint intention[J]. Artificial Intelligence, 1995, 75(2): 195-240.
[37] N.V.Findler, U.K.Sengupta: Multiagent collaboration in time-constrained domains[J]. Artificial Intelligence in Engineering, 1994, 9: 39-52.
[38] N.V.Findler: Multiagent coordination and cooperation in a distributed dynamic environment with limited resource [J]. Artificial Intelligence in Engineering, 1995, 9: 229-238.
[39] X. Miao: A Normative-Descriptive approach to hierarchical team resource allocation [J]. IEEE Trans on SMC, 1992, 22(3): 482-496.
[40] Wen-Ran Zhang. A Cognitive-Map Based Approach to the Coordination of Distributed Cooperative Agents [J]. IEEE Trans on SMC, 1992, 22(1): 103-113.
[2] 卢福禄: 关注地方财政风险[J]. 了望新闻周刊, 2000, 11(49):12-15.
[3] 张焕坤, 黄振清: 财政风险——积聚与对策.
http://www.e521.com/magazine/hbck/200103/0419104703.htm
[4] 傅志华, 黄运: 防范地方财政风险的国际经验与借鉴.
http://www.3idea.net/jryw/jryw02041817.htm
[5] 李俊生: 编制财政风险预算,防范财政风险[N]. 中国财经报, 2000-07-11(002).
[6] 王怀民: 基于Agent的分布计算环境[J]. 计算机学报, 1996, 10(3):17-22.
[7] N. V. Findler: Multiagent coordination and cooperation in a distributed dynamic environment with limited resource [J]. Artificial Intelligence in Engineering, 1995, 9:229-238
[8] Y.Shoham.AGENTO: A Simple Agent Language and its Interpreter. In: AAA1-91,1991.
[9] R.Goodwin: Formalizing Properties of Agents. Technical Report CMU-CS-93-159, School of Computer Science, Carmesie Mellon university, 1993.
[10] Demaze au: Coordination Patterns in Multi-Agent World Application to Computer vision and Robotics. IEEE Colloquium on Intellience Agents. Savoy place, 1991.
[11] J.Bats. A. B. Loyall, W. S. Reilly: An Architecture for action. Emotion and Social Behaviour. Technical Report CMU-CS-92-144, School of Computer Science, Carnegie Mellon university, 1992.
[12] M. H. Coen. SodaBot: A Software Agent Environment and Construction System. Technical Report 1493. MIT, Cambrigdge, 1994.
[13] D. C Smith etc. KidSim: Programming Agents without a Programming Language. Communication of the ACM, 1994, 37.
[14] 董红斌: 多Agent技术研究[J]. 计算机学报, 1999, 10(2): 29-32.
[15] 刘金琨, 尔联洁: 多智能体技术应用综述, 2001, 16(2).
[16] 赵晓: 中国财政风险及其纾缓.
http://www.unirule.org.cn/symposium/c200.htm
[17] 段清和: 防范和化解地方财政风险的对策探讨. 湖北经济报,2000-12-22(A03).
[18] 申嫦娥: 地方财政风险与财政制度改革, 财会研究, 2000(11).
[19] Katia P Sycara: Multiagent System. AI Magazine, 1998.
[20] 胡代平: 多Agent宏观经济决策支持系统的开发[J]. 中国矿业大学学报, 2001, 30(1):17-22.
[21] Ibrahim F Imam, Yves Kodratoff: Intelligent Adaptive Agents. AI Magazine, 1997.
[22] 史殿习, 王怀民, 邹鹏等: 一种基于Agent的分布式集成框架的设计与实现[J]. 计算机科学, 1998, 25(4): 42-45.
[23] Jennings N R, Sycara K P, Wooldriage M J: Agent technology: Foundations, applications and market [M]. Heidelberg: Springer-Verlag, 1997.
[24] Wittig T. ARCHON: An architecture for multi-agent systems [M]. Chichester England: Ellis Horwood, 1992.
[25] NilsJ. Nilsson: Artifical Intelligence: A new Synthesis. 北京机械工业出版社, 2000.
[26] 周笑波, 谢立, 周晓方: 一个基于多服务员系统的Internet市场服务管理模型[J]. 计算机学报, 1999, 22(4): 424-430.
[27] 张东摩, 李红兵: 人工智能研究动态与发展趋势[J]. 计算机科学, 1998, 25(2): 5-8.
[28] 朱山风, 陈莘萌等: 一种基于三层 C/S结构的MIS设计, 武汉大学学报, 1999.2: 53-56.
[29] 黄立, 杨林林, 王海: 三层结构技术及其在 Internet上的应用. 中国计算机报, 1998, 9: 80-81.
[30] 琚春华, 王光明: 基于C/S环境的商业决策支持系统研究与设计. 控制与决策, 2000, 1: 38-42.
[31] 姚国祥, 毛涛, 沈镇林: 基于 WWW的 Browser/ Server结构的信息系统研究与设计. 计算机科学, 1998, 6: 75-77.
[32] Sarit Kraus: Negotiation and cooperation in Multiagent environments [J]. Artificial Intelligence, 1997, 94: 79-97.
[33] Randall Davis, Reid G. Smith: Negotiation as a Mataphor for Distributed Problem Solving [J]. Artificial Intelligence, 1998, 20: 63-109.
[34] V R Lesser: Arestrospective of FA/C problem aolving [J]. IEEE Trans on SMC, 1991, 21(6): 81-96.
[35] 罗诩, 石纯一: Agent协作求解中形成联盟的行为策略[J]. 计算机学报, 1997, 20(11): 961-965.
[36] Jennings N R: Controlling cooperative problem solving industrial Multiagent systems using joint intention[J]. Artificial Intelligence, 1995, 75(2): 195-240.
[37] N.V.Findler, U.K.Sengupta: Multiagent collaboration in time-constrained domains[J]. Artificial Intelligence in Engineering, 1994, 9: 39-52.
[38] N.V.Findler: Multiagent coordination and cooperation in a distributed dynamic environment with limited resource [J]. Artificial Intelligence in Engineering, 1995, 9: 229-238.
[39] X. Miao: A Normative-Descriptive approach to hierarchical team resource allocation [J]. IEEE Trans on SMC, 1992, 22(3): 482-496.
[40] Wen-Ran Zhang. A Cognitive-Map Based Approach to the Coordination of Distributed Cooperative Agents [J]. IEEE Trans on SMC, 1992, 22(1): 103-113.