基于模糊神经网络的房产软件项目成本估算研究
|
摘要
随着经济的发展和城市化进程的加快,房产管理业务量迅速增加。近年来国家先后颁布了一系列房地产管理法案,传统的房产管理方式已满足不了社会各界,包括用户个人、企业、政府相关部门等对房产的管理与服务不断增长的需求。现状迫切要求房产管理部门进行行业的信息化建设,特别是信息系统的建设来提升管理水平。为了更好的管理和服务,房产管理信息系统的研究和开发迫在眉睫。
软件项目成本估算是项目成本管理的一个重要环节,它作为成本管理的起点,是成本控制等其他成本管理过程的依据和基础。目前,成本估算的方法有很多,但是还没有哪种方法能适应于所有的软件项目。常用的方法大多偏重于主观经验,但是模糊神经网络技术集合了模糊理论和神经网络的各自优点,使得估算结果相对准确。本文以探索的角度,结合实际需要,构建了基于模糊神经网络构建的软件成本估算模型,并将其用于房地产信息管理估算系统。
本文首先说明了对软件项目管理及软件成本估算的研究背景、研究目的和意义,并对国内外研究现状作了总结归纳,最后对文中涉及到模糊理论、神经网络及房产信息管理方面的相关基本理论进行了详细地阐述。本文构建了神经网络模型和模糊神经网络模型两种软件成本估算模型,以成熟的COCOMOⅡ模型为依据加以比较,得出模糊神经网络在软件成本估算方面优于神经网络模型,并将构建好的估算模型用于房产综合信息管理系统。本文验证了用模糊神经网络方法构建的软件成本估算模型综合了多方优势,能够快速有效的用于房产信息管理系统,为今后类似项目的开发维护提供了依据。
With the accelerated economic development and urbanization,the volume of business of property management is increasing rapidly.The state has promulgated a series of real estate management act in recent years,but traditional property management approach has failed to meet the growing demand of the community in management and service,such as individuals,businesses and government departments.The situation urgently requires the property management department for industry information construction to improve the management level.In order to better management and services,the research and development of real estate management information system is imminent.
The cost estimation of software project is an important part of the project cost management as a starting point,and it is the basis and foundation for cost control and other cost management processes.Currently,there are many cost estimation methods,but no single one can be adapted to all software projects.Common methods focus on subjective experience, but fuzzy Neural Network Technology makes relatively accurate estimates because it collects the respective advantages of the fuzzy theory and neural networks. Using exploration point of view and combined with the actual needs,it builds a software cost estimation model based on fuzzy neural network built,and be used to real estate information management estimating system.
This paper first describes the background,purpose,significance of software project management and cost estimation, and makes a summarized to research status,finally,it relates to the basic theory involved in the fuzzy theory,neural network and real estate information management described in detail.It constructs a model of neural networks and fuzzy neural network model comparing with mature COCOMOⅡmodel,and concludes that fuzzy neural network in software cost estimation is superior to neural network model,meanwhile,it builds a good estimation model.It shows that the software cost estimation model constructed by fuzzy neural networks combines multiple advantages.It can be used for real estate information management system quickly and efficiently,and provides a basis for the development and maintenance of similar projects in the future.
软件项目成本估算是项目成本管理的一个重要环节,它作为成本管理的起点,是成本控制等其他成本管理过程的依据和基础。目前,成本估算的方法有很多,但是还没有哪种方法能适应于所有的软件项目。常用的方法大多偏重于主观经验,但是模糊神经网络技术集合了模糊理论和神经网络的各自优点,使得估算结果相对准确。本文以探索的角度,结合实际需要,构建了基于模糊神经网络构建的软件成本估算模型,并将其用于房地产信息管理估算系统。
本文首先说明了对软件项目管理及软件成本估算的研究背景、研究目的和意义,并对国内外研究现状作了总结归纳,最后对文中涉及到模糊理论、神经网络及房产信息管理方面的相关基本理论进行了详细地阐述。本文构建了神经网络模型和模糊神经网络模型两种软件成本估算模型,以成熟的COCOMOⅡ模型为依据加以比较,得出模糊神经网络在软件成本估算方面优于神经网络模型,并将构建好的估算模型用于房产综合信息管理系统。本文验证了用模糊神经网络方法构建的软件成本估算模型综合了多方优势,能够快速有效的用于房产信息管理系统,为今后类似项目的开发维护提供了依据。
With the accelerated economic development and urbanization,the volume of business of property management is increasing rapidly.The state has promulgated a series of real estate management act in recent years,but traditional property management approach has failed to meet the growing demand of the community in management and service,such as individuals,businesses and government departments.The situation urgently requires the property management department for industry information construction to improve the management level.In order to better management and services,the research and development of real estate management information system is imminent.
The cost estimation of software project is an important part of the project cost management as a starting point,and it is the basis and foundation for cost control and other cost management processes.Currently,there are many cost estimation methods,but no single one can be adapted to all software projects.Common methods focus on subjective experience, but fuzzy Neural Network Technology makes relatively accurate estimates because it collects the respective advantages of the fuzzy theory and neural networks. Using exploration point of view and combined with the actual needs,it builds a software cost estimation model based on fuzzy neural network built,and be used to real estate information management estimating system.
This paper first describes the background,purpose,significance of software project management and cost estimation, and makes a summarized to research status,finally,it relates to the basic theory involved in the fuzzy theory,neural network and real estate information management described in detail.It constructs a model of neural networks and fuzzy neural network model comparing with mature COCOMOⅡmodel,and concludes that fuzzy neural network in software cost estimation is superior to neural network model,meanwhile,it builds a good estimation model.It shows that the software cost estimation model constructed by fuzzy neural networks combines multiple advantages.It can be used for real estate information management system quickly and efficiently,and provides a basis for the development and maintenance of similar projects in the future.
引文
[1]林师建.项目成本管理[M].北京：对外经济贸易大学出版社,2007.
[2]薛四新,郭军.软件项目管理[M].北京：机械工业出版社,2006.
[3]戚安邦.项目成本管理[M].天津：南开大学出版社,2006.
[4]Kirsch, L. J. Software Project Management:An Integrated Perspective for an Emerging Paradigm[M].Projceting the Future from the Past,2000:285-304.
[5]Bob Hughes.软件项目管理[M].北京：机械工业出版社,2010,9.
[6]帕维兹F.拉德.Project Estimating and Cost Management[M]北京：机械工业出版社,2005,1.
[7]Nicholas, J. Project Management for Business and Technology.Principles and Practice[M]. Prentice Hall:Upper Saddle River, NJ,2001.
[8]何晓阳.基于模型的软件成本估计方法[J].计算机研究与发展,2006,43(5)：777-783.
[9]康一梅.软件项目管理[M].北京：清华大学出版社,2010,4.
[10]Kathy Schwalbe.IT项目管理[M].北京：机械工业出版社,2004,11.
[11]Nasser Tadayon, Neural Network Approach for Software Cost Estimation. Proceedings of the IEEE International Conference on Information Technology:Coding and Computing, April 2005,2:815-818.
[12]Kathy Schwalbe. IT项目管理[M].北京：机械工业出版社,2008,12.
[13]McLaurin. The Comprehensive Impact of Offshore IT Software and Serbices Outsourcing on the U.S. Economy and the IT Industry[M]. Design Atuomation,2005.
[14]Mark A. Fuller. IT项目管理[M].北京：人民邮电出版社,2009,6.
[15]Blanchard, K., Fowler, S., and Hawkins, L. Self Leadership and the One Minute Magager:Increasing Effectiveness Through Situationsl Self Leadership[M]. New York:William Morrow,2005.
[16]周杰,杜磊.COMOMO软件项目管理中的成本估算方法[J].计算机应用研究,2000(11)：56-58.
[17]鲁光男.浅析软件成本估算的三种方法[J].长春大学学报,2006,18(2)：69-71.
[18]陈艳.ERP软件成本估算研究[D].硕士论文,武汉理工大学,2008.
[19]宋敬廷.软件成本估算模型的现状与展望[J].水利水电科学研究院,1994,2：14-19.
[20]王宏宇,党齐民.面向对象软件项目估算方法的研究[J].计算机与数字工 程,2005,07(33)：35-38.
[21]胡樱.软件成本估算[J].舰船电子工程,2005,25(6)：14-18.
[22]Briand LC, Wieczorek I. Resource Estimation in Software Engineering. In:Marcinak JJ, ed. Encyclopedia of Software Engineering. New York:John Wiley& Sons,2002,1160-1196.
[23]Capers Jones, Estimating Software Costs:Bringing Realism to Estimating. McGraw-Hill Osborne Media,2007.25-28.
[24]刘泽星,李建华.软件规模评估方法的研究与应用计算机工程[J].2002,12(28)：113-115.
[25]Saratchandran P. Performance evaluation of a sequential minimal radial basis function neural networks learning algorithm[J]. Neural Networks,1998,9(2):309-313.
[26]何晓阳.基于模型的软件成本估计方法[J].计算机研究与发展,2006,43(5)：777-783.
[27]方海光.基于范例库推理的软件成本估算模型研究[J].计算机工程,2006,32(19)：191-192.
[28]刘杰生.软件成本估算方法与技术现状探讨[J].舰船电子工程,2006,4：13-18.
[29]宋敬廷.软件成本估算模型的现状与展望[J].水利水电科学研究院,1994,2：14-19.
[30]陈冰.模糊神经网络研究[D].硕士论文,广东工业大学,2005.
[31]施彦,韩力群,廉小亲.神经网络设计与实例分析[M].北京：北京邮电大学出版社,2009,12.
[32]Wu S Q, Gao Y. A Fast Approach for Automatic Generation of Fuzzy Rules by Generalized Dynamic Fuzzy Neural Networks[J]. Fuzzy Systems,2001,9:579-592.
[33]Lee CH, Teng. Indentif ication and Control of Dynamic Systems Using Recurrent Fuzzy Neural Networks[J]. Fuzzy Systems,2000.8:349-366.
[34]钟珞,饶文碧,邹承明.人工神经网络及其融合应用技术[M].北京：科学出版社,2007,1.
[35]张吉礼.模糊-神经网络控制原理与工程应用[M].哈尔滨：哈尔滨工业大学出版社,2004,6.
[36]张乃饶,阎平凡.神经网络与模糊控制[M].北京：清华大学出版社,1998,10.
[37]H. Schwenk. Boosting Neural Networks[J]. Neural Computation, 2000,12 (8):1869-1887.
[38]Dorigo. M,施蒂茨勒.人工神经网络及其融合应用技术[M].北京：科学出版社,2007,1.
[39]周春光.计算智能[M].长春：吉林大学出版社,2009,1.
[40]史峰,王小川,郁磊Mat lab神经网络30个案例分析[M].北京：北京航天航空大学出版 社,2010,4.
[41]Y. Bengio, Low C B. Real-Time Implementation and Control of a Class of MIMO Nonlinear Systems[J]. Microprocessors and Microsystems,2002,26:457-461.
[42]Golnaraghi M F. Sensitivity Analysis of the Frequency Response of a Piecewiselinear System In a Frequency Island[J]. Journal of Vibration and Control,2004,10(2):187-190.
[43]Spevacek Christopher. Using A CMAC Neural Network in Noisy Environments[J]. Intenlligent Engineering Systems Through Artificial Neural Networks,2003,13:21-23.
[44]M F Abbood, D A Linkens, etc. A blackboard software architecture for intelligent control systems[J]. Kybernetes:The International Journal of Systems & Cybernetics,2001,29(8):1013-1015.
[45]李晓忠,汪培庄,罗承忠.模糊神经网络[M].贵阳：贵州科技出版社,1994.
[46]J.R.Quinlan, Bagging. In Proceedings of The Thirteenth National Conference on Atrificial Intelligence[J]. MIT Press,1996:725-730.
[47]Sundararajan N. Analysis of Minimal Radial Basis Function Neural Networks Algorithm for Realtime Identification of Nonlinear Dynamic Systems[J]. Control Theory,2000,147 (4):467-484.
[48]Vuolevi J,Rankonen T. Measurement Technique for Characterizing Memory Effects[J].On MTT-S,2001(49):1383-1388.
[49]Bhatt T M, McCain D. Matlab as A Development Environment for FPGA Design[M]. Design Automation, California,2005:607-609.
[50]刘莉霞,贺贵明.基于BP神经网络的模糊综合评判模型应用[J].计算机应用,2001,21(8)：61-67.
[51]刘增量,刘有才.模糊逻辑与神经网络——理论研究与探索[M].北京：北京航空航天大学出版社,1996.
[52]Magee DP. Matlab Extensions for The Development [M]. Design Atuomation, California, 2005:607-610.
[53]E. M. Cochrane, J. E. Beasley. The CO-adaptive Neural Network Approach to The Euc-lidean Traveling SalesmanProblem[J].NerualNetworks,2003,4(16):33-38.
[54]唐颖.软件项目成本估算研究[D].硕士学位论文.成都：电子科技大学,2006.
[55]杨泽中.房产测绘管理系统的设计与实现[J].黑河科技,2001,(4)：36-37.
[56]方志祥,黄全义.罗年学.房产测绘管理系统软件开发中的若干问题[J].测绘信息与工 程,2002,27(3)：26-2了.
[57]张桥平,张友静,许捍卫.房产分层分户图管理系统开发模式研究[J].测绘工程,2001,10(3)：38-40.
[58]徐爱俊,黄全义.房产测绘管理系统的设计与实现[J].城市勘测,2001,(4)：4-7.
[59]徐爱俊,放陆明.基于GIS的房产信息管理与办公自动化系统的设计与实现[J].矿山测量,2003,(1)：27-30.
[60]何红.房地产开发经营与管理[M].北京：化学工业出版史.2008,1.
[61]蒋海琴,蒋文明,闾国年.房产管理信息系统[M].北京：科学出版社,2007,4.
[62]姚星明.房地产项目管理[M].北京：化学工业出版社,2009,11.
[2]薛四新,郭军.软件项目管理[M].北京：机械工业出版社,2006.
[3]戚安邦.项目成本管理[M].天津：南开大学出版社,2006.
[4]Kirsch, L. J. Software Project Management:An Integrated Perspective for an Emerging Paradigm[M].Projceting the Future from the Past,2000:285-304.
[5]Bob Hughes.软件项目管理[M].北京：机械工业出版社,2010,9.
[6]帕维兹F.拉德.Project Estimating and Cost Management[M]北京：机械工业出版社,2005,1.
[7]Nicholas, J. Project Management for Business and Technology.Principles and Practice[M]. Prentice Hall:Upper Saddle River, NJ,2001.
[8]何晓阳.基于模型的软件成本估计方法[J].计算机研究与发展,2006,43(5)：777-783.
[9]康一梅.软件项目管理[M].北京：清华大学出版社,2010,4.
[10]Kathy Schwalbe.IT项目管理[M].北京：机械工业出版社,2004,11.
[11]Nasser Tadayon, Neural Network Approach for Software Cost Estimation. Proceedings of the IEEE International Conference on Information Technology:Coding and Computing, April 2005,2:815-818.
[12]Kathy Schwalbe. IT项目管理[M].北京：机械工业出版社,2008,12.
[13]McLaurin. The Comprehensive Impact of Offshore IT Software and Serbices Outsourcing on the U.S. Economy and the IT Industry[M]. Design Atuomation,2005.
[14]Mark A. Fuller. IT项目管理[M].北京：人民邮电出版社,2009,6.
[15]Blanchard, K., Fowler, S., and Hawkins, L. Self Leadership and the One Minute Magager:Increasing Effectiveness Through Situationsl Self Leadership[M]. New York:William Morrow,2005.
[16]周杰,杜磊.COMOMO软件项目管理中的成本估算方法[J].计算机应用研究,2000(11)：56-58.
[17]鲁光男.浅析软件成本估算的三种方法[J].长春大学学报,2006,18(2)：69-71.
[18]陈艳.ERP软件成本估算研究[D].硕士论文,武汉理工大学,2008.
[19]宋敬廷.软件成本估算模型的现状与展望[J].水利水电科学研究院,1994,2：14-19.
[20]王宏宇,党齐民.面向对象软件项目估算方法的研究[J].计算机与数字工 程,2005,07(33)：35-38.
[21]胡樱.软件成本估算[J].舰船电子工程,2005,25(6)：14-18.
[22]Briand LC, Wieczorek I. Resource Estimation in Software Engineering. In:Marcinak JJ, ed. Encyclopedia of Software Engineering. New York:John Wiley& Sons,2002,1160-1196.
[23]Capers Jones, Estimating Software Costs:Bringing Realism to Estimating. McGraw-Hill Osborne Media,2007.25-28.
[24]刘泽星,李建华.软件规模评估方法的研究与应用计算机工程[J].2002,12(28)：113-115.
[25]Saratchandran P. Performance evaluation of a sequential minimal radial basis function neural networks learning algorithm[J]. Neural Networks,1998,9(2):309-313.
[26]何晓阳.基于模型的软件成本估计方法[J].计算机研究与发展,2006,43(5)：777-783.
[27]方海光.基于范例库推理的软件成本估算模型研究[J].计算机工程,2006,32(19)：191-192.
[28]刘杰生.软件成本估算方法与技术现状探讨[J].舰船电子工程,2006,4：13-18.
[29]宋敬廷.软件成本估算模型的现状与展望[J].水利水电科学研究院,1994,2：14-19.
[30]陈冰.模糊神经网络研究[D].硕士论文,广东工业大学,2005.
[31]施彦,韩力群,廉小亲.神经网络设计与实例分析[M].北京：北京邮电大学出版社,2009,12.
[32]Wu S Q, Gao Y. A Fast Approach for Automatic Generation of Fuzzy Rules by Generalized Dynamic Fuzzy Neural Networks[J]. Fuzzy Systems,2001,9:579-592.
[33]Lee CH, Teng. Indentif ication and Control of Dynamic Systems Using Recurrent Fuzzy Neural Networks[J]. Fuzzy Systems,2000.8:349-366.
[34]钟珞,饶文碧,邹承明.人工神经网络及其融合应用技术[M].北京：科学出版社,2007,1.
[35]张吉礼.模糊-神经网络控制原理与工程应用[M].哈尔滨：哈尔滨工业大学出版社,2004,6.
[36]张乃饶,阎平凡.神经网络与模糊控制[M].北京：清华大学出版社,1998,10.
[37]H. Schwenk. Boosting Neural Networks[J]. Neural Computation, 2000,12 (8):1869-1887.
[38]Dorigo. M,施蒂茨勒.人工神经网络及其融合应用技术[M].北京：科学出版社,2007,1.
[39]周春光.计算智能[M].长春：吉林大学出版社,2009,1.
[40]史峰,王小川,郁磊Mat lab神经网络30个案例分析[M].北京：北京航天航空大学出版 社,2010,4.
[41]Y. Bengio, Low C B. Real-Time Implementation and Control of a Class of MIMO Nonlinear Systems[J]. Microprocessors and Microsystems,2002,26:457-461.
[42]Golnaraghi M F. Sensitivity Analysis of the Frequency Response of a Piecewiselinear System In a Frequency Island[J]. Journal of Vibration and Control,2004,10(2):187-190.
[43]Spevacek Christopher. Using A CMAC Neural Network in Noisy Environments[J]. Intenlligent Engineering Systems Through Artificial Neural Networks,2003,13:21-23.
[44]M F Abbood, D A Linkens, etc. A blackboard software architecture for intelligent control systems[J]. Kybernetes:The International Journal of Systems & Cybernetics,2001,29(8):1013-1015.
[45]李晓忠,汪培庄,罗承忠.模糊神经网络[M].贵阳：贵州科技出版社,1994.
[46]J.R.Quinlan, Bagging. In Proceedings of The Thirteenth National Conference on Atrificial Intelligence[J]. MIT Press,1996:725-730.
[47]Sundararajan N. Analysis of Minimal Radial Basis Function Neural Networks Algorithm for Realtime Identification of Nonlinear Dynamic Systems[J]. Control Theory,2000,147 (4):467-484.
[48]Vuolevi J,Rankonen T. Measurement Technique for Characterizing Memory Effects[J].On MTT-S,2001(49):1383-1388.
[49]Bhatt T M, McCain D. Matlab as A Development Environment for FPGA Design[M]. Design Automation, California,2005:607-609.
[50]刘莉霞,贺贵明.基于BP神经网络的模糊综合评判模型应用[J].计算机应用,2001,21(8)：61-67.
[51]刘增量,刘有才.模糊逻辑与神经网络——理论研究与探索[M].北京：北京航空航天大学出版社,1996.
[52]Magee DP. Matlab Extensions for The Development [M]. Design Atuomation, California, 2005:607-610.
[53]E. M. Cochrane, J. E. Beasley. The CO-adaptive Neural Network Approach to The Euc-lidean Traveling SalesmanProblem[J].NerualNetworks,2003,4(16):33-38.
[54]唐颖.软件项目成本估算研究[D].硕士学位论文.成都：电子科技大学,2006.
[55]杨泽中.房产测绘管理系统的设计与实现[J].黑河科技,2001,(4)：36-37.
[56]方志祥,黄全义.罗年学.房产测绘管理系统软件开发中的若干问题[J].测绘信息与工 程,2002,27(3)：26-2了.
[57]张桥平,张友静,许捍卫.房产分层分户图管理系统开发模式研究[J].测绘工程,2001,10(3)：38-40.
[58]徐爱俊,黄全义.房产测绘管理系统的设计与实现[J].城市勘测,2001,(4)：4-7.
[59]徐爱俊,放陆明.基于GIS的房产信息管理与办公自动化系统的设计与实现[J].矿山测量,2003,(1)：27-30.
[60]何红.房地产开发经营与管理[M].北京：化学工业出版史.2008,1.
[61]蒋海琴,蒋文明,闾国年.房产管理信息系统[M].北京：科学出版社,2007,4.
[62]姚星明.房地产项目管理[M].北京：化学工业出版社,2009,11.