Zhang-li YANG,Xiao-rong YOU,Dong XUSchool of Architecture and Materials,Chongqing College of Electronic Engineering,Chongqing 4013,China;
2Chongqing Machinery and Electronics Technician College,Chongqing 400037,China;
3Chongqing Tianhong Mining Co.,Ltd,Chongqing 401521,China
The popularization and generalization of network technology has brought brings a maximum convenience for fast speed processing conjuncture emergence event.Along with the swift increasing of net information,the network has become a great information base composed by heterogeneous data source which is difficult to make the stat.There is a lot of heterogeneous information among conventional database system and file system as well as semi-structure information system based on Web applications,such as HTML and XML and so on.Due to the difference on platform and database technology as well as communication protocol etc,it makes the operating among data source not only be difficult but also complex so as to form various information isolated island.The heterogeneous information includes being different in content and format and quality of data and even some data format not being transformed or information lost after conversion and so on,because of the different data coming from different avenue provided by different user in implementing data share process of city emergency system,specially in the applications of emergency event processing system.For instance,nowadays it has been become a hot topic on how to make the command department of emergency center,such as fire alarm system and electric power system etc,and to share data information on network better and to deal with conjuncture emergency event fleetly.The emergency event processing owns great uncertainty,once it happens that a lot of heterogeneous data must be processed in real time,and therefore it needs higher processing performance to emergency event processing system.Under the above background,the exchange and integration of heterogeneous data have been become more and more importance,currently it has been become a hot topic for researchers.
The essence of information integration is to implement the share of information and resource among networks which include not only the management information net but also the real time control net.It can be essentially divided into the following technologies so as to implement the system integration among control net and information net.
①Dynamic data exchange.Each application program can share the memory to exchange the information and implement dynamic data exchange between real data in control net and database in information net when the control and information net share the same work station or communication computer.
②Interconnection between control and information net is implemented by gateway and router.Net extending is implemented by bridge and repeaters.Web technology has got the practical application in interconnection.
③Long-distance communication technology.It is one of effective methods to implement the network integration when the geography distance is far from control and information net,such as data communication by modem,remote communication based on TCP/IP.
④Data access.Intranet connects into the control net through browser,and it can make information exchange with information net database in dynamic and interactive mode.
⑤Aiming at the software function integration of heterogeneous industrial control system group,the system can make seamless integration by OPC.
Along with farther popularizing of data integration technology,it appears two sorts of new data integration method successively,namely virtual database and data warehouse[1].It does not store any own data in the virtual database,the user’s query is first transformed into one or multi-data resource,after that it carries through synthesis processing response of user query for data resource,and finally the processing result is returned back to user.The data warehouse method means that the data copy from several data resources is stored in the unitary database named as data warehouse.Before it is stored to the data in the data warehouse,it has to make pretreatment,such as making the data filter and the relation connecting or assembling and so on,and make all the data mode accord with data mode of database.Whereas it is different between virtual database and data warehouse method in data integration tool and data analysis hierarchy,and still it has been widely used in various fields.But there are lots of problems needed to be solved.It is difficult to adapt the situation in data happened neatly because of the rule of data conversion and integration being fused in the customizing code.Due to the offered information being always the past information,it is difficult to get the accuracy information in real time.Also it is possible to make new data isolated island because of the data integration only through making various system to form middle database or centralized database.For adapting the new changing operation requirement,it has to pay more cost because of the frangibility in data integration scheme,and therefore it has to solve a series of technical puzzle[2].
As shown in Figure 1,the middleware is located between the application and system software.It is a kind of basic software that is independent of hardware or database manufacturer,and only a connector between client and service side.It belongs to category of reusable software,and it is needed to make secondary development for middle product.By dint of the middleware,it can implement the share among different software technology,specially for some distributed application software.It can manage the computing resource and net communication over the operating system of client server.Its function is that it provides the environment of running and development for application software located its upper-hierarchy,and helps user to develop and integrate the complicated application software in agility and high efficiency,and provides the currency services located among system hardware,operating system and applications.
The middleware can screen the differences among net hardware platform,heterogeneous character of operating system and net protocol.It makes application software to be able running smoothly over different hardware system or operating system,and at the same the application performance can greatly be upgraded in the aspects of load balance,connection management and scheduling,and therefore it can satisfy the requirement of key operation.The general service owns standard program interface and protocol.Aimed at the difference in operating system and hardware platform,the various implementations can be provided according with interface and protocol criterion,and satisfy a great lot of application demand.Generally speaking,it makes the system running on various platforms of hardware and operating system,and supports the distributed computing,standard protocol and interface,provides the transparence application or service interaction based on the platform by cross-net,cross-hardware and cross-operating system.
Figure 1.Middleware in system structure
It must be satisfied to the two basic conditions for going heterogeneous data exchange of non-structured database based on net application.The one is that it must be compatible to variously heterogeneous data format,whether it is structured or semi-structured data.The other is that it is easy to issue and exchange,after data exchanged it can be issued in multi-format expediently.Just the technology of XML can satisfy the above demand,not only become a sort of standard of data exchange among applications,but also be one of representation technology and an important information exchange standard in Internet[3].
For convenience of discussion,now it takes data exchange among Oracle,Sql server etc and access as an example,the operating system of data exchange platform applies Windows 2000 SP4,and the development environment selects ASP.NET and Visual C#,and then the platform structure is shown in Figure 2.
Figure 2.Platform archite cture
1)XML based data model
XML is a sort of semi-structured data model,and it can be used to describe anomalistic data and integrate the data from different data resources.The characteristic is that it can unlimitedly define a set of marker,provide independent of resolution with various features,and also it is extensible and structured,and owns higher efficiency.The data model in XML separates the data from display,and only modifying manner can change the display mode,and it can display in different mode by means of different demand[4].It can make self-describing,generally the file of XML includes a document type statement(DTD/Schema),the mode of representing data is truly independent of application and platform,and the document is viewed as the documental database and data document.To be independent of platform,the document of XML is a plain text and independent of platform and application,and the other system application can directly make data operation in XML file described by file itself.It is a standard cross-platform pattern for data exchange and operation.And it is able to implement mutual data operation in heterogeneous data integration.
2)Database and XML
Any document of XML has itself logic and physical structure.From the viewpoint of physical structure,the document is composed of cell called as entity,and the entity can quote the other entity and make it be included in the document.Logically the document consists of statement,element,note,symbol citation and process instruction.It is needed to own start mark and end mark for all the elements in the document of XML,the document has to include a root element,and it is able to be nested among elements.In the logical structure,the document of XML is a tree with hierarchy structure,and the root element is its tree root which content is viewed as tree page[5].Essentially the XML is the best data model to describe semi-structure data.The relative database model consists of three layer model which is respectively the database,table(relation)and field(column).The document of XML can be located as a tree for arbitrary hierarchy,and it can be used for representing the relative mode[6].The data is always processed as the text to be treated,and the data conversion middleware transforms the text in XML document into the data type in the other database,and vice versa.In this way,the document of XML can be considered as to be a text file of unified standard.It is unnecessary to consider the program or special format or data type of database for each application in communication.
3)Mode mapping
In the conversion process between document of XML and database,it deals with the concepts such as element,sub-element,attribute,processing instruction,entity,entity citation,name field and so on.Moreover in the database,it is to involve the concepts such as the table,record,field,keyword,main key,view,index and so on.Generally speaking,there exists mutual mapping of template and mode between document structure of XML and database.The mapping based on template drive is to embed into the command of SQL in a template,and it applies entity software data transfer,such as the middleware,to put up the treatment.It belongs to simple hierarchy mapping.It is based on the data result set that is the result of the command being executed in the SQL ,and not deals with the relative mode or object mode that can determine whether the database being existent or not.It can only be used for transferring the data between relative database and document of XML.This sort of mapping relation only embeds the database executing instruction into the template,and the instruction is processed by database conversion middleware.Therefore the template mapping provided a great agility.
When the data is transferred into the document of XML from database or the data is transferred into the database from the document of XML,it is implemented by a model to the mapping based on model drive.The main idea is that the document of XML is represented as a tree composed by the data object,and each element type corresponds to an object in the object model.It is a deep hierarchy mapping for the model drive mapping,and it is only a mapping between the modes.The current mainstream database always supports the technology of XML,and it includes the mode mapping of data exchange of heterogeneous database based on XML and data format mapping between the XML and relative database.But it has to pay attention that when making practical heterogeneous data exchange,it is needed to the data exchange rule dictionary.Firstly the user chooses the exchange rule on client end in terms of exchange requirement,and if there is not any needed rule then it has to customize the rule.After choosing the rule,the rule parsing module in system makes analyzing and decomposition for the selected rule,and through the pick-up and conversion and read in for data and so on,the data exchange can be completed.
The importance of electric power is evident in the modern metropolis society.Once the electric power system goes wrong somewhere,it has to make emergency processing.The originated unit of accident joins to the alarm system,and informs the happened fault position.The module of alarm system will automatically create data resource of XML in terms of alarm information.The data needs to pass a series of command processing before it reaches the GIS server end,including looking for the electric source of transformer,transformer load,influencing range of power supply after the transformer breaking off,and the best emergency maintenance solution.In the decision making system,the result command,such as receiving and transmission and so on,will be produced,the information processing flowchart is shown in Figure 3.
Figure 3.Process flowchart
In the process of electric analysis,the system calls the generated GIS server interface,and it completes the electric source analysis of fault point device.The system can automatically switch the map,and it labels the path from nonce fault device to electric source and shows the path.At the same time,the end of GIS returns analysis data to the alarm system to make the space analysis of load density,and connects all the load control system to the GIS net,and the page of GIS directly displays the distributing of each load control point.By means of querying and analyzing load information of each load control point,table account and archives of each load control device,and it can analyze the singularity instance and make decision automatically for optimal power-down isolation in least power-down range.When the fault happens it deals with the shortest path analysis in patrol and engineering emergency maintenance,the system offers optimal path analysis for patrol and emergency maintenance vehicle from one point to multipoint and multi-point to one point,not only displays the graph but also gives the dispatch path table.The emergency maintenance path is shown in Figure 4.
Figure 4.Shortest path analysis
The above interaction process deals with lots of heterogeneous data system,and the alarm system can utilize the analysis data of GIS system for each step to carry through data synthesis,and also collecting and generating the report forms so as to provide the reference for emergency maintenance.Due to the system applying unified technical frame,it can easily implement the integration among various isolated heterogeneous system.The better result is taken by this method for city emergency application,such as fire protection and emergency maintenance of power fault and so on.For instance,it can enhance the processing speed from five to ten times,to support dynamic appending of application and dynamic modifying of configuration parameter.Under the condition of no power-down for the whole system,it can extend the application service provided by the system.By use of data transmission platform constituted by bottom layer,it can make storing and sending data through the reliable message queue.Due to the function of encrypt and decrypt data,it ensured the data transmission security.The client end can update the data in real time without manual intervening,and the effect of system application demonstrated that it would be better in real time.
The actual application result demonstrates that it is an excellent choice to take middleware system based on XML to implement the information exchange among heterogeneous systems.Being as a data exchange tool for XML,it has been widely used in various aspects of heterogeneous system integration,and will be a very good application foreground in future.
[1]Zhang Ning,Jia Ziyan,Shi Zhongzhi.Research on Technology of ETL in Retail Trade Data Warehouse[J].Computer engineering and applications,2002(24): 213-216.
[2]Ai Haibing,Meng Lingkui,Lin zhiyong.XML-based Integrating and Sharing of Distributed Heterogeneous Geographic Information Databases[J].Remote sensing information,2002(4): 50-56.
[3]Dubuisson O.ASN.1-Communication between Heterogeneous Systems[S].Morgan Kaufmann,October,2000.
[4]Vakali A,Catania B,Maddalena A.XML data stores:emerging practices[J].Internet Computing,IEEE,2005,9(2):62-69.
[5]Alwardt A L.Using XML transactions to perform closedloop diagnostics in network centric support environments[J].Autotestcon,2005(9):707-713.
[6]Gao yang,Tan Liming.XML-Based Method of Information Interchanging between Relative Database and Object-Oriented Database[J].Computer Systems & Applications,2003(3):196-197.