Abstract: This paper introduces the structure and fault diagnosis methods of a Web-based fault diagnosis system, and describes the implementation of online diagnosis using ASP and MATLAB technologies, network diagnosis using LabVIEW's DataSocket technology, and how to design COM components using Visual C++ to implement a Web-based fault diagnosis method. Finally, the three network fault diagnosis methods are compared. Keywords: Fault diagnosis; Component; MATLAB; LabVIEW With the rapid development of Internet and Web technologies, the establishment of open and scalable remote fault diagnosis systems has become a reality. This is a product of the combination of Internet technology, communication technology, and mechanical equipment fault diagnosis technology. With the increasing popularity of the Internet, the monitoring and diagnosis mode of electromechanical equipment is undergoing a new revolution. After the development from stand-alone closed monitoring and diagnosis systems to distributed monitoring and diagnosis systems (DMDS) based on industrial local area networks, the research and application of wide area network-based remote fault diagnosis systems have received widespread attention. Structure of Web-based Remote Online Fault Diagnosis System Remote fault diagnosis systems employ a variety of technologies, and research in this field is increasingly in-depth both domestically and internationally, with multiple approaches to research scheme selection. This system adopts a B/S architecture. The diagnostic server integrates MATLAB diagnostic technology, LabVIEW diagnostic technology, and COM diagnostic component technology, and has explored some new approaches. The structure of the remote fault diagnosis system is shown in Figure 1. Figure 1: Monitoring and Diagnostic System Structure and Diagnostic Process The fault diagnosis process of the Web-based remote fault diagnosis system is as follows: The data acquisition system at the equipment operation site uses sensors to acquire online status information such as vibration, temperature, current, and voltage, based on the equipment's operating characteristics. Simultaneously, the computer monitoring system processes the acquired signals in real time, displays the values ​​and graphs of various signals, and determines whether the equipment is operating normally. If an abnormality occurs, an alarm signal is immediately issued. At this point, the user only needs to open a web browser and enter the URL address of the fault diagnosis center to access the fault diagnosis center server. This site provides web page guidance for users, who interactively input the relevant observation data to be diagnosed, select the relevant diagnostic methods, and submit the data to the web server. The web server then executes an ASP program to process the user request, calls the fault diagnosis component on the fault diagnosis server, performs diagnostic analysis on the data, and returns the processing results to the client browser in the form of a webpage. The user can then view the diagnostic results in their browser. The system structure is shown in Figure 1. Key Technologies Web-based fault diagnosis mainly utilizes dynamic webpage technology to create highly interactive webpages, which are launched by a remote client browser to diagnose the equipment. This article mainly introduces the key software implementation technologies: database access technology, fault diagnosis component development technology, and their calling technology in the webpage. Database Access Technology The database server is installed with Microsoft SQL Server 2000. Data obtained from the on-site online monitoring system is sent to the database server in real time via the network for storage and management. Currently, there are various technologies for accessing web databases in the Windows environment; here, ActiveX Data Objects (ADO) is used. ActiveX Data Objects (ADO) is the core technology of Universal Data Access (UDA). When a user requests an ASP homepage using a browser, the web server responds by calling the ASP engine to execute the ASP file, interpreting the scripting language (JavaScript or VBScript), using the ADO (ActiveX Data Objects) component to perform database access operations, and finally, the ASP generates an HTML page containing the data query results and returns it to the user's client for display. The key code for accessing a SQL Server database in ASP is as follows: <% set conn = Server.CreateObject(“ADODB.Connection”) // Establish connection object conn.Open con.open “PROVIDER=SQLOLEDB;DATASOURCE=SQL server name;UID=sa;PWD=database password;DATABASE=database name” sqlStr = “SQL statement '” // Specify the SQL command to execute rs = conn.Execute(sqlStr) rs = Server.CreateObject(“ADODB.RecordSet”) // Create recordset object rs.MoveNext // Point to the next record rs.close() // Close the connection conn.close() // Close the database %> The compilation of fault diagnosis components and their calling technology in web pages: MATLAB diagnostic technology. MATLAB stands for Matrix Laboratory. It is a high-performance numerical computation and visualization mathematical software launched by MathWorks in 1982. In system development, signal processing methods such as time-domain waveform analysis, power spectral analysis, correlation analysis, filtering, cepstral analysis, frequency refinement, curve fitting, and wavelet analysis are compiled into corresponding m-files and stored on an SQL Server database server, ready for use. Client users select any of these methods and submit them to the server for execution. The server then returns the analysis results to the client's browser. MATLAB provides a COM/DCOM interface, allowing users to execute MATLAB commands within ASP web applications. Below is the main code of the m-file for wavelet analysis written in MATLAB (the result is shown in Figure 2): Figure 2 Wavelet analysis using Matlab function wavelet(J, K, CH, FS, FILENAME, SCALE, OFFSET) // CH is the sampling channel number; FS is the sampling frequency DATA=load(FILENAME); [lc]=size(DATA); N=l; SIGNAL=DATA(:, CH)3SCALE+OFFSET; T1=0; T2=(N-1)/FS; T=T2-T1; [c, ll]=wavedec(SIGNAL, J, 'db10'); // Scale UD(:, 1)=wrcoef('a', c, ll, 'db10', J); For j=2:1:J+1 UD(:, j)=wrcoef('d', c, ll, 'db10', j-1); End t=(0:(N-1))/FS; P=UD(:, K); P=P'; P = mat2str(P3100, 3); The key code for calling the MATLAB application using ASP is as follows: <% SetMatlabapp = CreateObject(“MATLAB.Application”) // Create object String Return = Matlabapp.Execute(“MATLAB statement”) // Execute MATLAB algorithm SetMatlabapp = Nothing // Close object %> LabVIEW Diagnostic Technology LabVIEW is a development tool that uses graphical code to create applications instead of programming languages. It fully utilizes the advantages of PCs being inexpensive (compared to other measurement and control instruments) and powerful. Programs are written using a graphical language (G language), icons, and lines instead of text. Its biggest advantage is its simple programming and user-friendly interface, which makes it easy to present experimental data in graphical form. Virtual instruments can be designed in LabVIEW to display values ​​such as speed, displacement, and power. Users can access system settings and current operating status in a browser, and system operating status can be shared. GWebServer can publish the front panel of virtual instruments designed in the LabVIEW environment to an Intranet or the Internet, allowing remote users to use the virtual instruments running on-site through a browser. By utilizing GWebServer and combining it with the CGIVIs and CGIANimations programming tools provided by LabVIEW, user exchange functions such as start/stop, parameter passing, and password protection can be implemented within a browser. LabVIEW employs DataSocket technology to solve network programming problems. DataSocket is a new network programming technology based on the TCP/IP protocol provided by NI. It supports local file I/O operations, FTP and HTTP file transfer, real-time data sharing, and provides a universal API programming interface. It consists of two parts: DataSocket API and DataSocketServer. The former provides a unified programming interface, allowing clients to communicate with various data types in various programming environments. NI provides ActiveX controls for DataSocket, the LABWindows/CVIC library, and a series of LabVIEW Vis, so it can be used in any programming environment. This project uses ActiveX controls, published via DataSocketServer and the DataSocket API program, to broadcast real-time data to multiple remote clients at high speed. Furthermore, the DataSocket ActiveX control is further encapsulated to implement a functional ActiveX control and integrated into the web page, while simultaneously publishing the field data to DataSocketServer. In this way, users can access the web page through a browser. The filtering implemented using this method is shown in Figure 3. COM Component Diagnostic Technology: COM components are executable binary code written according to the COM specification and distributed as Win32 dynamic link libraries (DLLs) or executable files (EXEs). ActiveX controls are the most widely used COM components. Through ActiveX technology, programmers can assemble reusable software components into applications. A control consists of three parts: the control's appearance is public, visible to the user and interactive; the control's interface, including all its properties, methods, and events, is also public, used by any program containing an instance of the control; the control's private part is its implementation, i.e., the code that makes the control work. In other words, the control's implementation effect is visible, but the code itself is not. Users can inherit the control's private part and modify its visible part to match new application requirements. This project uses VC++ to create a custom ActiveX control to draw the waveform of the acquired signal. When developing ActiveX controls using VC++, there are two commonly used choices: Microsoft Foundation Class Library (MFC) and Active Template Library (ATL). Here, ATL is used for development. Figure 3 shows filtering using LabVIEW. ATL provides a small and sophisticated framework for generating ActiveX controls. This framework makes it easy to build small, fast, COM-based components. Using ATL gives software developers greater flexibility when implementing components, without relying on any auxiliary DLL files. ATL is a dynamic platform for developing ActiveX components, and it will gradually become the mainstream tool for developing ActiveX components as it continues to evolve. However, MFC's strength is also its weakness—its lack of support for general classes and tools. In the VC++ 6.0 development platform, create an ATL project using "ATLCOMAppWizard," insert a NewATL0bject, complete the object's property settings, click Next, and enter the control name Drboxt in the shortname field of the Name table. You will then see that the wizard automatically generates names for various items in other columns; these names can be modified. Click OK. Opening ClassView will then reveal a newly added class CDrboxt and an interface IDrboxt, prepared for the newly created control. The declarations of the control's properties and methods are added to the interface declaration, while the implementations are added to the newly added class. The interface declaration is in an IDL (InterfaceDefineLanguage) file. Since this control will be used on a network, its security must be configured. Here, IObjectSafety is used to mark the control as initialization/script safe. The definition of the CDrboxt class is shown in Figure 4. class ATL_NO_VTABLECDrboxt: public CComObjectRootEx , publicIObjectSafety, …… {public: STDMETHOD(get_Yunit)(/*[out,retval]*/short*pVal); STDMETHOD(start)(); UINTm_timer; ……} The key code of the ASP calling component is as follows: <% Set Obj=Server.CreateObject("Huabo.Drboxt")//Create object obj.start()//Call the method to draw waveform set? Obj=nothing//Release object Figure 4 Comparison of three diagnostic techniques for COM component implementation of waveform The above three technologies for implementing Web-based fault diagnosis software have the following characteristics: (1) MATLAB language is easy to learn and use, does not require users to have profound mathematical and programming language knowledge, and does not require users to have a deep understanding of algorithms and programming skills. MATLAB is both a programming environment and a programming language. This language, like C, FORTRAN and other languages, has its own rules, but MATLAB's rules are closer to mathematical representation, making it easier to use and allowing users to save design time and improve design quality. The application of MATLAB technology does not require the development of special software. It comes with a large number of signal processing tools that can be directly called, which greatly reduces the requirements for users' mathematical foundation and computer language knowledge. Moreover, the programming efficiency and calculation efficiency are extremely high. Results and beautiful graphic copies can be directly output on the computer, and the cost is low. (2) The advantages of virtual instrument technology are: no need to develop special client software, good openness, and users can join at any time; no need to write special gateway processing programs on the server side, which can be directly implemented using the VI of the toolkit; the original LABVIEW application can be upgraded to the network environment, and data and information can be automatically transmitted through FTP and SMTP, which is convenient for further remote analysis. Virtual instruments are not only diverse in function and accurate in measurement, but also have a user-friendly interface, simple operation, and low maintenance costs. It is convenient and flexible to integrate with other devices and has expandable functions. (3) COM component technology requires writing programs, which is difficult to develop and debug. It is technically difficult and has certain requirements for developers. The development cost is also higher than the other two methods. However, upgrading and modifying components does not require modifying the page, so it has good scalability. Moreover, after development, the components are easy to call, save code, have high running efficiency, and are easy to use and manage. Furthermore, the compiled components are distributed in binary form, enabling cross-platform use on Windows, and the source code is not leaked, effectively protecting the copyright of the component developers. Conclusion This project implemented online diagnostic methods using ASP and MATLAB technologies, network diagnostic methods using LabVIEW's DataSocket technology, and how to design COM components using Visual C++ to implement web-based fault diagnosis methods. It also explored how to integrate these three technologies to provide users with multiple diagnostic methods.