Remote Control of the FESTO MPS PA Compact Workstation for the Development of a Remotely Accessible Process Control Laboratory

—The technical education sector of the UAE adopts hands-on dominated curricula with the aim of meeting the industrial needs of highly qualified technical personnel with significant practical experience in their field. The utilization of remotely accessible laboratories is one of the technologies that is expected to have a significant role in improving the attainment of hands-on skills. This paper presents the design and implementation of a LabVIEW software interface to remotely control the FESTO MPS PA compact workstations. This workstation is commonly used as laboratory equipment and for training in the process-control filed. To showcase the developed remotely accessible process control laboratory, Proportional-Integral-Derivative (PID) and ON/OFF controllers were designed and implemented on the system. The proposed remotely accessible process control laboratory offers practical and industrial training over the Internet. This feature could open doors for sharing the laboratory's resources between higher educational institutes of the UAE. Which is expected to result in the optimal utilization of the available resources. The proposed remotely accessible process control laboratory is expected to have a major cost reduction of the usually high education system cost through the efficient utilization of the available laboratory equipment.


Introduction
In today's modern world of rapidly evolving technology, it is difficult to assemble and maintain a high-performance well-educated workforce without proper education and training. One of the challenges that hinder educational organizations' abilities to provide proper practical education and training is the cost of training equipment. Educational institutes within a country or a region could overcome this problem with the

System components
The MPS PA CWS combines four control loops which are level, pressure, temperature, and flow rate control loop. The actuators that are used in the system are centrifugal pump, proportional valve, and heating element. The sensors included in the MPS PA CWS are an ultrasonic sensor, flow rate sensor, temperature sensor, and a pressure sensor. These sensors are used for measuring the level, flow rate, temperature, and pressure, respectively. The MPS PA CWS also has multiple proximity sensors and float switches. The MPS PA CWS has a terminal connection with 24 pins. This connection allows communication between the MPS PA CWS and a personal computer by using the EasyPort USB interface card. The Components of the MPS PA CWS is shown in the P&ID given in Figure 2. A short description of the aforementioned components is given in Table 1.

Easyport USB data acquisition card
The EasyPort card, developed by Festo, allows bidirectional transmission of signals from sensors and to actuators. The EasyPort is supported by several software packages that are dedicated to the simulation and development of control applications such as Easy Veep, Fluid Lab-PA, and Fluid SIM [19]. The aforementioned software packages do not allow operators to create their own control algorithms. In this work, the communication interface between the PC and the USB EasyPort will be developed based on the ActiveX elements within the LabVIEW programming environment. The EasyPort communicates with the MPS compact workstation by connecting panels dedicated to handling digital and analog signals. Eight digital inputs and outputs of the card are connected to the MPS panel by two plug sockets with 24 contacts Syslink according to IEEE 488 standard. While the four analog inputs from sensors and two analog outputs to control the pump and the proportional valve of the MPS compact workstation communicate to EasyPort by a plug socket with 15-pin Sub D connector. Analog/digital conversion of these signals is performed with a resolution of 12 bits at a sampling frequency of 5 kHz [19]. The development of applications with EasyPort can be performed with three different methods as shown in Figure 3. ActiveX controls are used to interface with the EasyPort card because they offer accessible methods and events for communication with EasyPort modules when LabVIEW is used as a development platform. ActiveX controls are software blocks based on invoke nodes shown in Figure 4. These software blocks can be used by different components of a program to communicate with each other and reuse their code. ActiveX controls are built in COM (Component Model Object) which allows them to be used in other programming languages other than the one in which they were created, among these languages stand out C++, Java, and Visual Basic.

LabVIEW Programming
This section describes the developed Virtual Instrumentations (VIs) used for operating the MPS PA CWS remotely over the Internet. Several LabVIEW VIs were created to manipulate the actuators and obtain information from the sensors of the MPS PA CWS as well as video streaming for the real-time monitoring of the stations. The user can select the desired control loop from the "Control loop selection" panel with the level being the default option. The controller design is described in Section 3. The designed panel also allows the user to choose between PID and ON/OFF controllers. The controller selection is achieved by typing-in the number of the desired controller in the "Choose a Controller" panel. The user can also adjust the parameters of the selected controller such as the differential gap, proportional gain, integral gain, derivative gain, and sampling time. As can be seen in Figure 5, the front panel as accessed over the remotely using the Windows Internet Explorer browser. The front panel shows the experimental results on the graphs at the bottom left of Figure 5. iJOE -Vol. 16, No. 5, 2020 Connect/disconnect to/from EasyPort: The invoke node "Connect' method of the FESTO ActiveX control is used in order to initiate the login phase. This method checks all serial ports found within the system for EasyPort modules as shown in Figure 6.   Analogue sensor reading: The mapping of the output voltage of the level sensor located in tank B102 is shown in Figure 8. The reading of the level sensor is obtained using the Invoke Node method with GetAnalogOutput0 action. The output of the four analog sensors is an integer between the values 0 to 32767 due to the 12-bit resolution analog/digital conversion. In order to map the output voltage to a value between 0 and 10 V, the sensors' output is divided by 3276.7. Finally, some mathematical operations were used in order to convert the reading to the desired unit which is cm in the case of the level sensor. The same invoke node method is used for reading the flow rate, pressure, and temperature sensors, by changing the action to GetAnalogInput1, GetAnalogInput2 and GetAnalogInput3 respectively. Figures 9, 10, and 11 shows the LabVIEW diagrams that are used for reading the flow, pressure, and temperature sensors, respectively.

Sending signals to the actuators of the MPS PS CWS:
The Invoke Node method with SetAlanogOutpt0 illustrated in Figure 12 is used to set the desired supply voltage for operating the pump. The proportional valve voltage is set in a similar way by changing the method in the Invoke Node to SetAlanogOutpt1, see Figure 13.

Remote operation
The remote control is accomplished through the web server tool of LabVIEW which allow guests to log in to a local host and remotely operate the MPS PA CWS. The system user can monitor the operations in real time through a camera that is mounted on the system, as shown in Figure 1.
Web publishing: This subsection addresses the implementation of a web server on LabVIEW software. The web server is needed for the remote access to the MPS PA CWS. Figure 14 is an explanatory diagram of the process instrumentation and remote monitoring to be carried out on the MPS CWS. The remote monitoring is realized by the IP camera that is used by the operator to observe the state of the MPS PA CWS. Therefore, it is necessary to explain the configurations and concepts of certain related elements of hardware and software as follows. http://www.i-joe.org

Implementation of Web server:
The main requirement for the configuration of the web server within LabVIEW VI is to have a functioning online version of the front panel used to control the MPS PA CWS. First, the "Tools" option in the LabVIEW toolbar is selected. Then one has to choose "Options" and finally select "Web Options"; a window opens as displayed in Figure 15. In this window, we proceed to enable the "Enable Remote Panel Server" option and then configure the server communication port. After configuring the server port, "Web Publishing Tool" option is selected, a window will open as shown in Figure 16, where the target VI to be published on the web server is selected. One can also configure the type of interaction between the server and client, enabling or disabling the "Request Control when the connection is established" option. If this option is enabled, the customer must request control of the VI after the connection, otherwise the user takes control of the VI immediately once connected to the server has been established. Finally, the window shown in Figure 17 will appear, this window specifies the address where the file will be saved on the computer that will serve as a server and also the route used to enter the control panel of the plant via the web.

System identification
It is well known that the plant mathematical model should resemble the dynamics of the system as close as possible. This is desired in order to achieve a good controller response as the controller is designed based on the obtained model. The system in hand is a laboratory scale; therefore, there will be no significant change in the system's dynamics. Thus, the system can be approximated by a single transfer function. The MATLAB system identification (Ident) tool is used to estimate the system transfer function form open-loop experimental date. The following process was used to obtain the mathematical model for the level, flow rate, pressure, and temperate control loops. However, only the level control loop system identification is presented here. The centrifugal pump was operated on open-loop mode with Pseudorandom Binary Sequence (PRBS) input. The PRBS was chosen because it persistently stimulates the system; i.e. the signal PRBS provide sufficient energy to the process to overcome the noise and disturbances. In addition, the PRBS signal operates between two random fixed limits, and when injected into the process it excites the fast and slow process dynamics. The PRBS is shown in Figure 18. The liquid level on the upper tank is sensed with the ultrasonic level sensor. The PRBS input and the sensor output are logged in an excel file that is used with the Ident tool for obtaining the system transfer function. The system identification window is shown in Figure 19, after importing the time-domain data for the experiments which are the PRBS input and the measured level. The obtained system model is validated as follows; the validation process starts by dragging the input signal to the box "Validation Data" on the main screen of the system identification tool. Then by selecting "Model Output" which will open the window shown in Figure 20. Figure 21 shows the transfer function estimation report for the best fit and Figure 22 shows the output of the transfer function with the highest confidence level along with the experimental data.   The maximum rate of 92.42% overlap is a transfer function of 2 poles and 2 zeros. The transfer function exported to MATLAB Workspace is given in equation 1. The system identification results are summarized in Table 2. (1)

Control design
ON/OFF controller: The ON/OFF controller is a digital controller, which is the simplest form of control. The ON/OFF control action turns the pump ON or OFF based on the level setpoint. The output frequently changes according to actual level readings. To prevent rapid switching of the control action from happening, a level band called hysteresis is created between the ON and OFF operations. The implemented ON/OFF controller is shown in Figure 23.

PID controller:
The MATLAB Control System Designer Toolbox provides multiple options for the automatic tuning of PID controllers. For example, controllers can be tuned using classical tuning methods, graphically by varying the root locus, or the frequency responds. Figure 24 shows the main screen of the Control System Designer Toolbox. The system transfer function obtained in the previous section is imported to the Control System Designer Toolbox and the automatic PID tuning with robust response time is used to obtain the PID compensator transfer function as shown in Figure  25.  The resulted transfer function is given in Equation 2. ( The values of the controller gains are , and .
These values were obtained from fine-tuning after implementing the designed gains on the system. The implemented discrete form of the PID controllers is given in Equation  3. ( The implemented PID controller in LabVIEW is shown in Figure 26.

Experimental Results and Discussions
The PID and ON/OFF controllers that were designed in the previous sections were implemented on the developed software interface. The experiential results for the ON/OFF controller and the PID controller are presented in this section. The experimental result for the ON/OFF controller is shown in Figure 27. The setpoint is 8 cm and the differential gap is set to 2 cm. The ON/OFF controller is operated as follows: the ON/OFF controller is selected by typing number '2' in the "choose a controller" panel. Then, the setpoint and differential gap values can be selected. The PID controller for the level control system is operated as follows: the PID controller is selected by typing number '1' in the "choose a controller" panel. The PID controller parameters are selected and can be tuned online as desired; then, the value of the setpoint is selected. In the provided example, the setpoint was chosen to be 8 cm. The experimental result for the PID controller is depicted in Figure 28. The maximum overshoot is 10% and the settling time is 10 sec.

Conclusion
In conclusion, this paper presented a special online experimental platform that allows remote access to the FESTO MPS PA CWS used for education and training in the process control and the mechatronics fields. Custom-built LabVIEW VIs were used to interface with the MPS PA CWS through the EasyPort data actuation board. After designing the components that allow reading the sensors of the MPS PA CWS and sending data to its actuators; PID and ON/OFF controllers were implemented on the system. The designed platform is published using the Web Server tool of LabVIEW to allow remote access while viewing the system operations in real time through a camera that is mounted on the system. The proposed remotely accessible process control laboratory can be used to conduct online experiments and /or training to students at other universities in the UAE. In addition, this platform is expected to result in a significant cost reduction and efficient utilization of the available laboratory equipment at IAT.