This method allows software applications to become more accurate at predicting outcomes . For example , Neural Network which is one of a machine learning tools that can be applied to learn the dynamics and behaviour of a chemical process [ 2 ]. The Neural Network model can be utilised to design an efficient process to control the chemical process .
The concept of machine learning has been included in chemical engineering teaching and learning in XMUM . In the Process Control and Instrumentation course , students are given an assignment to develop a dynamic model and process control for chemical processes . In this assignment , the concept of machine learning can be applied where an open loop simulation is carried out to analyse the dynamic behaviour of a process . The simulation can be performed using the Scilab XCOS software . The analysis of process dynamics provides an insight into the stability , nonlinearity , and other characteristics of the chemical processes .
After the analysis , a process controller can be designed , and a closed loop simulation be performed . Figure 12 shows a closed loop simulation of a fluidised bed reactor for Polyethylene production performed by students . Figure 13 shows the results of these simulations including the temperature profile of different PID tuning settings for the fluidised bed reactor . It has been shown that the PID controller can successfully control the reactor temperature . The behaviour of a Biodiesel model can also be analysed and used as a reference to design a suitable PID controller . As shown in Figure 14 , students simulate the transesterification process for Biodiesel production . Even though the process is nonlinear , the Biodiesel concentration can be controlled efficiently by a PID controller as shown in Figure 15 . Students are able to apply the machine learning concept , which is a modelbased control system , by analysing the model ’ s behaviour prior to designing a controller .
Simulation Integrated Pilot Plant
In addition to embracing specialist software to prepare our students for Industry 4.0 , at XMUM , we are constantly engaged in incorporating new digital technologies in our curriculum . In traditional chemical engineering education , practical training is incorporated in the curriculum in the form of industrial training as well as industrial visits . As students are not qualified engineers , they are generally not allowed to perform essential operations in these environments . More importantly , the chemical plants in the industry often involve equipment with elevated temperature and pressure , which could be a considerable safety risk to students . Therefore , they are constrained to cognitive enhancement with limited psychomotor training in the chemical plant .
In 2018 , XMUM started a joint-development project with Qinhuangdao Bohe Science and Technology Development Co ., China , to develop a Simulation Integrated Process Plant ( SIPP ) on our campus . SIPP consists of scaled-down equipment in a chemical plant . The operations of the equipment are controlled using an industrial grade Distributed Control System ( DCS ) coupled with an advanced computer algorithm that allows them to mimic the actual operation in an industrial environment . Students will be able to perform actual operations under various plant environments such as normal operation , startup , shut down , and various commonly occurred industrial incidents without worrying about their safety and any damage to the equipment that might invite any financial repercussions .
Figure 16 : 3D Model of SIPP
To further enhance Industry 4.0 in our curriculum , SIPP is equipped with a 3D Virtual Factory mode . A Virtual Reality ( VR ) simulator is used to allow students to experience the process plant in VR mode and access critical information of the process plant virtually .
39