Makoto Nakaya

Application of Tracking Simulator to Reforming Process

We are researching advanced operation support system which can help operators and provide them with novel knowledge. Advanced operation support system needs more essential information of plants such as temperature of reactor inside, real-time data of composition, prediction of the process behavior and so on. We use simulators as a way of getting such essential information of plants from the experience that we have developed process simulator. Our process simulator is strictly based on chemical engineering principle, namely, rigorous model which makes accurate simulation possible. However, it is hard to keep simulation highly-accurate against many reasons, e.g., change of process characteristics with degradation of plant equipments or catalysts. For this reason, we have proposed tracking simulator which runs in parallel with an actual process and adjust its parameters online. We developed an experimental steam reforming plant generating hydrogen from methane gas and the corresponding model for demonstration of tracking simulator. And then, we run tracking simulator with the process to visualize composition of reactor in real-time as an example of operation support

MODEL PARAMETER ESTIMATION BY TRACKING SIMULATOR FOR THE INNOVATION OF PLANT OPERATION

Abstract This paper proposes a new application of the tracking simulator relevant to future plant operation. The tracking simulation technique will enable matching of physical and virtual worlds. The simulator receives sensor data from the plant and the model parameters in the simulator are constantly adjusted to match the actual plant behavior. Model parameter estimation is one application of the tracking simulator. Here, the tracking simulator showed good simulation accuracy with model parameters to estimate the functions of its related variables. A fuel cell simulator was used for model parameter estimation in this study.

Innovative plant operations by using tracking simulator

Plant design and offline analysis of plant behavior are increasingly being carried out using plant simulators. However, online simulators are rarely used in large plants. The dynamic simulator Visual Modeler developed by Omega Simulation Co., Ltd. is a well-established Operator Training Simulator, and is widely used in training. It also provides similar or higher-speed computation performance than real-time computation in large plants. Using this technology, an online dynamic simulator, which can keep adjusting the parameters of a plant model to correspond with the actual plantpsilas parameters as closely as possible, is called a tracking simulator. We have developed a dynamic data reconciliation technology that enables the dynamic simulator to adjust the parameters of a plant model to the actual plantpsilas parameters even when plant behavior is not steady. This technology will facilitate early fault detection in processes and prediction of future plant behavior, thereby assisting innovation of plant operations.

An Application of On-Line Tracking Simulator to a PEMFC

The polymer electrolyte membrane fuel cell (PEMFC) presents number of control problems due to its complex dynamics. Water management is known to be one of the critical issues for efficient operation of the PEMFC. But water transport process inside PEMFC is a complex phenomenon, making it hard to predict output voltage at highly humidified condition by use of off-line simulator. Recently physical model-based controls of PEMFC are proposed to achieve efficient operation of PEMFC. However to realize the model-based control of PEMFC at high energy regions, the construction of high precision, high speed dynamical simulator with water process model becomes indispensable. We have developed an on-line tracking simulator, simultaneously running with actual process to adjust model parameter of the simulator to the actual process in real-time. By applying on-line tracking simulator to PEMFC experiment, we have performed an experimental study of on-line tracking simulator to a PEMFC. The experimental study validated the use of on-line tracking simulator to predict accurately the output voltage of PEMFC even at highly humidified condition where uncertain liquid water processes becomes apparent

Hybrid plant model of physical and statistical model with robust updating method

This paper introduces an architecture of hybrid model of physical and statistical model with adaptive updating method based on Kalman Filter and just-in-time (JIT) modeling. When the real time reference output is accessible, the powerful Kalman filter is recommended to update the statistical linear model. In addition, an effective initial value calculating method is proposed to perfect the powerful Kalman filter updating mechanics. When the reference output is not available, Kalman filter can not work anymore, just-in-time (JIT) updating method is brought in to update statistical models. The validity of this architecture is clarified through a case study of Methane-Steam Reforming process.

Process Education and Development Infrastructure Building by Virtual Plant in DCS Manufacturing Company

Abstract When distributed control system (DCS) was first developed, there were a number of control engineers with deep knowledge of customers process in DCS manufacturing companies. At present, however, engineers working in DCS companies are strongly involved in new technologies, such as networking and security. Therefore, DCS engineers are somewhat isolated from the customers process. To gain a better understanding of the customers process and operation, a vinyl acetate monomer (VAM) virtual plant was developed. Using this virtual plant, we allowed engineers involved in the development of operation support software to experience plant operation. This training was effective because the engineers gained insights into both advantages and disadvantages of their designs. The VAM virtual plant was utilized as not only a training tool but also as a development infrastructure. This paper reports that the virtual plant contributes to training of young engineers and facilitates product development in a DCS manufacturing company.

Online Process Simulator with Hybrid Model of Physical Model and Just-In-Time Model

Abstract A tracking simulator that runs in parallel with the plant and perfectly simulates the plant behavior has been developed. This tracking simulator is currently in the validation phase at a commercial plant site. The tracking simulator strongly depends on the physical model based on physical and chemical laws, and it cannot simulate unknown phenomena in the plant. A hybrid model consisting of a physical model and a statistical model using the historical plant operation data was proposed. The estimation accuracy of the statistical model was improved using the virtual input from the tracking simulator. Although the Kalman filter technique is applied to conventional software sensors for quality estimation as a model updating function, deterioration of estimation accuracy is often seen with marked changes in plant behavior and the analyzers shut down. To resolve these issues, we applied Just-In-Time modeling based on the historical data modeling technique with the tracking simulator, and the plant model coverage of the tracking simulator was expanded for various processes.

Parameter estimation of PEMFC by on-Line tracking simulator

The tracking simulator receives the sensor-measured data directly from the plant and the model parameters inside the simulator are adjusted in order that the output of simulator agrees with data from the plant. Therefore its possible to estimate the simulation model parameters which changes with an elapse of time and operational status by the tracking simulator. We report that more precise simulation can be done to the fuel cell by utilizing the tracking simulator.

Application of economic MPC to a CSTR process

This paper presents an application of Economic Model Predictive Control (EMPC) to a continuously stirred tank reactor (CSTR). As a kind of nonlinear MPC, mechanism models are employed to predict and optimize. The control strategy minimizes the economic objective function under the process constraints composed by mechanism models, bounds, quality and security etc. Several kinds of objective functions for different demands are also analyzed and discussed. The effectiveness of EMPC is illustrated with some numerical simulations.