Adnan Khalid

Dynamic System Modeling of Industrial Boiler using Fuzzy Model

Industrial processes are commonly nonlinear in nature and include time varying parameters and uncertainties. The basic approach for modeling the system is to use first principle methods e.g. differential equations. One of the major hurdles in developing differential equation model is the complete understanding of the dynamics of the process. The first principle models cannot replicate the true behavior of the system due to many factors. Another approach is based on heuristics. The difficulty behind heuristics based design is that situation arise when unexpected process variables are present and are outside the operator`s experience. These drawbacks motivate us towards model based designing of fuzzy controllers.

Fuzzy models become useful when a system cannot be defined in precise mathematical terms. The non-fuzzy or traditional representations require a well structured model and well defined model parameters. Even if the structure is known, numerical model representations usually become irrelevant and computationally inefficient as the complexity increases. Moreover, there may be a lot of uncertainties, unpredictable dynamics and other unknown phenomena that cannot be mathematically modeled at all.

Therefore, when a system cannot be modeled with traditional methods for the reasons stated above, the fuzzy modeling approach offers an efficient mathematical tool in handling many practical problems. The main contribution of fuzzy system identification techniques is its ability to handle many practical problems that cannot be adequately modeled by conventional modeling techniques.

This thesis will present an adaptive system identification method based on Takagi-Sugeno fuzzy inference systems. A NARX model will be used for the dynamic modeling of an industrial boiler. The modeling knowledge will be applied in cases when the process is under severe changes and the knowledge to control the process is no longer available. This approach uses fuzzy model in Boiler Operation control where the fuzzy model is used to capture human expert knowledge about the process.