| This thesis is concerned with development of embedded adaptive controllers for industrial applications. Many industrial processes present challenging control problems such as high nonlinearity, time-varying dynamic behaviors, and unpredictable external disturbances. Conventional controllers are too limited to successfully resolve these problems. Therefore, the adaptive control strategy, an advanced control theory, is applied to overcome deficiencies of the conventional controllers.;Through the thesis, an embedded adaptive controller is designed and implemented for the specific case study, a gasoline-refining plant. The adaptive controller design is initially achieved in continuous-time space and then converted to discrete-time space by using z-transform. It is finally implemented on an advanced reduced instruction set computer machine (ARM) processor. A plant simulator written in C++ executes functions of the gasoline-refining plant. Therefore, an integrated testing environment is developed in order that the embedded adaptive controller can interact in real-time fashion with the plant simulator located in a remote computer. In all system tests, the embedded adaptive controller successfully controlled the remote plant simulator and fully satisfied all control objectives. |
