Embedded systems, utilizing microprocessing devices to perform specific tasks, are rapidly finding their ways in a variety of machines such as medical devices, automobiles, aircrafts, home appliances, toys, mobile phones, MP3 players and many more. A successful strategy for the development of embedded control systems should consider cost efficiency, optimal performance, reliability, reduced time to market and fast deployment. In practice, several iterations of computationally expensive, complex and time-consuming procedures for realization of an embedded control system are commonly performed before finding an acceptable model of the system and achieving the desirable control objectives.; In this dissertation, a novel real-time, adaptive, model-based, optimal control is designed for embedded systems. The proposed strategy is based on a new recursive subspace identification and receding horizon optimal control. The recursive subspace identification strategy can be parameterized in terms of recursive approximation of subspace intersections and adaptive estimation of state sequences. Ultimately, the integrated modeling-control strategy can be implemented, in real time, with minimum knowledge of the controlled system. The real-time optimal strategy has the following properties: (a) It is implementable in real time; (b) It is adaptive, in that the identified model captures the changes in the system; (c) It is an appropriate framework for control of multivariable systems with both state and control constraints.; The proposed real-time adaptive optimal controller is then implemented in simulation and hardware implementation for both single-input-single-output (SISO) and multiple-input-multiple-output (MIMO) systems for verification and demonstration. Actual hardware experiments on active vibration control and active acoustic noise cancellation problems using a digital signal processor (DSP) such as dSPACE DS1104 controller board have been carried out in order to verify the performance of the proposed methodology.
展开▼
