PENGEMBANGAN TEKNIK KOMPUTASI UNTUK SINTESIS SISTEM DISIPATIF

• Main Author: Kusuma, Idris
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/1583
• Institution: Institut Teknologi Bandung

In general, the behaviour of system is nonlinear and we can't always expect the system behaves as linear, so that the designing theory of linear control can't be directly applied on the nonlinear system. Therefore, it is needed the designing theory of control served as globally. One of the theory is a theory of dissipative system. The problem faced on the dissipative system is no availability of a solution enabling the taking place of the system analysis and synthesis. Here we realized that it is necessarily developed a computational technique as a tool. The computational technique developed uses a scheme of finite difference with an approach method of value space. This computational technique is implemented for the controlling synthesis of state feedback and output feedback. In the synthesis of state feedback, the computational technique has been developed into the two iteration processes, min-max and max min iterations. The results of the iteration processes show that the increase of energy in the system with the min-max iteration is smaller for the linear system and it is nearly same for the nonlinear system compared with the max min iteration. However, the process of minmax iteration is more practical and natural for the profile of storage function and the resulting control law. For the reason, the min-max iteration is used for the analysis of the performance of supply rate on the dissipative system. The result of the analysis of the quadratic supply rate indicates that to synthesize the dissipative controller there isa designing parameter represented by the value of 0, it can be adjusted to get several the system's performances. The computational technique is developed for the problem of output feedback based on the Certainty Equivalence Principle (CEP). The analysis of output feedback is conducted in each point in the space of information state through the state estimation to establish the control law from output signal measurement.