Analog circuit optimization on basis of control theory approach

• Published in: Compel 2014-01, Vol.33 (6), p.2180-2204
• Main Author: Zemliak, Alexander
• Format: Article
• Language: English
• Subjects: Analog circuits; Circuit design; Circuits; Colleges & universities; Computer simulation; Control theory; Decomposition; Electronic circuits; Genetic algorithms; Mathematical analysis; Mathematical models; Methodology; Methods; Microprocessors; Optimization; Optimization algorithms; Optimization techniques; Searching; Strategy; Transistors
• ISSN: 0332-1649; 2054-5606
• DOI: 10.1108/COMPEL-10-2013-0324

Purpose - The purpose of this paper is to define the process of analog circuit optimization on the basis of the control theory application. This approach produces many different strategies of optimization and determines the problem of searching of the best strategy in sense of minimal computer time. The determining of the best strategy of optimization and a searching of possible structure of this strategy with a minimal computer time is a principal aim of this work. Design/methodology/approach - Different kinds of strategies for circuit optimization have been evaluated from the point of view of operations' number. The generalized methodology for the optimization of analog circuit was formulated by means of the optimum control theory. The main equations for this methodology were elaborated. These equations include the special control functions that are introduced artificially. This approach generalizes the problem and generates an infinite number of different strategies of optimization. A problem of construction of the best algorithm of optimization is defined as a typical problem of the control theory. Numerical results show the possibility of application of this approach for optimization of electronic circuits and demonstrate the efficiency and perspective of the proposed methodology. Findings - Examples show that the better optimization strategies that are appeared in limits of developed approach have a significant time gain with respect to the traditional strategy. The time gain increases when the size and the complexity of the optimized circuit are increasing. An additional acceleration effect was used to improve the properties of presented optimization process. Originality/value - The obtained results show the perspectives of new approach for circuit optimization. A large set of various strategies of circuit optimization serves as a basis for searching the better strategies with a minimum computer time. The gain in processor time for the best strategy reaches till several thousands in comparison with traditional approach.