An evolutionary-based methodology for symbolic simplification of analog circuits using genetic algorithm and simulated annealing

摘要

In this paper, an evolutionary-based multi-objective criterion is introduced for simplified symbolic small-signal analysis of analog circuits containing MOSFETs. After circuit analysis via modified nodal analysis technique, the derived exact symbolic transfer function of the circuit behavior is automatically simplified. In contrast to traditional simplification criteria, the main objective of our criterion is to control the final simplification error rate. The proposed simplification methodology can be performed by such optimization algorithms as local-search algorithms, heuristic algorithms, swarm intelligence algorithms, etc. In this paper, a hybrid algorithm based on genetic algorithm and simulated annealing is applied to validate the proposed methodology. It is remarkable that all steps including netlist text processing, symbolic analysis, post-processing, simplification, and numerical analysis are consecutively derived in an m-file MATLAB program. The proposed methodology was successfully tested on three analog circuits, and the numerical results were compared with HSPICE.