基于自适应遗传算法的模拟电路的优化设计方法

摘要

针对在模拟电路设计中参数调整复杂性带来的困难,提出了一种新的针对CMOS模拟运算放大器参数优化方法.其特点是把模拟电路设计知识与遗传算法相结合,通过对遗传算法的自适应改进帮助其解决多目标优化和收敛的问题,并根据不同的性能指标要求,在相同结构下优化出不同用途的运算放大器.实验结果证明,该方法在相同结构下与其他优化方法相比较可以精确而有效地优化出高增益、高带宽、低噪声的运算放大器.该方法适用于模拟电路优化设计:由于其基于Hspice仿真结果,更贴近于实际电路设计,具有实用价值.%A new method for optimizing the parameters of a CMOS operational amplifier based on an adaptive GA ( genetic algorithm) was presented in order to solve the difficulty caused by parameter optimization in analog circuit design. The main advantage of the method is that the problems of convergence and multiple objective optimization tasks can be solved through combining the useful features of manual analog circuit design, and adjusting the GA with the evolution process. Operational amplifiers for different uses can also be developed depending on various performance specifications. The simulation results show that this method can accurately achieve high DC-gain, high bandwidth, low noise, and low power operational amplification, and it efficiently compares with other optimizing methods having the same circuit structure. The method is suitable for CMOS analog circuit optimization. Because it is based on the simulation results of Hspice, it is much more similar to actual circuit design and therefore more useful.