In this work, a study case is presented in which the design of the layout for a CMOS sensor cell is partially automated by implementing a metaheuristic algorithm to find the best tradeoff between two conflicting objectives (two quantitative opposite and not totally independent yet desired performance or design qualities) among the set of feasible layout and electronic device configurations within a constricted search space. The feasibility of a solution (a particular configuration) and its capability to fulfill every requested objective, is determined by its compliance to the CMOS-MEMS design rules and fabrication process. Any given solution besides showing optimal or very near-to-the-optimal characteristics, must be suitable to be fabricated in the CMOS conventional process which for this case is a \(0.5\,\mu \hbox {m}\), 3-metal 2-poly N-well fabrication, beside this, since monolithic inertial sensors generally contains embedded movable electromechanical parts a surface micromachining must be considered. Simulation data and behavior of the bio-inspired metaheuristic algorithm used during the design process are presented, as well as electromechanical simulation results based the automatic-generated solutions.

在这项工作中，提出了一个研究案例，其中通过实施元启发式算法在两个相互矛盾的目标（两个相对的，相对不完全独立但尚未达到预期性能的目标）之间找到最佳折衷方案，从而使CMOS传感器单元的布局设计部分自动化。设计质量）在狭窄的搜索空间内的一组可行布局和电子设备配置中。解决方案（特定配置）的可行性及其实现每个要求目标的能力，取决于其对CMOS-MEMS设计规则和制造工艺的遵守程度。除了显示最佳或非常接近最佳特性之外，任何给定的解决方案都必须适合在CMOS常规工艺中制造，在这种情况下，该工艺为\（0.5 \，\ mu \ hbox {m} \）此外，由于单片惯性传感器通常包含嵌入式可移动机电部件，因此需要进行3微金属2多晶硅N阱制造，因此必须考虑表面微加工。给出了设计过程中使用的生物启发式元启发式算法的仿真数据和行为，以及基于自动生成的解决方案的机电仿真结果。