微机电系统(MEMS)器件在军用和民用领域应用都比较广泛,其设计研究涉及到机械、电子、磁场等领域,属于多学科问题,对其进行优化设计比较困难。文中以微加速度计ADXL150的优化设计为例以SUGAR软件作为分析器提出采用导重法对微机电系统器件进行优化设计。该方法与国外采用的遗传算法和退火算法等自然模拟优化算法相比,计算效率与计算精度大大提高,采用遗传算法和退火算法需要上万次分析,计算时间达4~5 h,而该方法仅需六七次优化迭代几分钟时间即可找到最优解。 Microelectromechanical systems (MEMS) devices are widely used in military and civil fields. Their design and research involve mechanical, electronic, magnetic field and other fields, which are multidisciplinary issues. It is difficult to optimize them. In this paper, the micro-accelerometer ADXL150 optimization design as an example to SUGAR software as the analyzer proposed using guided-weight method to optimize the design of MEMS devices. Compared with the natural simulation optimization algorithms such as Genetic Algorithm and Annealing Algorithm used in foreign countries, this method greatly improves the computational efficiency and accuracy. Genetic algorithms and annealing algorithms need tens of thousands of analyzes, and the calculation time is up to 4 ~ 5 h. However, this method The optimal solution can be found in just six or seven optimization iterations for a few minutes.