Genetic algorithm and response surface optimization of multi-scale sensor structure
In order to design a low-range capacitive accelerometer with high overload resistance and high sensitivity,seven sets of size parameters of the sensor are optimized by combining response surface analysis and genetic algorithm.The functional relationship between the size parameter and the sensor output is determined by using multiple sets of sampling points in design of experiment(DOE)and quadratic fitting.A penalty factor is then introduced to this function to perform multiple genetic algorithm iterations.Finally,through the results of the response surface optimization and genetic algorithm,a set of optimal size parameters are obtained.The optimization results show that compared to before,the equivalent stress of the sensor is reduced by about 18％,the equivalent displacement is increased by about 12％ and the design sensitivity is increased by about 20％.And the maximum stress under 30000 gn overload is reduced by 224.94 MPa,which proves the anti-overload ability of the sensor has a significant improvement.
response surface optimizationgenetic algorithms(GA)anti-overloadlow range
万方数据
维普
