The Robust Design Method Research On MEMS Pressure Sensor

The MEMS sensor use silicon as the material, and use micro-machining technology and integrated circuit manufacturing technology to achieve miniaturization and integration of sensor structure. MEMS sensors are widely used in industrial automation, aerospace, electronic communications, automotive electronics, environmental monitoring, agriculture, bio-medical, military and other fields.The paper designs a MEMS piezoresistive pressure sensor, its range of0to100Kpa. The structure size of the silicon and the parameter of force-sensitive resistance is mainly designed; The simulation model of the output of the sensor is build, which using finite element method and the path integral method, and complete the simulation and calculation of the sensor output; The are two methods be used to complete the robust design of MEMS pressure sensors, which is Taguchi robust design method and robust design method based on stochastic modelsAt first, The basic theoretical knowledge of sensor be used, Such as the theory of piezoresistive effect of silicon, the elastic deformation of the diaphragm deflection, Wheatstone bridge, analyzed the working principle of the sensor, on this basis, to complete the design of the sensor silicon cup size parameters and structural dimensions of force sensitive resistors. The three-dimensional solid model of the sensor to be build, using the finite element simulation software in the external elastic membrane sensor load stress analysis, and combined with the path integral method to establish the simulation model of the sensor output voltage, according to the simulation results, analysis the main structure size and performance parameters which will affect the performance of sensor; Analysis of all possible uncertainties factor which may be arise in the process of design or manufacture, to determine the key parameters which will affect the main performance of the sensor under different conditions, select the appropriate robust design method to establish robust design model, and to design the parameters of sensor.In this paper, Taguchi robust design methods and robust design method based on stochastic model to be used, in the Taguchi robust design, to determine the level of controllable factors and noise factors to determine the orthogonal test program based on orthogonal table, obtained by simulation of various factors impact on target performance parameters of its design analysis of variance to obtain the best combination of parameters when the quality characteristics. In the robust design of the stochastic model to determine the design parameters, the noise factor, quality characteristics, constraints, a mathematical model based on stochastic model of robust design, and programming calculations, obtain robust design parameters. In the robust design which based on stochastic model, to determine the design parameters, noise factor, quality characteristics and the constraint conditions, and to build the mathematical model using robust design method which based on stochastic model, and programming calculation, to obtain the robust design parameters.Comparison the results of Taguchi robust design method and general optimization, can be seen in the parameters of the new robust design, quality characteristics of sensor is superior to the general optimization. In the robust design base on Stochastic model not only consider the problem of tolerance of the design variables, but also fully consider noise factor which will impact the performance of the sensor, through build the stochastic model, and calculated using stochastic simulation program, to obtain the parameters with tolerance, this is not possible with other methods of design, the method can make the sensor has robust performance when variation occur in manufacturing, using and nature of the materials, so it help improve the quality of the sensor.