Measurements And Applications Of Micro-machined Accelerometer

Accelerometer is one of the main inertial components in inertial measuring and navigation systems. As the microel etronic processing technology permeated into inertial instruments, micromachined accelerometer has come into existence in this field and attracted much attention. As a new style of inertial sensor, the micromachined accelerometer has many advantages over the traditional accelerometer , for example, in bulk, weight, cost, power consumption, rel iabi li ty and. length of life. The model ADXL05 mieroSi accelerometer, a kind of products of AD Corporation of USA, is a right sample. This paper introduced its structure design and operating theory, measured its important performance data, and presented the measured results and necessary mathematical analysis. Using the testing data, the mathematical model is also given.Comparing with the dominant mechanical inertia] sensor, the great disadvantage of mieroSi sensor is low precision. In order to solve the problem of nonlinearity and other errors brought by environment reasons, some compensative and corrective measures should be taken. Because of the ability of nonlinear approximation, self-adaptation and learning, two methods based artificial neural networks are presented to realize the nonlinear correction. Each of them has certain feature and suits for different conditions. The result of simulation shows that artificial neural networks are nimbler in use, simpler in algorithm and more practical. In fact, the flexibility of these means automatically compensates any variation of the sensor response occurring due to change in environmental conditions.In the end, according to the idea of inertial sensor group that comprises several same kind micromachined sensors, the fusion estimated algorithm of multisensor data is presented in this paper.It is theoretically proved that this algorithm has least mean square errors. As a result, more precise and reliable data can be obtained from the integrated output. The simulation results also show the effectiveness of the algorithm to practical application.