Error Modeling And Analysis Of MEMS Gyroscope In The Two-axis Platform Stabilization System

Platform stabilization system can isolate the disturbances of the movement vectorto the detection system, real-time monitor the posture and the position changes of theplatform, and also supply a reverse torque to compensate the interference from outside,accurately keep the benchmark of the detector dynamic posture unchanged, finally,realize monitoring the fixed object or auto-track the moving target through the photodetector of the platform. Gyroscopes are the core sensitive components of theplatform stabilization system. MEMS gyroscopes are widely used in military andcivilian fields for their characteristics, such as lighter weight, smaller size, lower cost,lower power and consumption, higher reliability. However, due to the peculiarities ofthe processing materials and limitations of the processing technology level, themeasurement accuracy level of MEMS gyroscope is still low in the world today. It isthe key to miniaturize the platform stabilization system that improving themeasurement accuracy of the MEMS gyroscope.A signal sample system of MEMS gyroscope, which using in two-axis platformstabilization system of micro-optical pod, was designed. In this system, we designed agyroscope zero adjustment circuit, which based on digital potentiometer. The zerocenter of the gyro can be adjusted through the output of the digital potentiometer,which was controlled by DSP.The change of environment temperature is one of the most important reasonswhich affect the measurement accuracy of MEMS gyroscope. Due to the miniaturizationof the size and the internal structure of MEMS gyroscope, its performance is very sensitive tothe environment temperature changes. Among the performances, the gyroscope zerooutput and the Scale factor are greatest affected by temperature changes. Therefore,we designed a full temperature test experiment to get the characteristics of gyroscope,and also designed an experiment, which was simulated by the power source, about therelationship between gyroscope zero output changes and the adjustment of the digitalpotentiometer. Through the research of the two experiments, we got a gyroscope zerocompensation model, also realized the adjustment of zero center by controlling thedigital potentiometer. The temperature compensation experiments showed that usingthe method of full temperature test and the compensation of digital potentiometer, the output stability of gyroscope was greatly improved, and also the method had somepractical value. We designed the speed calibration experiments in differenttemperatures, using three method to fit the scale factor, in the end, the optimal speedsegment fitting method of gyroscope scale factor was reached.The gyroscope random drift error is one of the main sources of the measurementerror. Here we used time series-analysis to build the ARMA model for the randomdrift data of MEMS gyroscope. AR(1) model was the best model, which wasdetermined by the least AIC criterion. Using Kalman filter to filter the gyroscoperandom drift data on the base of the model built before. In the end, the gyroscoperandom errors before and after filtering were evaluated by Allan variance analysis.The results showed that Kalman filter method was effective and feasibility in filteringof the gyroscope random drift error.