Research On Compensation Technology About MIMU Temperature Drift Error

With the continuous application and innovation of MEMS technology, MEMS inertial devices such as accelerometers, gyroscopes and its combinations MIMU are widely used in military and many civilian fields due to its advantage of small size, low power consumption and light weight. However, since MEMS inertial device is made of the silicon material which is sensitive to the temperature, temperature variation of environment will have a greater impact on its measurement. That is, temperature drift error is one of the main factors affecting the accuracy of MEMS inertial measurement device, consequently, how to reduce this type of error has become an urgent problem. The method currently used mainly attributed to two types: hardware control and software compensation. This paper studied on the temperature drift error of MIMU by analysis, testing and modeling, eventually compensate its error. Special details are discussed as follows:Firstly, the composition structure,mechanical properties and working principle of silicon micro-machined vibratory gyroscope and capacitive micro-accelerometer were analysed. On this basis the temperature error mechanism is described, including the changes of its performance parameters such as resonant frequency, quality factor and mechanical sensitivity due to temperature changed, these changes ultimately affect the drift of its bias and scale factor, and these theory lay a foundation for temperature modeling.Then, MIMU calibration is the basis of error modeling, analysis of temperature drift error is proceeded on the premise of calibration. An error mathematical model of MIMU was established, and 24 parameters including the bias, scale factor and installation misalignment coefficient was determined by designing appropriate calibration tests. The relationship between acceleration measurement, angular rate measurement and output voltage was established, and supplied the theoretical basis for the subsequent analysis modeling and compensation of temperature error.Finally, the temperature characteristics of 24 parameters including the bias and scale factor, installation misalignment coefficient were analysed by temperature test, thus establishing the temperature error model of MIMU which compensate the output of MIMU by software method. The strategy of compensation is that establishing the compensation relationship among acceleration measurement, angular rate measurement, bias,scale factor,installation misalignment coefficient and output voltage. Experimental results show that the error compensation model can effectively reduce acceleration and angular velocity measurement error in variable temperature of MIMU.