| Silicon micromachined gyroscopes has important application value in civil and military fields due to their small size,light weight and low cost.With the improvement of micromachining technology and the optimization of control circuit,it is possible for micromachined gyroscopes to replace expensive large-scale high-end gyroscopes in some fields.The disk resonator gyroscope(DRG)with axisymmetric structure has the advantages of good structural symmetry,high energy transfer efficiency,good vibration resistance and so on.It has become an important candidate for high-performance micromachined gyroscope.The bias error of gyroscope reflects the core performance parameters such as detection resolution and long-term stability,which determines the detection accuracy of gyroscope.In this thesis,the Cobweb-like DRG(CDRG)designed by our group is taken as the research object to improve the bias stability and detection resolution of gyroscope.The error factors affecting the bias output are analyzed,and the methods to solve these problems in the measurement and control technology are given.The main research work and innovations are as follows:(1)This thesis analyzes the error factors that affect the gyroscope detection output,and provides ideas for further improving the performance of gyroscope.This thesis analyzes the error factors from the aspects of structural asymmetry,structural electric feedthrough,control circuit and working environment changes,which provides some breakthrough points for solving these problems.It mainly includes frequency splitting caused by stiffness asymmetry,quadrature and in-phase interference caused by vibration shape deflection,signal feedthrough introduced by coupling capacitance between electrodes,phase shift and noise in control circuit,frequency and Q value change caused by environmental temperature,etc.(2)The force to rebalance(FTR)closed-loop detection system is constructed,and the transfer function under the FTR closed-loop is derived,which provides a theoretical basis for the analysis of the closed-loop characteristics.Starting from the transfer function,the expressions of scale factor and bandwidth in the closed-loop FTR are given,and compared with the open-loop detection.This thesis focuses on the analysis of the influence of the frequency and Q value of the harmonic oscillator on the detection output under the FTR closed-loop detection,and proves that the FTR closed-loop can well suppress the scale factor and bias drift error caused by the frequency splitting and Q value change,that is,it can improve the robustness of the detection output to the structural asymmetry and the environmental temperature change.(3)A phase-shift compensation method based on FTR closed-loop detection is proposed to reduce the influence of the quadrature error on the bias output.Phase is an important factor in resonator control.Therefore,in the modulation-demodulation FTR detection system,the influence of the phase shift of the circuit on the detection output is analyzed.It is proved that the sense mode detection output is related to the phase shift of the drive mode detection circuit when the drive and the sense mode detection circuit are the same.When the phase shift compensation is not accurate,the quadrature coupling will be introduced into the detection output,resulting in a large bias and drift with the temperature.On the basis of phase analysis,a circuit phase shift self-compensation system is proposed.The experimental results show that this method can accurately calculate and compensate the phase shift of the circuit,greatly reduce the influence of quadrature coupling on the detection output,and improve the performance indexes such as bias instability and bias temperature sensitivity.(4)A system noise model based on FTR closed-loop detection is proposed to reduce the influence of noise on the bias output.The system noise directly affects the accuracy of detection output.Therefore,in the modulation-demodulation FTR detection system,the influence of four kinds of noises on the detection output is analyzed,and the noise equivalent rotation rate expression of the detection output is given in the frequency domain.Different from the existing closed-loop noise model,this model takes into account the electric feed-through of noise at the carrier frequency in the feedback circuit when the electromechanical amplitude modulation(EAM)detection technology is used,and describes the transfer path of two kinds of noise in the feedback circuit.Based on the theoretical noise model,a method is proposed to improve the adverse effects of various noises on the detection output.The experimental results show that,for the adopted circuit system,when the feed-through noise is suppressed,the noise level of the detection output can be effectively reduced,and the performance indexes such as angle random walk(ARW)can be improved.Using FPGA to realize the closed-loop detection system of FTR,and adopting the above optimization scheme,the bias instability is reduced from 7.663°/hr to 0.617 °/hr,the ARW is reduced from 0.209°/?hr to 0.017°/?hr,and the bias temperature sensitivity in the range of 0?-70? is reduced from 0.055°/s/°C to 0.001 °/s/°C.The performance of bias output has been improved obviously,and the expected goal has been achieved,which lays a good foundation for the subsequent realization of high-precision measurement. |
PDF Full Text Request