Systems and methods for a polarization matched resonator fiber optic gyroscope |
|||||||
申请号 | EP15150865.2 | 申请日 | 2015-01-12 | 公开(公告)号 | EP2896935B1 | 公开(公告)日 | 2018-10-03 |
申请人 | Honeywell International Inc.; | 发明人 | Sanders, Glen A.; Srandjord, Lee K.; Qiu, Tiequn; Wu, Jianfeng; | ||||
摘要 | Systems and methods for a polarization matched resonator fiber optic gyroscope are provided. In one embodiment an RFOG comprises: a light source; a fiber optic ring resonator; a photodetector that outputs an electrical signal that varies as a function of optical intensity; and an input light polarization servo. A light beam from the servo is launched into the resonator ring in a first direction of circulation. The input polarization servo comprises a birefringence modulator that modulates a phase shift between two components of an input polarization state of the light beam at É m , the modulator is controlled to drive towards zero a 1 st harmonic of É m as measured in the electrical signal. The servo further comprises a tunable ½ waveplate that adjusts an amplitude of the two components of the input polarization state relative to each other. The tunable ½ waveplate is controlled to maximize a peak optical intensity as measured in the electrical signal. | ||||||
权利要求 | |||||||
说明书全文 | In commercial navigation application, there is a continuing need for developing low cost, small sized navigation grade gyroscopes. The resonator fiber optic gyro (RFOG) is one gyroscope technology that can satisfy those needs. In the RFOG, a ring resonator is formed using a fiber optic coil and couplers to couple light into and out of the ring resonator in clockwise (cw) and counterclockwise (ccw) directions. At least two input light waves are frequency-tuned to the resonances of the ring resonator in the cw and ccw directions respectively. After measuring the resonance frequencies in the two directions by tuning each input beam to them, the input beam frequencies are compared, and the difference is proportional to the rotation rate of the resonator coil. In an RFOG bias instability is manifested as an indicated rotation rate output even when the gyroscope is not rotating. Patent document number For the reasons stated above and for other reasons stated below which will become apparent to those skilled in the art upon reading and understanding the specification, there is a need in the art for improved systems and methods to address RFOG bias instability. The embodiments of the present invention provide methods and systems for polarization matched resonator fiber optic gyroscope and will be understood by reading and studying the following specification. In one embodiment, a polarization matched resonator fiber optic gyroscope (RFOG) comprises: a first light source; a fiber optic ring resonator coupled to the first light source; a first photodetector coupled to an output of the fiber optic ring resonator, where the first photodetector outputs an electrical signal that varies as a function of optical intensity measured by the first photodetector; and a first input light polarization servo; wherein the fiber optic ring resonator is coupled to the first light source via the first input light polarization servo such that a light beam from the first input light polarization servo is launched into the fiber optic resonator ring in a first direction of circulation; wherein the first input polarization servo comprises a birefringence modulator that modulates a phase shift between two components of the input polarization state of the light beam at a modulating frequency ωm, wherein the birefringence modulator is further controlled to drive towards zero a 1st harmonic of the modulating frequency One issue with achieving this matched condition is that there is polarization cross-coupling within the input beams paths to the resonator that keeps it from being in the exact desired polarization state at the input to the resonator. The existence of this cross-coupling in the input leads, as well as thermally induced fluctuations in the birefringence of the input leads, can cause the input light wave polarization state to fluctuate over time, causing the coupling of input light into the polarization state of the resonator to vary. The magnitude of the cross-coupling itself can vary over time. Furthermore, even if the input light state were completely stable at the entrance of the resonator, the polarization state inside the resonator at its input can vary over time, causing the overlap with the input polarization state to vary. The various embodiments described herein provide systems and methods for very accurately launching the input light into one polarization mode within the ring resonator reducing bias instability within the RFOG. As described in detail below, one or more of these embodiments use active feedback to constantly adjust the input polarization state of the cw and ccw inputs beams to track and match the desired polarization state of the ring resonator. For simplicity, only components related to the CW operation of RFOG 100 are discussed in detail below. However, as shown in As shown in The light within resonator ring 110 is already highly (but not perfectly) polarized so that the amplitude of the Y component may be defined by ε (where ε is ≪ 1) which is equal to the relative amplitude of the Y component as compared to the amplitude of the X component. Thus the resonator polarization state may be represented by the normalized matrix: The light 113 entering input light polarization servo 120 is also already polarized into a similar polarization state, but the polarization state of light beam 113 will be mismatched with respect to the resonator polarization state. That is, ε for the polarization state of light 113 is off by Δε, while ϕ is off by Δϕ. The polarization state of light 113 entering input light polarization servo 120 can therefore be characterized by the normalized matrix: Thus if input light polarization servo 120 is operated to drive both Δε and Δϕ of light beam 113 to zero, the polarization state of the launch light 116 will have the same polarization state as exists inside ring resonator 110. To address Δϕ, input light polarization servo 120 comprises a birefringence modulator 124 positioned within the path of the light beam 113 prior to the ring resonator 110. Birefringence modulator 124 modulates the phase difference between the X-component and the Y-component of the launch light 116 (which is referred to herein as the input polarization state). There are various means for implementing birefringence modulator 124. For example, in one embodiment birefringence modulator 124 comprises a lithium niobate electro-optic phase modulator positioned along a principal axis of the input fiber 117 (which may comprise a length of polarization maintaining fiber). In another embodiment, an electro-optic transparent ceramic material is used, such as those manufactured at Boston Applied Technologies Incorporated (BATI). In the example embodiment shown in More specifically, the output intensity variation from ring resonator 110 at the peak of the resonance is sensed at ωm. It would be appreciated that signal processing methods known in the art may be used to control the cw and ccw input wave frequencies to the resonances centers of the cw and ccw resonances, respectively. Optical intensity observed by photodetector 114 is converted into an electrical signal which is fed into the first demodulator 126. The first demodulator 126 is also referred to herein as the 1st harmonic demodulator, because it demodulates the electrical signal from photodetector at the modulation frequency ωm. The output of the first demodulator 126 is a DC voltage or digital value proportional to Δϕ. The output of the first demodulator drives an integrator 150. In one embodiment, integrator 150 comprises a digital accumulator. The integrator 150 output ramps up in response to a non-zero output of the first demodulator 126, which results, in steady state, in a DC voltage component Δø0 that is imparted to the birefringence modulator 124 to adjust Δϕ such that Δϕ=0. If the input polarization state misalignment angle Δϕ is equal to 0 originally, the output of the first demodulator is zero and the input to the integrator 150 is zero. Thus, the integrator 150 output is zero in steady state. If there is a change in Δϕ, such that Δϕ≠0, then there exists a phase difference between the X-component and the Y-component of the input polarization state of launch light 116 that is different than that of the resonator polarization state. Therefore, birefringence modulator 124 is driven to minimize changes in intensity (i.e., minimize Δø0) at the phase modulation frequency ωm so that a phase difference between the X and Y components is applied. In this case, when Aϕ≠0, in the detected output of the photodetector 114 at ωm will be temporarily non-zero, the demodulated output of demodulator 126 will be temporarily non-zero, and the integrator 150 output will adjust the value of Δø0 until steady state is reached. In steady state, the Δϕ=0, and therefore the detected signal at ωm will be zero, as well as the demodulated output of the first demodulator. When Δø0 is adjusted such that Δϕ is driven to zero, variation at ωm will vanish, indicating that the phase shift ϕ for the input polarization state of launch light 116 and the phase shift ϕ for the resonator polarization state match and Δϕ has been driven to zero or minimized. As mentioned above, input light polarization servo 120 is operated to drive both Δε and Δϕ of light beam 113 to zero. To ensure maximum matching of the input polarization state to the resonator polarization state, the relative amplitudes of the x-component and the y-component of the input polarization state may be adjusted to drive Δε towards zero. To minimize Δε, input light polarization servo 120 also comprises a tunable ½ waveplate 122 positioned within the path of the light beam 113 prior to the ring resonator 110. In the example embodiment shown in This driving voltage that controls tunable ½ waveplate is derived from a feedback signal responsive the peak optical intensity of light received at photodetector 114. Optical intensity observed by photodetector 114 is sampled and converted into an electrical signal which is fed into a sampling analog-to-digital converter 171 within the input light polarization servo 120. The output of photodetector 114 as a function of time will comprise a DC component with a sinusoidal ripple riding on the DC component due to the application of modulation frequency ωm by modulator 124. Input light polarization servo 120 is configured to capture the maximum optical intensity level observed during each cycle of modulation frequency ωm and average that maximum optical intensity level over time to arrive at what is referred to herein as the peak optical intensity. Averaging the measured maximum optical intensity levels over time provides the benefit of averaging out random noise sources that may cause spikes observed during individual cycles. More specifically, analog-to-digital converter (A/D) 171 generates optical intensity sample data that is stored in memory 172. Samples are captured and stored at a sufficient frequency to capture details such as the waveform peak of the electrical signal. For example, in one embodiment, samples of the electrical signal are obtained at a rate of 1000x per cycle of the modulation frequency ωm. Memory 172 stores sample data covering a sufficient number of cycles so that averaging the cycles of sample data stored in memory 172 effectively averages out random noise captured during individual cycles. For example, in some embodiments, memory 172 stores between 10 to 100 cycles of sample data. Alternately, in other embodiments, memory 172 may store an average of previously captured cycles and a weighted average is performed with data samples for the most recently captured cycle or cycles. The averaging of sample data stored in memory 172 is performed by processor 174 in order to arrive at the value of the peak optical intensity. Processor 172 further drives tunable ½ waveplate 122 via a waveplate controller 175 to maximize the peak optical intensity of light received at photodetector 114. That is, waveplate controller 175 produces the driving voltage that controls tunable ½ waveplate 122 based on a control signal from processor 172. Maximizing peak optical intensity of light received at photodetector 114 results in driving Δε to zero. In one embodiment, processor 172 maximizes peak optical intensity by observing the effects adjusting tunable ½ waveplate 122 has on the observed peak optical intensity. For example, if rotating the axis of tunable ½ waveplate 122 in one direction produces an increase in peak optical intensity, then peak optical intensity prior to the adjustment was not maximized. When a control point is found such that rotating the plane of linear polarization in either direction from that control point will produce a decrease in peak optical intensity, then peak optical intensity at that control point is maximized. In one embodiment the 1st feedback loop used for controlling Δϕ is operated at higher bandwidth than the 2nd feedback loop used for controlling Δε. Keeping the two loops separated in frequency space can be utilized to avoid conflicts between them. For example, in one embodiment where the 1st harmonic feedback loop operates with 100Hz bandwidth, the 2nd harmonic feedback loop operates with 1-10Hz since variations in phase error may be expected to occur at a greater frequency that variations in the relative amplitude error. As explained above, the two major variables in the input polarization state of launch light 116, the polarization components' relative phase and amplitude, are adjusted to match that of the resonator polarization state of ring resonator 110. By controlling these variables in the launch light to match the resonator eigenstate, maximum power is coupled into the ring resonator 110 for maximum signal to noise, bias instabilities associated with interference from excitation of the second eignestate of the resonator are minimized or eliminated, and intensity fluctuations inside the resonator that give rise to optical Kerr effect variations (another bias instability mechanism) may be minimized. The method begins at 210 with launching a polarized light beam into a fiber optic ring resonator. The polarized light may originate from a light source, such as a tunable laser that produce monochromatic light beams. The method continues to 220 with detecting an optical intensity from an output of the fiber optic ring resonator. Observing the signal from the output of the ring resonator is believed to provide the best representation of how well the input light polarization state is matched to the resonator's true polarization state. The method proceeds to 230 with modulating a phase shift between two components of the polarization state of the polarized light beam with a birefringence modulator at a modulation frequency of ωm. In one embodiment, the birefringence modulator is driven by a voltage having two components. The first component is characterized by the time varying function Δømcos(ωmt), where Δøm is a selected amplitude and ωm is the modulation frequency. The second voltage component is the DC offset voltage, Δø0 derived from a feedback signal derived from the optical intensity measurements obtained in block 220. Accordingly, the method proceeds to 240 with driving the birefringence modulator to minimize a 1st harmonic of the modulation frequency ωm as observed from the optical intensity of the output of the fiber optic ring resonator. As discussed above in The method proceeds to 250 with controlling a tunable ½ waveplate to adjust a relative amplitude of a first polarized component of the polarized light beam with respect to a second polarized component of the polarized light beam. To ensure maximum matching of the input polarization state to the resonator polarization state, the relative amplitudes of the x-component and the y-component of the input polarization state can be adjusted to drive Δε to zero. In the example embodiment shown in Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that any arrangement, which is calculated to achieve the same purpose, may be substituted for the specific embodiment shown. Therefore, it is manifestly intended that this invention be limited only by the claims. |