Short-time multi-time efficient four-position north-seeking method fusing MEMS gyroscope noise characteristics

Abstract

The invention discloses a short-time multi-time efficient four-position north-seeking method fusing MEMS gyroscope noise characteristics, which comprises the following steps: taking corresponding time t0 when the noise of a gyroscope north seeker is mainly reflected as white noise as standing time of four-position north-seeking, and solving a group of azimuth angles by using MEMS gyroscope output data at four positions; and a final north-seeking result is obtained by adopting a mode of multiple north-seeking and averaging, so that the influence of the white noise of the gyroscope on north-seeking errors is reduced. According to the method, the Allan variance is reduced to times of the original Allan variance, so that the north-seeking precision is effectively improved while the influence of the low-frequency error of the gyroscope on the north-seeking result is inhibited.

Description

Technical Field

[0001] The present invention belongs to the field of MEMS gyroscope north finding, and specifically relates to a "short-time multiple times" efficient four-position north finding method integrating the noise characteristics of a MEMS gyroscope. Background Art

[0002] A MEMS gyroscope is a chip-based angular velocity sensor manufactured using micro-electromechanical technology. It offers advantages such as low cost, compact size, and low power consumption. With the continuous improvement of its performance, high-precision MEMS gyroscopes can achieve north-finding and orientation by sensing the horizontal component of the Earth's rotational angular velocity. Currently, north-finding devices based on MEMS gyroscopes mostly use a two-position or four-position method to find north. In theory, multi-position differential analysis can suppress the gyroscope's temperature drift and obtain azimuth information.

[0003] In the actual north-seeking process, the MEMS gyroscope's numerous noise sources and temperature drift errors restrict its north-seeking accuracy. The gyroscope noise characteristics can be characterized by its Allan standard deviation, see Figure 2 During time t1, the gyro noise is mainly manifested as white noise, which can be suppressed by extending the sampling time; between t1 and t2, the noise will mainly manifest as 1 / f noise, at which time the Allan standard deviation curve of the gyro output signal reaches its lowest; and after time t2, the Allan standard deviation of the gyro output signal manifests as short-term drift, at which time the gyro error will be mainly affected by the temperature drift error.

[0004] When using a MEMS gyroscope for north-finding using a static north-finding method based on multi-position differentials, its north-finding accuracy is directly related to the performance of the MEMS gyroscope. Chinese Patent CN115993114A discloses a two-position smoothed precision north-finding method based on a MEMS gyroscope. This method improves north-finding accuracy by performing multiple north-finding processes and ultimately smoothing the multiple sets of north-finding results. However, this method does not specify the static sampling time for each position. Chinese Patent CN111765880A discloses a high-precision four-position north-finding method based on a single fiber optic gyroscope. By comparing the angular random walk of the fiber optic gyroscope at different sampling times, the static time corresponding to the minimum angular random walk is selected as the static time for the north-finding instrument during the final north-finding to ensure the best single-shot north-finding accuracy. However, the noise characteristics of fiber optic gyroscopes and MEMS gyroscopes are different. The north-finding accuracy of a MEMS gyroscope is affected not only by the angular random walk but also by temperature drift errors. For a MEMS gyroscope, static sampling at the time corresponding to the minimum angular random walk cannot achieve optimal north-finding accuracy. Summary of the Invention

[0005] The purpose of this invention is to provide a "short-time multiple times" efficient four-position north-finding method integrating the noise characteristics of MEMS gyroscopes, by analyzing the noise characteristics of the gyroscope output signal and selecting the appropriate static sampling time for each position, thus avoiding gyroscope temperature drift and Low-frequency errors such as noise affect the north-seeking accuracy of MEMS gyroscopes.

[0006] The technical solutions for achieving the purpose of the present invention are:

[0007] A "short-time, multiple-time" efficient four-position north-finding method that integrates the noise characteristics of MEMS gyroscopes is proposed. The time t0 corresponding to when the gyroscope north-finding instrument noise mainly manifests as white noise is used as the static time of the four-position north-finding. The output data of the MEMS gyroscopes at four positions are used to solve a set of azimuth angles. The final north-finding result is obtained by averaging multiple north-finding attempts to reduce the influence of gyroscope white noise on the north-finding error.

[0008] Compared with the prior art, the present invention has the following significant advantages:

[0009] Under the condition of certain performance of MEMS gyroscope, the present invention reduces temperature drift error and The impact of low-frequency errors such as noise on north-seeking accuracy; in addition, this north-seeking method effectively breaks through the bottleneck problem that limits the north-seeking accuracy due to the performance of MEMS gyroscopes. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 The figure is a flow chart of a four-position north-finding method integrating the noise characteristics of MEMS gyroscope.

[0011] Figure 2 This is a typical gyroscope Allan standard deviation curve.

[0012] Figure 3 Schematic diagram of the gyro Allan standard deviation curve.

[0013] Figure 4 This is a schematic diagram of the north-seeking results using the conventional four-position gyro north-seeking method.

[0014] Figure 5 This is a schematic diagram of the north-seeking results using the "short-time multiple" four-position gyro north-seeking method. DETAILED DESCRIPTION

[0015] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0016] Combine Figure 1 The present invention provides a "short-time multiple times" efficient four-position north-finding method integrating the noise characteristics of MEMS gyroscopes, comprising:

[0017] Place the MEMS gyro north finder stationary, collect the gyro output signal and perform Allan standard deviation analysis. The Allan standard deviation can be expressed as:

[0018]

[0019] Where N represents the angle random walk coefficient of the gyroscope, B represents the zero bias instability coefficient of the gyroscope, and K represents the speed random walk coefficient of the gyroscope. The three represent the gyroscope white noise, Noise and short-term drift level. From the above formula, we can see that the gyro white noise can be suppressed by extending the sampling time τ; Noise is a type of noise that comes from the gyro circuit and whose power spectrum density decreases with the increase of frequency f. This noise is not affected by the sampling time τ. However, the temperature drift error of the gyro increases with the extension of the sampling time τ. Figure 3 Conventional research believes that the ideal static sampling time for each position should be the time t1 corresponding to the minimum value of the gyro Allan standard deviation curve. At this time, the gyro measurement error is minimal, and theoretically, a higher north-seeking accuracy can be achieved. However, in the actual north-seeking process, the gyro is affected by temperature changes to different degrees in different environments, and the time t1 corresponding to the minimum value of the Allan standard deviation curve is different. This makes the gyro susceptible to temperature drift errors, and the gyro will be affected by temperature drift errors. The influence of errors such as noise.

[0020] To avoid temperature drift and Low-frequency errors such as noise affect north-seeking accuracy. This invention proposes selecting the time t0 corresponding to when the gyro noise is mainly white noise as the static sampling time for each position during north-seeking to suppress the impact of low-frequency errors on the gyro north-seeking accuracy. The selection process of t0 is as follows: static sampling is performed on the MEMS gyro north-seeker, and the Allan standard deviation curve of the gyro is drawn based on the obtained gyro angular velocity output and the static sampling rate. Figure 3 ; Take the slope of the Allan standard deviation curve as The time t0 corresponding to the farthest intersection of the characteristic curve and the Allan standard deviation curve is the static sampling time of the gyroscope, which ensures that the north-seeking accuracy is not affected by the low-frequency noise of the gyroscope while making the single north-seeking accuracy the highest.

[0021] The MEMS gyro north finder is kept at a preset initial zero position for a set sampling time t0, and the gyro angular velocity output ω1 is stored; the MEMS gyro north finder is rotated 90° clockwise (counterclockwise) from the initial zero position to the second position using a rotating mechanism, and the gyro angular velocity output ω2 is continued to be collected for a time period t0; then the MEMS gyro north finder is rotated 90° clockwise (counterclockwise) from the second position to the third position using the rotating mechanism again, and the gyro angular velocity output ω3 is continued to be collected for a time period t0; finally, the MEMS gyro north finder is rotated 90° clockwise (counterclockwise) from the third position to the fourth position using the rotating mechanism, and the gyro angular velocity output ω4 is collected for a time period t0. Using the MEMS gyroscope data at the four positions, a set of azimuth angles is solved, and the calculation method is as follows:

[0022]

[0023] Based on the four-position method, the azimuth angle results will not be affected by errors such as geographic latitude, gyro bias, and scale factor. Repeat the above four-position north-finding steps to solve multiple sets of azimuth angle results and store them. Since the errors of multiple sets of azimuth angle results are mainly affected by gyro white noise, the multiple sets of azimuth angle results are summed and averaged to obtain the final north-finding result:

[0024]

[0025] Relative to the i-th azimuth result α i (i=1,...,n), the north-seeking error is reduced to Where σ is the single-shot north-seeking error, and n is the number of repeated north-seeking attempts. This shows that when the gyro is primarily affected by white noise, the number of north-seeking attempts, n, and the sampling time, τ, have equivalent effects on suppressing gyro measurement errors. Therefore, the "short-term, multiple-shot" north-seeking method can effectively improve the accuracy of MEMS gyro north-seekers.

[0026] When the conventional four-position north-finding method and the "short-time multiple" four-position north-finding method are used for gyro north-finding, the gyro output diagram is shown in Figure 4 and Figure 5 , Figure 4 The sampling time of the conventional four-position north-finding method is t1, and the total time of a single north-finding is 4t1. Figure 5 The sampling time for the "short-time, multiple-pass" four-position north-finding method is t0, and the total duration of a single north-finding attempt is 4t0. In the figure, t1 and t0 satisfy the relationship t1 = nt0, indicating the time it takes to complete one north-finding attempt using the conventional four-position north-finding method. The "short-time, multiple-pass" four-position north-finding method can complete n north-finding attempts while maintaining the same total north-finding time. The "short-time, multiple-pass" four-position north-finding method can reduce the north-finding error by smoothing the multiple north-finding attempts.

[0027] In summary, the present invention reduces temperature drift and The impact of low-frequency errors such as noise on north-seeking accuracy; in addition, this north-seeking method effectively breaks through the bottleneck problem that limits the north-seeking accuracy due to the performance of MEMS gyroscopes.

[0028] Through the description and drawings, the present invention provides typical embodiments of specific structures of specific embodiments. Based on the spirit of the present invention, other transformations can be made. Although the above invention provides the existing preferred embodiments, these contents are not intended to be limiting.

Claims

1. A "short-time, multiple-time" efficient four-position north-finding method integrating the noise characteristics of MEMS gyroscopes, characterized by: The time t0 corresponding to when the gyro north finder noise mainly manifests as white noise is used as the static time of four-position north seeking. The output data of the MEMS gyroscopes at four positions are used to solve a set of azimuth angles. The final north seeking result is obtained by taking the average of multiple north seeking attempts to reduce the influence of gyro white noise on the north seeking error.

2. The "short-time multiple times" efficient four-position north-finding method integrating MEMS gyroscope noise characteristics according to claim 1 is characterized in that: The selection process of time t0 is as follows: static sampling is performed on the MEMS gyro north finder, and the Allan standard deviation curve of the gyro is drawn according to the obtained gyro angular velocity output and the static sampling rate; the slope of the Allan standard deviation curve is selected. The time t0 corresponding to the farthest intersection of the characteristic curve and the Allan standard deviation curve is the static sampling time of the gyroscope.

3. The "short-time multiple times" efficient four-position north-finding method integrating MEMS gyroscope noise characteristics according to claim 1 or 2, characterized in that: The process of finding north in four positions is as follows: The MEMS gyro north finder is kept at a preset initial zero position for a set sampling time t0, and the gyro angular velocity output ω1 is stored; the MEMS gyro north finder is rotated 90° from the initial zero position to a second position, and the gyro angular velocity output ω2 is continued to be collected for a time period t0; then the MEMS gyro north finder is rotated 90° from the second position to a third position, and the gyro angular velocity output ω3 is continued to be collected for a time period t0; finally, the MEMS gyro north finder is rotated 90° from the third position to a fourth position, and the gyro angular velocity output ω4 is continued to be collected for a time period t0; a set of azimuth angles are solved using the MEMS gyroscope data at the four positions.

4. The "short-time multiple times" efficient four-position north-finding method integrating the MEMS gyroscope noise characteristics according to claim 3 is characterized in that: The azimuth angle is calculated as follows: Where α1 is the first set of azimuth angles to be solved.

5. The "short-time multiple times" efficient four-position north-finding method integrating the MEMS gyroscope noise characteristics according to claim 4 is characterized in that: The final north-seeking result is: where α i is the i-th group of azimuths to be solved, and n is the number of repeated north-seeking.

Images