A search and score strategy-based fiber-optic gyroscope zero bias estimation method

Abstract

The application discloses a fiber-optic gyroscope zero offset estimation method based on a search and scoring strategy, and comprises the following steps: S1, collecting original angular rate output data of high-precision fiber-optic inertial navigation under quasi-static conditions, integrating the angular rate output data of the gyroscope to obtain a quasi-static period attitude angle sequence, calculating the mean value of the attitude angle, and further obtaining an attitude deviation sequence; S2, determining a search space and a search step of the gyroscope zero offset; S3, generating all candidate zero offset values, which are used for correcting the attitude deviation sequence to obtain a compensated attitude deviation sequence corresponding to each candidate zero offset value; S4, calculating a score value of the compensated attitude deviation sequence corresponding to each candidate zero offset value; and S5, based on the score value corresponding to each candidate zero offset value obtained in S4, taking the candidate zero offset value corresponding to the minimum score value as the optimal zero offset of the gyroscope. The method can accurately estimate the gyroscope zero offset of high-precision fiber-optic inertial navigation.

Description

Technical Field

[0001] This invention relates to the field of precision measurement technology for rail transit, specifically to a method for zero-bias estimation of fiber optic gyroscopes based on a search and scoring strategy. Background Technology

[0002] When high-precision fiber optic inertial navigation systems (INS) undergo initial alignment or zero-bias calibration under static or quasi-static conditions, the INS attitude angle is obtained by integrating the gyroscope's output angular rate. During this process, the main error sources for the gyroscope are zero bias and angular random walk. Ideally, the attitude angle of a stationary base is constant, and the attitude change after integration should primarily be caused by zero bias. Therefore, the gyroscope zero bias can be estimated using methods such as linear fitting of the attitude sequence.

[0003] However, in real-world applications (such as mounting the inertial navigation system on a railway rail inspection platform), the equipment is not absolutely stationary. Operator touches, the control screen, automatic rotation of the total station, and environmental factors such as wind and mechanical vibration can all cause random, irregular, short-lived, and variable attitude disturbances. Figure 1 As shown, these perturbations cause "mutations" or "drifts" in the attitude sequence due to non-zero bias, and the timing, amplitude, and duration of these changes are unpredictable.

[0004] If linear fitting or averaging is directly applied to the attitude data throughout the entire time period, these interfering data will severely pollute the zero-bias estimation results, causing the estimated values ​​to deviate from the true zero bias, thereby affecting the subsequent navigation accuracy.

[0005] Traditional methods typically detect and remove data from periods of interference by setting thresholds (such as angular velocity thresholds or attitude change rate thresholds). However, this method has the following significant drawbacks:

[0006] 1) The threshold is difficult to set adaptively; if it is too strict, effective data will be lost, and if it is too wide, interference cannot be filtered out.

[0007] 2) Poor detection performance for irregular, non-pulse-type slow disturbances (such as slow tilting caused by prolonged pressing);

[0008] 3) The robustness of the method depends on the scenario and experience in parameter tuning. Summary of the Invention

[0009] To address the problems existing in the prior art, this invention provides a method for estimating the zero bias of fiber optic gyroscopes based on a search and scoring strategy, which can accurately estimate the zero bias of high-precision fiber optic inertial gyroscopes.

[0010] Therefore, the present invention adopts the following technical solution:

[0011] A method for zero-bias estimation of fiber optic gyroscopes based on a search and scoring strategy is described below:

[0012] S1. Acquire the raw angular rate output data of the gyroscope under quasi-static conditions using a high-precision fiber optic inertial navigation system; integrate the raw angular rate output data of the gyroscope to obtain the attitude angle sequence of the quasi-static time period; calculate the mean value of the attitude angles in the attitude angle sequence of the quasi-static time period, and then calculate the attitude deviation value of each attitude angle relative to the mean value to obtain the attitude deviation sequence.

[0013] S2, determine the search range and search step size for zero bias of the gyroscope;

[0014] S3, generate all candidate zero bias values ​​within the search interval according to the search step size, and correct the attitude deviation sequence obtained in S1 based on each candidate zero bias value to form a compensated attitude deviation sequence corresponding to each candidate zero bias value.

[0015] S4, calculate the score value of the compensated attitude deviation sequence corresponding to each candidate zero bias value obtained in S3;

[0016] S5. Based on the score value corresponding to each candidate zero bias value obtained in S4, the candidate zero bias value corresponding to the minimum score value is taken as the optimal zero bias of the gyroscope.

[0017] In step S1 above:

[0018] The time interval for collecting data is ;

[0019] The quasi-static time-period attitude angle sequence is as follows: ,in, The total length of the attitude angle sequence during the quasi-static time period. The attitude angle obtained by integrating at time step is ;

[0020] The mean of the attitude angles is , ;

[0021] attitude angle attitude deviation value relative to the mean for:

[0022] ,

[0023] The attitude deviation sequence is .

[0024] In step S1 above: the time interval Consistent with the inertial navigation sampling rate.

[0025] In step S2 above: based on the technical specifications of the gyroscope used, a search interval of 5 to 10 times the technical specifications is selected. The search step size is 1 / 20 to 1 / 10 of the aforementioned technical indicators. .

[0026] The specific operation of step S3 above is as follows:

[0027] According to the search step size Generate the search interval All candidate zero bias values, total The first one, record the number The candidate zero bias value is ,in, ;

[0028] Based on candidate zero bias Correct the attitude deviation sequence obtained in step S1 For each attitude deviation value, obtain the corresponding candidate zero bias value. The corresponding compensated attitude deviation sequence:

[0029] ,

[0030] in, Attitude deviation value after time compensation .

[0031] The specific operation of step S4 above is as follows: calculate the absolute value of each attitude deviation value in the compensated attitude deviation sequence to obtain... Then, sort them in ascending order to form an absolute value attitude deviation sequence; based on prior knowledge, set a percentage threshold. Select the first absolute value attitude deviation sequence The values ​​are calculated, and their mean is used as the current candidate zero bias value. The score is denoted as Resulting in a sequence of score values .

[0032] In step S4 above: the percentage threshold This represents the proportion of data with little or no interference, and is adjusted within the range of 70% to 90%.

[0033] The specific operation of step S5 above is as follows:

[0034] The score sequence is obtained based on S4. Find the minimum value among the score values. , The corresponding candidate zero bias value is the optimal zero bias of the gyroscope. .

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. Strong robustness: This invention can automatically suppress the influence of irregular, unknown-amplitude short-term or long-term interference (such as personnel operation, equipment rotation, environmental vibration, etc.) on the zero bias estimation without setting any detection threshold.

[0037] 2. High precision: Through fine-grained search and a statistically based scoring strategy, this invention can accurately approximate the true zero bias value even in the presence of interfering data.

[0038] 3. Simple and reliable: The core of this invention is search and ranking scoring, with clear logic, controllable computational complexity, and easy engineering implementation.

[0039] 4. Wide adaptability: This invention is not sensitive to the type, amplitude, or duration of interference, as long as the proportion of interference in the total duration does not exceed a preset limit. ), and it will work effectively.

[0040] 5. Make full use of data: This invention avoids the problem that the threshold method may mistakenly remove valid data or miss removing interfering data. It smoothly distinguishes between "high-probability static data" and "suspected interfering data" by sorting by percentage. Attached Figure Description

[0041] Figure 1 A schematic diagram of the attitude angle sequence during a quasi-static time period subject to external disturbances;

[0042] Figure 2 This is a flowchart of the gyroscope zero bias estimation method of the present invention;

[0043] Figure 3 This is a sequence diagram of the entire attitude angle obtained by integration in an embodiment of the present invention;

[0044] Figure 4 for Figure 3 Enlarged view of the attitude angle sequence during the quasi-static period;

[0045] Figure 5 To Figure 4 A comparison of attitude angle sequences during the mid-quasi-static period after using different candidate zero-bias compensation. Detailed Implementation

[0046] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and embodiments. Obviously, the following embodiments are only some embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] like Figure 2 As shown, this invention proposes a method for estimating the zero bias of a fiber optic gyroscope based on a search and scoring strategy. The specific steps are as follows:

[0048] S1, Data Acquisition and Preprocessing.

[0049] Acquire high-precision fiber optic inertial navigation system raw gyroscope angular rate output data (e.g., heading angle) under quasi-static conditions, with a sampling time interval of [missing information]. (Typically 0.005s, consistent with the inertial navigation sampling rate), integrate the original angular rate output data of the gyroscope to obtain the quasi-static time period attitude angle sequence. The total length of the quasi-static time period attitude angle sequence is The quasi-static time period attitude angle sequence records The attitude angle obtained by integrating at time step is .

[0050] Calculate the mean of attitude angles in the quasi-static time period attitude angle sequence. :

[0051] ,

[0052] calculate Attitude deviation of the attitude angle relative to the mean at any given time :

[0053] ,

[0054] Obtain the attitude deviation sequence .

[0055] S2 determines the search range and search step size for zero bias of the gyroscope.

[0056] Based on the technical specifications of the gyroscope used (such as nominal zero-bias stability), a search range of 5 to 10 times the specified technical specifications is used. The search step size is 1 / 20 to 1 / 10 of the aforementioned technical indicators. For example, for a nominal zero-bias stability of... The spinning top, setting the search range Set the search step size .

[0057] S3, candidate gyroscope zero bias compensation.

[0058] According to the search step size Generate the search interval All candidate zero bias values, total The first one, record the number The candidate zero bias value is ,in, .

[0059] Based on candidate zero bias Correct the attitude deviation sequence obtained in step S1 For each attitude deviation value, obtain the corresponding candidate zero bias value. The corresponding compensated attitude deviation sequence:

[0060] ,

[0061] in, Attitude deviation value after time compensation .

[0062] S4, Calculate the score: Calculate the zero bias value obtained in S3 for each candidate. Corresponding compensated attitude deviation sequence The scoring value is determined as follows:

[0063] Calculate the absolute value of each attitude deviation value in the compensated attitude deviation sequence to obtain... Then sort them in ascending order to form an absolute value attitude deviation sequence.

[0064] Based on prior knowledge, a percentage threshold is set. (For example, 80%) This threshold represents the approximate proportion of data with no or very little interference. Parameter It can be flexibly adjusted based on the prior knowledge of "static purity" in a specific scenario, for example, 70%~90%.

[0065] Select the first absolute value attitude deviation sequence The values ​​are calculated, and their mean is used as the current candidate zero bias value. The score is denoted as .

[0066] The smaller the score, the smaller the attitude fluctuation over most of the time period after compensation with the candidate zero bias value; that is, the closer the candidate zero bias value is to the true value, and the better the compensation effect. Interference data (regardless of size) will likely fall into the lower rank after sorting due to their large deviation from the mean. The part that is not explicitly detected or removed is naturally excluded from the scoring calculation.

[0067] The scoring value sequence is obtained through the above operations. .

[0068] S5, determine the optimal gyroscope zero bias.

[0069] The score sequence obtained based on S4 Find the minimum score value , The corresponding candidate zero bias value is the optimal zero bias of the gyroscope. .

[0070] Example

[0071] Scenario: A gyroscope has a nominal zero-bias stability index of... The fiber optic inertial navigation system was installed on the railway rail inspection trolley platform, with the trolley stationary. Static data acquisition was conducted for 30 minutes. During this period, the operator made 5 screen clicks and 2 equipment adjustments, and the total station automatically rotated 3 times, despite occasional wind.

[0072] The method for estimating gyroscope bias includes the following steps:

[0073] S1, Data Acquisition and Preprocessing.

[0074] Record the angular rate output of the gyroscope's Z-axis (azimuth axis) at sampling intervals. The quasi-static time-period attitude angle sequence is obtained through integration. The total length of the quasi-static time period attitude angle sequence is [missing information]. , Attitude angle obtained by time-integration .

[0075] Calculate the average value of the attitude angles in the quasi-static time period attitude angle sequence: .

[0076] calculate Attitude deviation of the attitude angle relative to the mean at any given time: .

[0077] Obtain the attitude deviation sequence .

[0078] S2 determines the search range and search step size for zero bias of the gyroscope.

[0079] Search range: ,

[0080] Search step size: .

[0081] S3, candidate gyroscope zero bias compensation.

[0082] Generate all candidate zero bias values ​​within the search interval according to the search step size. Each candidate zero bias value. Correct the attitude deviation sequence obtained in step S1 This forms a zero bias with each candidate. Corresponding compensated attitude deviation sequence ,in, Attitude deviation value after time compensation .

[0083] S4, Calculate the score.

[0084] In this embodiment, a percentage threshold is set. Calculate the absolute value of each attitude deviation value in the compensated attitude deviation sequence to obtain... They are sorted in ascending order to form an absolute value attitude deviation sequence.

[0085] The candidate zero bias is obtained by averaging the first 80% of the absolute value attitude deviation sequence. Corresponding rating value .

[0086] This leads to the score sequence. .

[0087] S5, determine the optimal gyroscope zero bias.

[0088] The score sequence obtained based on S4 Find the minimum score , The corresponding candidate zero bias value is the optimal zero bias of the gyroscope. .

[0089] In this embodiment, Corresponding candidate zero bias value If the zero bias is 0.012° / h, then the optimal estimated gyroscope bias is 0.012° / h.

[0090] Figure 3 This embodiment demonstrates the attitude angle changes in the entire attitude angle sequence obtained through integration. Figure 4 Enlarged display Figure 3 The attitude angle changes during the quasi-static period. Figure 5 This embodiment demonstrates three candidate zero bias values ​​and the optimal zero bias value pair. Figure 3 The image shows the attitude angle changes after compensation for the quasi-static time period attitude angle sequence. It can be seen that after compensation using the gyroscope zero bias value estimated by the method of this invention, the attitude angle drift during the quasi-static time period is significantly reduced, and long-term stability is improved.

Claims

1. A method for estimating the zero bias of a fiber optic gyroscope based on a search and scoring strategy, characterized in that, The specific steps are as follows: S1. Acquire the raw angular rate output data of the gyroscope under quasi-static conditions using a high-precision fiber optic inertial navigation system; integrate the raw angular rate output data of the gyroscope to obtain the attitude angle sequence of the quasi-static time period; calculate the mean value of the attitude angles in the attitude angle sequence of the quasi-static time period, and then calculate the attitude deviation value of each attitude angle relative to the mean value to obtain the attitude deviation sequence. S2, determine the search range and search step size for zero bias of the gyroscope; S3, generate all candidate zero bias values ​​within the search interval according to the search step size, and correct the attitude deviation sequence obtained in S1 based on each candidate zero bias value to form a compensated attitude deviation sequence corresponding to each candidate zero bias value. S4, calculate the score value of the compensated attitude deviation sequence corresponding to each candidate zero bias value obtained in S3; S5. Based on the score value corresponding to each candidate zero bias value obtained in S4, the candidate zero bias value corresponding to the minimum score value is taken as the optimal zero bias of the gyroscope.

2. The fiber optic gyroscope zero-bias estimation method according to claim 1, characterized in that, In step S1: The time interval for collecting data is ; The quasi-static time-period attitude angle sequence is as follows: ,in, The total length of the attitude angle sequence during the quasi-static time period. The attitude angle obtained by integrating at time step is ; The mean of the attitude angles is , ; attitude angle attitude deviation value relative to the mean for: ， The attitude deviation sequence is .

3. The fiber optic gyroscope zero-bias estimation method according to claim 2, characterized in that, In step S1: the time interval Consistent with the inertial navigation sampling rate.

4. The fiber optic gyroscope zero-bias estimation method according to claim 1, characterized in that, In step S2: Based on the technical specifications of the gyroscope used, a search range of 5 to 10 times the specified technical specifications is established. The search step size is 1 / 20 to 1 / 10 of the aforementioned technical indicators. .

5. The fiber optic gyroscope zero-bias estimation method according to claim 4, characterized in that, The specific operation of step S3 is as follows: According to the search step size Generate the search interval All candidate zero bias values, total The first one, record the number The candidate zero bias value is ,in, ; Based on candidate zero bias Correct the attitude deviation sequence obtained in step S1 For each attitude deviation value, obtain the corresponding candidate zero bias value. The corresponding compensated attitude deviation sequence: ， in, Attitude deviation value after time compensation .

6. The fiber optic gyroscope zero-bias estimation method according to claim 5, characterized in that, The specific operation of step S4 is as follows: calculate the absolute value of each attitude deviation value in the compensated attitude deviation sequence to obtain... Then, sort them in ascending order to form an absolute value attitude deviation sequence; based on prior knowledge, set a percentage threshold. ; Select the first absolute value attitude deviation sequence The values ​​are calculated, and their mean is used as the current candidate zero bias value. The score is denoted as Resulting in a sequence of score values .

7. The fiber optic gyroscope zero-bias estimation method according to claim 6, characterized in that, In step S4: the percentage threshold This represents the proportion of data with little or no interference, and is adjusted within the range of 70% to 90%.

8. The fiber optic gyroscope zero-bias estimation method according to claim 6, characterized in that, The specific operation of step S5 is as follows: The score sequence is obtained based on S4. Find the minimum value among the score values. , The corresponding candidate zero bias value is the optimal zero bias of the gyroscope. .

Images