High-precision fiber-optic gyroscope temperature drift monitoring and error suppression system and method

Abstract

The invention relates to a high-precision fiber-optic gyroscope temperature drift monitoring and error suppression system and method. The system is characterized in that a reference gyroscope is responsible for noise output; the monitoring gyroscope is used for monitoring a zero offset mean value; the two gyroscopes are arranged on the same carrier plane, and sensitive axes are parallel; the rotary heating cover is arranged outside the two gyroscopes, consists of an upper hemisphere and a lower hemisphere, and is rotationally supported on the fixed carrier through a supporting point at the vertex of the two hemispheres, so that the heating cover freely rotates clockwise and anticlockwise around the normal of the plane of the carrier; the rotary heating cover is composed of ball belt heating sheets with the same width in the latitude direction, and a heating resistor is distributed on each heating sheet; the comparator is used for collecting output signals of the reference gyroscope and the monitoring gyroscope and carrying out differential operation; and the temperature compensation unit adopts a neural network training model to realize the determination and addressing of the conduction of the heating resistor, and outputs the heating temperature and the heating time to realize the rotary heating control of the rotary heating cover. According to the invention, temperature drift monitoring and error suppression of the high-precision fiber-optic gyroscope are well realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of fiber optic gyroscopes, and particularly relates to a high-precision fiber optic gyroscope temperature drift monitoring technology, an error suppression method and a system. Background Art

[0002] With the wide application of fiber optic gyroscopes, the demand for the accuracy of gyroscopes is increasing day by day, especially in some high-reliability and high-precision application scenarios. Since the fiber optic gyroscope itself uses a fiber optic loop with a relatively large spatial size, when the ambient temperature changes, there is a certain temperature field non-uniformity in the loop, resulting in thermal drift. The most significant drift is the gyro drift caused by the Shupe thermal effect, that is, the Shupe error. In recent years, the zero bias of fiber optic gyroscopes changing with temperature is the main reason restricting the further development of fiber optic gyroscopes towards high-precision applications. In order to effectively eliminate the above errors, in addition to improving the symmetry and precision of loop winding, an effective method is to eliminate the temperature field non-uniformity. The temperature field non-uniformity can be eliminated by using temperature control and a rotating heating cover rotation method to achieve temperature field uniformity through "stirring" the spatial thermal field. The advantage is that the principle is simple, but the disadvantage is that the mechanical execution device has a relatively large volume, there is a certain hysteresis in operation control, resulting in an ineliminable time delay, and a part of the temperature gradient still acts on the gyroscope to cause drift. It is necessary to suppress or eliminate the temperature drift caused by the above hysteresis.

[0003] Since high-precision fiber optic gyroscopes usually have a relatively large loop size (hundreds of millimeters in diameter) and a relatively long loop length, the success rate of precise winding is very low. The survival rate of qualified loop drift is usually less than 20%, especially the survival rate of ultra-high-precision loops is even lower. Therefore, in the case where a loop with good symmetry cannot be obtained, how to effectively utilize the high resolution of the loop and avoid the low temperature drift characteristics of the loop is particularly important for still being able to effectively research high-performance gyroscopes when the current process cannot be quickly broken through. Summary of the Invention

[0004] The present invention aims at the deficiencies of the prior art and provides a high-precision fiber optic gyroscope temperature drift monitoring and error suppression system and method.

[0005] One of the purposes of the present invention is achieved through the following technical solutions:

[0006] A high-precision fiber optic gyroscope temperature drift monitoring and error suppression system includes a reference gyroscope, a monitoring gyroscope, a rotating heating cover, a synchronous carrier, a comparator, and a temperature compensation unit;

[0007] The reference gyroscope, which uses a gyroscope with high precision and a large temperature drift coefficient, is responsible for noise output;

[0008] The monitoring gyroscope, which uses a gyroscope with low precision and a small temperature drift coefficient, is responsible for monitoring the zero bias mean value;

[0009] The monitoring gyroscope and the reference gyroscope are placed on the same carrier plane, and the sensitive axes of the two gyroscopes are arranged in parallel.

[0010] The synchronous carrier is used to install the reference gyroscope, the monitoring gyroscope and the rotating heating cover.

[0011] The rotating heating cover is arranged outside the reference gyroscope and the monitoring gyroscope. The rotating heating cover is spherical and consists of two upper and lower hemispheres. A support point is arranged at the vertex of each of the upper and lower hemispheres. The two support points are rotatably supported on the fixed carrier through bearings, so that the heating cover can rotate freely along the normal line of the carrier plane in the clockwise and counterclockwise directions. The rotating heating cover is composed of spherical zone heating sheets with equal width in the latitude direction, and heating resistors are arranged on each heating sheet.

[0012] The comparator is used to collect the output signals of the reference gyroscope and the monitoring gyroscope and perform differential operation.

[0013] The temperature compensation unit is used to determine and address the conduction of the heating resistor by using a neural network training model according to the differential operation structure of the comparator, and output the heating temperature and heating time to realize the rotation heating control of the rotating heating cover.

[0014] Moreover, the ratio of the accuracy of the monitoring gyroscope to that of the reference gyroscope is not less than 5; the ratio of the temperature sensitivity coefficient of the monitoring gyroscope to that of the reference gyroscope does not exceed 1 / 5.

[0015] The second object of the present invention is achieved by the following technical solutions:

[0016] A method for monitoring the temperature drift and suppressing the error of a high-precision fiber optic gyroscope, based on the above high-precision fiber optic gyroscope temperature drift monitoring and error suppression system, includes the following steps:

[0017] Step 1: The comparator is used to collect the outputs of the reference gyroscope and the monitoring gyroscope in real time and perform difference operation. When the ratio of the absolute value of the difference between the outputs of the two gyroscopes to the absolute value of the output drift value of the monitoring gyroscope is greater than the set zero-bias threshold for heating start, it is determined that the gyroscope drifts due to temperature excitation.

[0018] Step 2: When it is confirmed in Step 1 that the gyroscope drifts, the temperature compensation unit performs rotation heating control on the rotating heating cover; the rotating heating cover rotates at a fixed speed, and the rotation axis is perpendicular to the carrier plane; the rotating heating cover is heated by zone heating, and the heating sheets near the carrier plane of the gyroscope are heated cyclically. After the temperature of the zone heating reaches the set value, the rotating heating cover rotates periodically and continuously from side to side, and the rotation start-stop duty ratio is 1:1 until the temperature field is uniform.

[0019] Step 3: When the ratio of the absolute value of the output difference between the two gyroscopes to the absolute value of the output drift value of the monitoring gyroscope does not exceed the set bias threshold, the output combination of the two gyroscopes is used as the effective gyroscope output; the combination method is: the average value of the monitoring gyroscope output is used as the bias mean value, and the reference gyroscope noise output is used as the gyroscope noise;

[0020] Step 4: Output and record the gyroscope data obtained in Step 3.

[0021] Moreover, in Step 1, the expression of the set bias threshold k is:

[0022] k = |E 参考 - E 监控 | / |E 监控 | ≥ 5;

[0023] where E is the gyroscope output bias stability, E 误差 is the stability of the rotational speed difference between the reference gyroscope and the monitoring gyroscope output, and E 监控 is the stability of the monitoring gyroscope output rotational speed.

[0024] Moreover, in Step 2, the rotation heating cover is heated by constant temperature heating, and the heating temperature and heating time are generated by the temperature compensation unit using a neural network training model. The inputs of the neural network training model are the initial environmental temperature T0 and the gyroscope bias stability error E 误差 ; T out and t form an output matrix, which are the output heating temperature and heating time respectively. IW and LW are the neural network weight factors, as shown in the following formula:

[0025]

[0026] The advantages and positive effects of the present invention are:

[0027] The present invention proposes a complementary output method for two gyroscopes. One is used as the main output gyroscope - the reference gyroscope, which is responsible for noise output, and the other is used as the auxiliary gyroscope - the monitoring gyroscope, which is responsible for monitoring the bias mean value. The complementary output of the two gyroscopes corrects the temperature drift and the current noise change. According to the output difference between the two gyroscopes, the gyroscope drift caused by temperature excitation can be obtained in a timely manner, and the movement and heating of the rotation heating cover are controlled by the temperature compensation unit, which preferably realizes the high-precision fiber optic gyroscope temperature drift monitoring and error suppression, and has good controllability. Brief Description of the Drawings

[0028] Figure 1 is a schematic structural diagram of the gyroscope drift monitoring system of the present invention;

[0029] Figure 2 is a flowchart of the gyroscope drift monitoring and error suppression of the present invention. DETAILED DESCRIPTION

[0030] The structure of the present invention is further described below with reference to the accompanying drawings and by way of examples. It should be noted that the present examples are descriptive rather than restrictive.

[0031] The design idea of ​​the present invention is as follows: there is no shortage of high-temperature performance gyroscopes among medium-precision gyroscopes, and practical experience shows that the temperature performance of medium-precision rings is generally better than that of high-precision rings. The main reason is that the process of medium-precision fiber optic gyroscope rings is mature, and there are many samples. The more important reason is that the medium-precision rings have smaller ring sizes, limited number of layers and turns, and lower symmetry weight factors than high-precision rings. Therefore, rings with excellent temperature performance can be screened out from medium-precision gyroscopes. However, the core reason why medium-precision gyroscope rings cannot achieve high precision is that the ring size is not large enough, resulting in insufficient sensitivity, and this fundamental threshold of insufficient sensitivity cannot be crossed. This threshold is precisely what high-precision large-size rings are inherently equipped with, so even if the high-precision ring temperature performance is poor, it can still be competent for sensitivity to short-term noise, and two such gyroscopes with different performances are easy to obtain.

[0032] Based on the excellent temperature performance of the medium-precision gyro compared to the high-precision ring and the high sensitivity of the high-precision gyro compared to the medium-precision gyro, the present invention proposes a two-gyro complementary output monitoring system, one as the main output gyro-reference gyro, which is mainly responsible for noise output, and one as an auxiliary gyro-monitoring gyro, which is mainly responsible for monitoring the zero bias mean. The two gyros complementarily output to correct temperature drift and current noise changes. Gyro monitoring technology, that is, using a low-cost small gyro as a monitor, the gyro responds to temperature changes as environmental changes. Among them, the reference gyro temperature performance can be poor, but the gyro resolution is higher and the cost is higher. The monitoring gyro itself has excellent temperature performance, is insensitive to temperature, and has a low cost. In order to better play the mean value characteristics, both gyros are placed in a partitioned temperature-controlled rotating heating cover with good temperature performance, the monitoring gyro is in a vibration isolation working state, and the two gyros are on the same synchronous carrier. The monitoring gyro extracts relatively low-frequency signals, such as slow speed changes and medium- and long-term magnetic field changes. The reference gyro can effectively propose high-frequency disturbance information in the environment, such as vibration noise and magnetic field excitation.

[0033] In order to effectively improve the error change of the monitoring gyro, the compensation amount is obtained by multiplying the gyro output comparison amount by the compensation coefficient matrix. The compensation amount is multiplied by the ambient input temperature Tin and finally fed back to the rotating heating cover. The rotating heating cover sets the heating plate temperature according to the feedback network and changes the rotation speed at the same time to quickly offset the temperature shock. The compensation coefficient matrix can use a neural network system or a linear fixed-length coefficient matrix. The above compensation coefficient matrix is ​​obtained through multiple test training compensation.

[0034] The specific positional relationship of the monitoring system is as follows: The monitoring gyroscope and the reference gyroscope are placed on the same carrier plane, and the sensitive axes of the two gyroscopes are parallel. The rotating heating cover is spherical and consists of two hemispheres, the upper and the lower. There is a support point at the vertex of each of the upper and lower hemispheres, and it is rotatably supported on the fixed carrier through a ball bearing, enabling the heating cover to freely rotate clockwise and counterclockwise around the normal line of the carrier plane.

[0035] Before determining that large temperature fluctuations cause gyro drift, it is necessary to effectively identify that the zero-offset change of the monitoring gyroscope is due to temperature excitation rather than carrier movement. The basic principle is: In the case of carrier movement, both gyroscopes can sense it, and at this time, when the output rotation speeds of the two gyroscopes are subtracted, there is no additional drift.

[0036] After excluding the zero-offset change of the gyroscope caused by carrier movement, it is necessary to determine a zero-offset threshold criterion, and based on this threshold, the heating cover is started to achieve temperature equalization. Generally, the monitoring gyroscope will have a small temperature drift, and the reference gyroscope has a temperature drift sensitivity coefficient inversely proportional to its accuracy. Assume the accuracy of the monitoring gyroscope is σ 监控 = 0.001° / h, and the accuracy of the reference gyroscope is σ 参考 = 0.0002° / h, and the accuracy ratio of the two gyroscopes is 5. Generally, the sensitivity coefficient of the reference gyroscope and the monitoring gyroscope to temperature is negatively correlated with the accuracy ratio. Take an example to illustrate the problem. Assume that the zero-offset change of the reference gyroscope caused by the same ambient temperature change is 0.0003° / h, but the influence of the above temperature fluctuation on the monitoring gyroscope does not exceed 0.0003°

[0037] / h / 5 = 0.00006° / h. In this case, after deducting the influence of carrier movement, the zero-offset change of the combined gyroscope of the two gyroscopes does not exceed 0.00036° / h, and still can reach a relatively high accuracy. However, when the accuracy drift change of the reference gyroscope caused by external temperature change reaches 0.001° / h, at this time, the temperature drift of the monitoring gyroscope reaches 0.001° / h / 5 = 0.0002° / h. At this time, the worst-case combined accuracy of the two gyroscopes reaches 0.0012° / h, which can no longer meet the requirement of accuracy improvement. Therefore, small temperature disturbances will not affect the noise of the monitoring gyroscope in the short term, but when there is a relatively high-frequency temperature excitation, the monitoring gyroscope is excited to have a large temperature drift, which will seriously affect the gyroscope accuracy. In this case, it is necessary to equalize the ambient temperature to reduce the temperature drift error of the monitoring gyroscope. When the external temperature causes drift, the reference gyroscope will be more sensitive than the monitoring gyroscope, and an output deviation E will be generated after subtracting the outputs of the two gyroscopes 误差 = E 参考 - E 监控 , and the criterion for starting the heating cover for equalization can be preset as: k = E 误差 / E 监控, that is, when the above deviation threshold is exceeded, this temperature drift needs to be eliminated. The method of elimination is to adopt the method of rapid rotation and soaking to eliminate the uneven temperature field gradient. Based on the gyro monitoring invention, the present invention proposes to collect and monitor the change of the gyro zero bias and compare it with the change of the reference gyro zero bias. By comparing the size of the zero bias change and the threshold value, the rotation of the partitioned rotating heating cover is controlled for soaking, so as to achieve a drift suppression method for offsetting the external temperature change.

[0038] Based on the above content, the implementation steps of the high-precision fiber optic gyro temperature drift monitoring and error suppression method of the present invention include the following five steps:

[0039] (1): Fabricate a rotating frame and a rotating heating cover. Two hemispherical heating covers are fixed on the isolation carrier through a frame connected by upper and lower bearings to form a rotating heating device;

[0040] (2): Select two combinable gyroscopes, one with high precision but large temperature drift coefficient as the reference gyroscope, and one with low precision but small temperature drift coefficient as the monitoring gyroscope, and place them on the same carrier plane inside the rotating heating cover, where the precision ratio of the low-precision gyroscope to the high-precision gyroscope is not less than 5, and the temperature sensitivity coefficient of the low-precision gyroscope to the high-precision gyroscope does not exceed 0.2.

[0041] (3): Conduct rotating heating control: Judge to conduct heating control according to the ratio of the stability of the absolute value of the output difference between the two gyroscopes being greater than the heating start criterion and the stability of the absolute value of the output of the monitoring gyroscope; the rotating heating cover rotates at a fixed speed, the rotation axis is perpendicular to the carrier plane, and the rotation speed is 6° / s. The rotating heating cover is composed of equally wide spherical zone heating sheets in the latitude direction, and the width of the heating sheet is w. Heating resistors are arranged on each heating sheet, and the conduction of the heating resistors is determined and addressed by the temperature compensation unit. The heating temperature and heating time are calculated and determined by the temperature compensation unit.

[0042] (4): Collect and process gyro data: When the ratio of the stability of the absolute value of the output difference between the two gyroscopes does not exceed the temperature heating criterion, the output combination of the two gyroscopes is used as the effective gyro output; the combination method is to use the average value of the monitoring gyro output as the zero bias mean value, and the noise output of the reference gyro as the gyro noise.

[0043] (5): Output and record the gyro data.

[0044] The above specific implementation method of rotating heating control:

[0045] Assume that the temperature performance of the two gyroscopes changes greatly in a short period of time. At this time, the comparator collects a large difference. The gyroscope achieves improvement in the influence of temperature on the gyroscope through rotation and local area heating. The specific angular rate of rotation and heating measures are as follows:

[0046] Typical values of the angular rate are 6° / s. Assuming the rotation time is t seconds, then it stops for t seconds, and the start-stop cycle is repeated until the temperature field is uniform. The rotating heating cover uses a zoned heating method. The rotating heating cover rotates along a fixed rotating axis, and the heating elements are evenly laid on the heating rotating cover in the form of a spherical belt surface. The electrodes of the heating elements are powered in the form of contact brushes. The specific heating temperature can adopt constant temperature control and variable temperature control. Among them, the temperature of the variable temperature control is determined by the compensation coefficient matrix. For the sake of simplicity of description, the present invention preferably adopts constant temperature control (the temperature of the heating element is T out °C). The typical heating curve is as shown in the figure. After the zoned heating temperature reaches the set value, the rotating heating cover rotates continuously left and right periodically, and the start-stop duty cycle is 1:1. It can be but is not limited to the typical event that the start and stop each account for t seconds.

[0047] Perform a differential filtering on the monitoring gyro, that is, perform a difference on the gyro output to obtain the gyro noise. Perform a smoothing filtering on the reference gyro, that is, output the average value of the gyro; add the digital quantity obtained after amplifying the zero bias value of the monitoring gyro to the noise of the monitoring gyro to obtain the composite gyro digital quantity, and this composite digital quantity is used as the final output of the composite gyro. Extract the difference between the drifts of the two gyros as the feedback quantity, send the feedback quantity to the temperature compensation unit, and calculate the heating temperature value and the heating time required through the neural network unit. Heat the position to be heated (heating element) to the specified temperature through the rotating heating cover. The heating position adopts a rotation cycle method, that is, the heating elements near the gyro carrier plane are heated cyclically until the temperature field is uniform. Thus, the balance of the temperature field is realized, and then the temperature drift compensation of the gyro is carried out. In this way, the zero bias mean error of the gyro is reduced.

[0048] The implementation method of the above temperature compensation unit:

[0049] Use a neural network training model to generate a matrix, where T out and t form the output matrix, which are the output heating temperature and heating time respectively. IW and LW are the neural network weight factors, and here 15 neurons are taken for training respectively. The input is a 1*2 input matrix composed of the initial temperature T0 and the gyro error E 误差 (the difference between the outputs of the two gyros). The matrix feedback network is characterized by the weight matrix IW*LW.

[0050]

[0051] An implementation example of the weight matrix of 15 neurons is as follows:

[0052]

[0053] Although embodiments and drawings of the present invention are disclosed for illustrative purposes, those skilled in the art will understand that various substitutions, changes, and modifications are possible without departing from the spirit of the present invention and the appended claims. Therefore, the scope of the present invention is not limited to the content disclosed in the embodiments and drawings.

Claims

1. A high-precision fiber optic gyroscope temperature drift monitoring and error suppression system, characterized in that: It includes a reference gyro, a monitoring gyro, a rotating heating cover, a synchronization carrier, a comparator, and a temperature compensation unit; The reference gyro is a gyro with high precision and large temperature drift coefficient, which is responsible for noise output; The monitoring gyro adopts a gyro with low precision and small temperature drift coefficient, which is responsible for monitoring the zero bias mean; The monitoring gyro and the reference gyro are placed on the same carrier plane, and the sensitive axes of the two gyros are arranged in parallel; The synchronization carrier is used to realize the installation of the reference gyro, the monitoring gyro and the rotating heating cover; The rotating heating cover is arranged outside the reference gyro and the monitoring gyro. The rotating heating cover is spherical and consists of two upper and lower hemispheres. A support point is arranged at the apex of each of the upper and lower hemispheres. The two support points are rotatably supported on the fixed carrier through bearings, so that the heating cover can freely rotate clockwise and counterclockwise around the normal line of the carrier plane. The rotating heating cover is composed of spherical heating plates with equal width in the latitude direction, and each heating plate is provided with a heating resistor. The comparator is used to collect the output signals of the reference gyroscope and the monitoring gyroscope and perform differential operations; The temperature compensation unit is used to determine and address the conduction of the heating resistor according to the differential operation structure of the comparator and adopt a neural network training model, and output the heating temperature and heating time to realize the rotary heating control of the rotary heating cover.

2. The high-precision fiber optic gyro temperature drift monitoring and error suppression system according to claim 1, characterized in that: The accuracy ratio of the monitoring gyro to the reference gyro is not less than 5; the temperature sensitivity coefficient ratio of the monitoring gyro to the reference gyro is not more than 1 / 5.

3. A high-precision fiber optic gyroscope temperature drift monitoring and error suppression system method, based on the high-precision fiber optic gyroscope temperature drift monitoring and error suppression system according to claim 1 or 2, comprising the following steps: Step 1: The outputs of the reference gyro and the monitoring gyro are collected in real time by a comparator, and a difference operation is performed. When the ratio of the absolute value of the difference between the outputs of the two gyros to the absolute value of the output drift of the monitoring gyro is greater than the zero bias threshold set for heating start, it is determined that the temperature excitation causes the gyro to drift. Step 2, when it is confirmed in step 1 that the gyro drifts, the temperature compensation unit is used to control the rotation heating of the rotating heating cover; the rotating heating cover rotates at a fixed speed, and the rotation axis is perpendicular to the carrier plane; the rotating heating cover adopts zone heating to cyclically heat the heating plate near the gyro carrier plane, and after the zone heating temperature reaches the set value, the rotating heating cover periodically rotates left and right continuously, and the rotation start-stop duty ratio is 1:1, until the temperature field is uniform; Step 3: When the ratio of the absolute value of the difference between the outputs of the two gyros and the absolute value of the output drift of the monitoring gyro does not exceed the set zero bias threshold, the output combination of the two gyros is used as the effective gyro output; the combination method is: the average value of the monitoring gyro output is used as the zero bias mean value, and the reference gyro noise output is used as the gyro noise; Step 4: Output and record the gyro data obtained in step 3.

4. The high-precision fiber optic gyro temperature drift monitoring and error suppression system method according to claim 3, characterized in that: In step 1, the expression of the zero bias threshold k is set as: k=|E 参考 -AND 监控 | / |And 监控 | ≥ 5; Where, E is the gyro output bias stability, E 误差 is the stability of the difference between the output speed of the reference gyro and the monitoring gyro, E 监控 To monitor the stability of the gyro output speed.

5. The high-precision fiber optic gyro temperature drift monitoring and error suppression system method according to claim 3, characterized in that: In step 2, the rotating heating cover is heated at a constant temperature. The heating temperature and heating time are generated by the temperature compensation unit using a neural network training model. The input of the neural network training model is the initial temperature T0 of the environment and the gyro error E 误差 ; T out and t form the output matrix, which are the output heating temperature and heating time respectively. IW and LW are the neural network weight factors, as shown in the following formula:

Images