A laser gyro correction system based on temperature adaptation

Abstract

The invention relates to the technical field of laser gyroscopes, and in particular to a temperature-adaptive laser gyroscope correction system, comprising a gyroscope body and a correction unit, wherein the correction unit comprises: a temperature acquisition module, which is arranged on the gyroscope body and acquires current working temperature data of the gyroscope body; a data operation and processing module, which acquires the acquired temperature data, performs operation and processing through an established functional relationship model between angular velocity and temperature data, and outputs expected angular velocity data; an error correction module, which acquires angular velocity data actually output by the gyroscope body, performs operation and processing on the angular velocity data compared with the expected angular velocity data, obtains current angular velocity error parameters, and calculates angular velocity compensation parameters based on an error evolution model; and a feedback adjustment module, which outputs the angular velocity compensation parameters to a control circuit of the gyroscope body based on the angular velocity compensation parameters, so as to dynamically correct the angular velocity output of the gyroscope body and realize continuous stability and precision of the angular velocity output of the laser gyroscope.

Description

Technical Field

[0001] The invention relates to the technical field of laser gyroscopes, and in particular to a laser gyroscope correction system based on temperature self-adaptation. Background Art

[0002] As a high-precision inertial navigation device, laser gyroscope has a wide range of applications in aerospace, military and civil fields. However, in practical applications, there are still some challenges. During long-term operation, the workload of the internal components of the traditional laser gyroscope causes heat accumulation, which causes the temperature of the laser gyroscope to change, thus affecting the optical properties of the laser gyroscope, causing the angular velocity output of the laser gyroscope to drift or error, affecting the measurement accuracy of the laser gyroscope. Therefore, in order to ensure the measurement accuracy of the laser gyroscope, it is necessary to calibrate the angular velocity output of the laser gyroscope.

[0003] In the prior art, the correction system of the laser gyroscope mostly adopts static correction methods. These methods usually perform one-time correction during manufacturing or verification, and cannot adapt to the dynamic changes of ambient temperature. For example, Chen Jian, Liu Liying and others mentioned in the relevant document "Research on Temperature Compensation Model of High-Precision Laser Gyroscope" that the output zero bias of the high-precision laser gyroscope will drift with temperature changes, which limits its application. In order to reduce and compensate the influence of temperature on the zero bias of the gyroscope, improve the application performance of the gyroscope, and reduce the equipment cost, a temperature compensation model considering multiple influencing factors is established, and a corresponding temperature test is designed; the test results show that the compensation effect of the temperature compensation model is significant, and the trend of zero bias changing with temperature can be basically eliminated. However, when correcting, it mainly relies on fixed correction parameters and lacks the ability to respond to temperature changes in real time. In other correction systems, although neural network models are also used to construct multi-dimensional variables and serve as the input of the correction model to realize the correction of the angular velocity of the laser gyroscope, its model training is relatively complex, and in actual application, it is not friendly enough to application scenarios with high real-time requirements, and it is difficult to correct the angular velocity output in a timely and accurate manner.

[0004] In view of the limitations of existing technologies in laser gyroscope correction, proposing a laser gyroscope correction system based on temperature adaptation has become a technical problem that needs to be urgently solved in the industry. Summary of the invention

[0005] The present invention aims to provide a laser gyro correction system based on temperature adaptation, which is used to solve the static and linear problems of temperature correction in the prior art, improve the real-time and adaptive capabilities of the system, and ensure the accuracy and reliability of the correction signal by introducing advanced signal conversion circuits and drive control circuits, thereby significantly improving the measurement accuracy and stability of the laser gyro.

[0006] Specifically, the present invention is achieved through the following technical solutions:

[0007] A temperature-adaptive laser gyro correction system includes a gyro body and a correction unit, wherein the correction unit includes:

[0008] The temperature acquisition module is arranged on the gyroscope body and is used to collect the temperature data of the gyroscope body when it is currently working;

[0009] The data calculation and processing module obtains the collected temperature data, performs calculation and processing through the pre-established functional relationship model between the angular velocity and the temperature data, and outputs the final predicted angular velocity data;

[0010] The error correction module obtains the angular velocity data actually output by the gyroscope body, performs a comparison operation with the final predicted angular velocity data output by the data operation processing module, obtains the current angular velocity error parameter, and calculates the angular velocity compensation parameter based on the error evolution model;

[0011] The feedback adjustment module outputs the angular velocity compensation parameter to the closed-loop control circuit of the gyroscope body based on the angular velocity compensation parameter to dynamically correct the angular velocity output of the gyroscope body.

[0012] In a further solution, the temperature acquisition module includes a first temperature sensor and a second temperature sensor, the first temperature sensor is located inside the resonant cavity of the gyroscope body, and the second temperature sensor is located on the outer shell of the gyroscope body, and the first temperature sensor and the second temperature sensor respectively collect internal and external temperature data of the gyroscope body.

[0013] In a further solution, the temperature acquisition module also includes a computing component, which pre-processes the internal and external temperature data of the gyroscope body to eliminate the influence of data noise.

[0014] In a further solution, the preprocessing includes cleaning the collected temperature data, removing outliers, and performing weighted average processing to reduce random noise in the data and make the data smoother. The calculation process of weighted average processing is:

[0015] ;

[0016] in, It means that the temperature acquisition module is Output temperature data, It means in time The internal temperature data collected, It means in time The collected external temperature data, are the weighting coefficients of the internal temperature data and the external temperature data, which satisfy .

[0017] In a further solution, in the data operation and processing module, the functional relationship model of the angular velocity and the temperature data is a hybrid model, which includes a polynomial regression model and a neural network model. The polynomial regression model outputs preliminary angular velocity prediction data based on the collected temperature data. The neural network model inputs the preliminary angular velocity prediction data, the temperature data and the temperature dynamic feature vector to obtain the correction data of the preliminary angular velocity prediction data. The hybrid model calculates the preliminary angular velocity prediction data and the correction data to obtain the final predicted angular velocity data, and the calculation process satisfies:

[0018] ;

[0019] in, It refers to the mixed model in time The final predicted angular velocity data output is Refers to the preliminary angular velocity prediction data output by the polynomial regression model, It refers to the neural network model in time Output the correction data of the preliminary angular velocity prediction data.

[0020] In a further scheme, the polynomial regression model describes the effect of temperature on angular velocity by fitting the correlation coefficient, thereby capturing the linear and low-order nonlinear relationship between temperature data and angular velocity, and calculating and outputting preliminary angular velocity prediction data. The calculation process satisfies the following:

[0021] ;

[0022] in, are the coefficients of the polynomial regression model, corresponding to the temperature terms of different powers, where is a constant term, arrive is corresponding to 1 to the power of The coefficient of the power, Indicates temperature of Power and corresponding regression coefficient The product of .

[0023] In a further solution, the neural network model is a recurrent neural network structure, which includes an input layer, a hidden layer and an output layer, wherein the input layer takes the angular velocity output value of the polynomial regression model as input, the hidden layer includes multiple feature extraction layers, and the multiple feature extraction layers are used to extract temperature data and temperature dynamic feature vectors from the collected temperature data. The output layer inputs the dynamic feature information of the temperature data and the preliminary angular velocity prediction data output by the polynomial regression model based on the activation function, and performs calculation processing to obtain the correction data of the preliminary angular velocity prediction data, and the calculation process is:

[0024] ;

[0025] in, , Refers to time The temperature change rate and temperature change acceleration, that is, the dynamic eigenvector, refers to the weight matrix, where is the number of feature extraction layers; is the bias vector, Is the activation function, which is specifically the Sigmoid activation function.

[0026] In a further solution, the error correction module obtains the angular velocity data actually output by the gyroscope body, and compares and processes it with the final predicted angular velocity data output by the data operation and processing module to obtain the current angular velocity error parameter. The operation and processing process is:

[0027] ;

[0028] in, Refers to the time The current angular velocity error parameter obtained is: Refers to the time The angular velocity data actually output by the gyroscope body.

[0029] In a further scheme, after obtaining the current angular velocity error parameter, the error correction module calculates the angular velocity compensation parameter based on the error evolution model, wherein the error evolution model is established based on the historical angular velocity error parameter sequence, and the historical angular velocity error parameter is used to train the model to obtain the evolution law of the error over time, and the current angular velocity error parameter is used as the model input to calculate and predict the angular velocity error parameter of the next time step, and the angular velocity error parameter of the next time step and the current angular velocity error parameter are used as calculation factors for input operation together, so as to obtain the final angular velocity compensation parameter; the operation process is:

[0030] ;

[0031] ;

[0032] in, is the predicted angular velocity error parameter for the next time step, is the constant term of the error evolution model, is the order of the evolution model, that is, the number of historical angular velocity error parameters, is the autoregression coefficient, Indicates the delay step number of the historical angular velocity error parameter, is the final angular velocity compensation parameter, is the proportionality coefficient.

[0033] In a further scheme, the closed-loop control circuit includes a signal conversion circuit and a drive control circuit, the signal conversion circuit includes an MCU chip, a D / A converter and an amplifier, wherein the MCU chip is used to receive angular velocity compensation parameters and convert them into correction analog signal output, the D / A converter is used to convert the correction analog signal output by the MCU into a correction electrical signal, the amplifier is used to amplify the analog correction electrical signal, and the drive control circuit is an H-bridge drive circuit, which is used to receive the amplified correction electrical signal and convert it into a control voltage to correct the angular velocity output signal of the gyroscope body.

[0034] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0035] 1. The present invention realizes comprehensive collection and monitoring of the overall temperature distribution data of the gyroscope body through the deployment of multiple temperature sensors. At the same time, the internal and external temperature data after collection are cleaned to remove abnormal values, and random noise in the data is reduced by weighted average processing, so that the data is smoother and the data quality is improved. Different weighting coefficients are used to more accurately reflect the real temperature state of the gyroscope, thereby ensuring that the temperature data of the gyroscope body is more accurate, so that the correction system can more accurately reflect the influence of temperature on angular velocity, thereby significantly improving the accuracy of the correction result;

[0036] 2. The present invention introduces a hybrid model formed by polynomial regression and a neural network model in the data operation and processing module, which combines the linear prediction ability of polynomial regression and the nonlinear fitting ability of the neural network. Specifically, the linear relationship and low-order nonlinear relationship between temperature and angular velocity are captured by the polynomial regression model to preliminarily establish a functional relationship between temperature and angular velocity and generate preliminary angular velocity prediction data, while the neural network model uses a powerful nonlinear fitting ability to input preliminary angular velocity prediction data, temperature data and temperature dynamic feature vectors to deeply explore complex temperature dynamic features and perform high-precision correction on the preliminary prediction data, thereby obtaining correction data of the preliminary angular velocity prediction data and obtaining the final predicted angular velocity data, which further improves the complexity and accuracy of the model and significantly improves the prediction accuracy and correction effect of the system;

[0037] 3. The present invention further introduces an error evolution model into the error correction module, and the error evolution model is established based on the historical angular velocity error parameter sequence, and the model is trained using historical data to obtain the evolution law of the error over time. By using the current angular velocity error parameter as the model input, the system can predict the angular velocity error parameter of the next time step, which not only considers the current error state, but also considers the historical evolution trend of the error, so that the angular velocity compensation parameter is more accurate and forward-looking, and the predicted error value of the next time step is combined with the current error parameter, so that the correction system can calculate a more accurate angular velocity compensation parameter, thereby significantly improving the dynamic response capability and long-term stability of the correction;

[0038] In summary, this system not only improves the temperature data acquisition accuracy and angular velocity prediction accuracy through precise temperature data acquisition, application of a hybrid model of polynomial regression and neural network, and calculation of angular velocity compensation parameters based on the error evolution model, but also significantly improves the correction effect of the laser gyroscope and the overall performance of the system through dynamic compensation calculation, thereby achieving highly accurate correction of the laser gyroscope angular velocity data. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of this application, and do not constitute a limitation of the embodiments of the present invention. In the drawings:

[0040] Figure 1 It is a schematic diagram of the overall structure of the present invention;

[0041] Figure 2 It is a schematic diagram of the process of the correction unit of the present invention.

[0042] Marks and corresponding parts names in the attached drawings:

[0043] 1. Gyro body; 10. Resonant cavity; 11. Outer shell; 2. Correction unit; 20. Temperature acquisition module; 200. First temperature sensor; 201. Second temperature sensor; 21. Data operation and processing module; 22. Error correction module; 23. Feedback adjustment module. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and drawings. The schematic implementation modes and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention. It should be noted that the present invention is already in the actual development and use stage.

[0045] Laser gyroscopes are widely used in high-precision measurement fields such as aerospace, navigation, and earthquake monitoring. These fields require extremely high measurement accuracy, and any small error will have a significant impact on the performance and reliability of the system. In actual applications, they are often affected by temperature changes, thereby affecting their measurement accuracy. In order to ensure that the laser gyroscope can maintain high-precision angular velocity measurement in different temperature environments, the existing technology mostly uses a correction system to perform temperature compensation on the laser gyroscope. The common temperature compensation method is mainly to install a temperature sensor inside the gyroscope and perform simple linear or nonlinear compensation based on the measured temperature data. However, this single compensation method can often only handle part of the temperature error and cannot fully cover the impact of all temperature changes on the performance of the gyroscope, resulting in limited compensation effect. In addition, when compensating and correcting, a fixed model is often used for temperature compensation, which lacks the ability to adapt to different working environments and temperature conditions. When dealing with rapid temperature changes, the dynamic response capability is poor, and it is difficult to track and compensate for the errors caused by temperature changes in real time. This results in the measurement accuracy of the laser gyroscope being unable to be effectively guaranteed in an environment with rapid temperature changes.

[0046] In view of this, in order to overcome the defects of the existing correction system, this scheme proposes a laser gyro correction system based on temperature adaptation, which can comprehensively solve the defects of the existing technology in temperature error correction through accurate temperature acquisition of multi-sensor fusion, dynamic error correction of hybrid model, adaptive model optimization and efficient and simple system design, and realize high-precision and real-time compensation of laser gyro temperature error, significantly improve the measurement accuracy and system stability of laser gyro, and meet the needs of high-precision applications. The specific implementation method is recorded as follows.

[0047] Example

[0048] Please refer to the attached Figure 1 to Figure 2The present embodiment provides a temperature-adaptive laser gyro correction system, comprising a gyro body 1 and a correction unit 2, wherein the correction unit 2 comprises: a temperature acquisition module 20, arranged on the gyro body 1, for acquiring temperature data of the gyro body 1 when it is currently working; a data operation and processing module 21, for acquiring the acquired temperature data, performing operation and processing through a pre-established functional relationship model between the angular velocity and the temperature data, and outputting final predicted angular velocity data; an error correction module 22, for acquiring the angular velocity data actually output by the gyro body 1, performing operation and processing on the angular velocity data compared with the final predicted angular velocity data output by the data operation and processing module 21, obtaining a current angular velocity error parameter, and calculating an angular velocity compensation parameter based on an error evolution model; a feedback adjustment module 23, based on the angular velocity compensation parameter, outputting it to a closed-loop control circuit of the gyro body 1, so as to dynamically correct the angular velocity output of the gyro body 1.

[0049] It should be understood that when the laser gyro is working, its measurement accuracy is significantly affected by temperature changes, that is, the temperature changes inside the resonant cavity 10 of the gyro body 1 and the outer shell will directly affect the propagation path of the laser and the overall performance of the gyro. If only a single temperature sensor is used to collect the temperature data of the resonant cavity 10 or the outer shell of the gyro body 1, it is impossible to fully reflect the temperature changes of the internal working environment and the external environment, thereby resulting in incomplete temperature compensation, affecting the measurement accuracy and stability of the laser gyro. Only a single point temperature collection inside the resonant cavity 10 of the gyro body 1 or on the outer shell cannot fully capture the temperature changes of the gyro body 1 under different working environments and working conditions. For the gyro body 1, its internal and external temperature changes have a significant gradient effect, and only collecting temperature data at a single position cannot reflect this gradient effect. The inside of the resonant cavity 10 may cause local overheating due to high energy density, while the temperature of the outer shell 11 may be lower. This internal and external temperature gradient has an important impact on the laser propagation path and frequency stability.

[0050] Therefore, for this embodiment, a further preferred implementation is: the temperature acquisition module 20 includes a first temperature sensor 200 and a second temperature sensor 201, the first temperature sensor 200 is located inside the resonant cavity 10 of the gyroscope body 1, and the second temperature sensor 201 is located on the outer shell of the gyroscope body 1, and the first temperature sensor 200 and the second temperature sensor 201 respectively collect internal and external temperature data of the gyroscope body 1. By setting the first temperature sensor 200 and the second temperature sensor 201, the temperature data of the inside of the resonant cavity 10 and the outer shell 11 are collected respectively, so as to fully reflect the temperature changes of the gyroscope body 1 at different positions, capture the temperature gradient and environmental interference factors, and establish a more accurate and comprehensive temperature compensation model, thereby improving the measurement accuracy and stability of the laser gyroscope. For the first temperature sensor 200 and the second temperature sensor 201, a further preferred implementation method is as follows: a plurality of first temperature sensors 200 are provided and are evenly distributed inside the resonant cavity 10 of the gyroscope body 1; a plurality of second temperature sensors 201 are provided and are arrayed on the outer shell 11 of the gyroscope body 1, so as to realize comprehensive temperature collection of the gyroscope body 1.

[0051] Preferably, the temperature acquisition module 20 further includes a computing component, through which the internal and external temperature data of the gyro body 1 are preprocessed to eliminate the influence of data noise. The preprocessing includes cleaning the collected temperature data, removing abnormal values, and performing weighted average processing to reduce random noise in the data to make the data smoother. The computing process of the weighted average processing is:

[0052] ;

[0053] in, It means that the temperature acquisition module 20 is Output temperature data, It means in time The internal temperature data collected, It means in time The collected external temperature data, are the weighting coefficients of the internal temperature data and the external temperature data, which satisfy . Based on the above, by cleaning and removing outliers and weighted averaging, the random noise in the temperature data can be effectively reduced, making the data smoother and more accurate, thereby improving the accuracy and reliability of the subsequent temperature compensation model, and ensuring the measurement stability and accuracy of the gyro body 1 under different temperature environments. That is, by weighted processing the temperature data inside the resonant cavity 10 and the outer shell 11 of the gyro body 1, more accurate and comprehensive temperature data can be obtained. Specifically, the weighted average processing can not only reduce the random noise in the collected data, improve the smoothness and stability of the data, but also comprehensively reflect the temperature changes inside and outside the gyro body 1, thereby providing more accurate overall temperature distribution information. It effectively indicates the temperature state of the gyro body 1 under different working conditions and environments, improves the representativeness and accuracy of the temperature data, and ultimately improves the angular velocity measurement accuracy and reliability of the laser gyroscope. It should be added here that the weighting coefficient refers to the importance and influence of the internal temperature data and the external temperature data of the gyro body 1. In order to better reflect the overall temperature state of the gyro body 1 when it is working, the corresponding weighting coefficient can be set based on actual needs to balance the contributions of the two. For example, in a high-energy-density working environment, the temperature change inside the resonant cavity 10 has a greater impact on the laser gyro. At this time, the weight of the internal temperature data can be set higher, that is, , while the weight of external temperature data is lower, i.e. If the ambient temperature changes significantly and there are more external interferences in the working environment, the external temperature data on the outer shell 11 has a greater impact on the gyroscope body, then you can set , .

[0054] As a further preference of the above embodiment, in the data operation processing module 21, the functional relationship model of the angular velocity and the temperature data is a hybrid model, and the hybrid model includes a polynomial regression model and a neural network model. The polynomial regression model outputs preliminary angular velocity prediction data based on the collected temperature data, and the neural network model inputs the preliminary angular velocity prediction data, the temperature data and the temperature dynamic feature vector to obtain correction data of the preliminary angular velocity prediction data. The hybrid model calculates the preliminary angular velocity prediction data and the correction data to obtain the final predicted angular velocity data, and the calculation process satisfies:

[0055] ;

[0056] in, It refers to the mixed model in time The final predicted angular velocity data output is Refers to the preliminary angular velocity prediction data output by the polynomial regression model, It refers to the neural network model in time Output the correction data of the preliminary angular velocity prediction data.

[0057] It should be understood that the hybrid model used in the present technical solution combines the polynomial regression model and the neural network model, which significantly improves the accuracy and robustness of the angular velocity prediction. Among them, the polynomial regression model can quickly capture the linear and low-order nonlinear relationship between temperature data and angular velocity, and provide preliminary prediction results; secondly, the neural network model has a strong nonlinear fitting ability, which can deeply explore the complex temperature dynamic characteristics and perform high-precision correction on the preliminary prediction data, thereby compensating for the shortcomings of the polynomial regression model. It not only retains the advantages of the polynomial regression model's efficient calculation, simplicity and intuitiveness, but also integrates the advantages of the neural network model in processing complex nonlinear relationships, ensuring that high-precision, low-error angular velocity prediction can be provided under a wide range of working environments and temperature changes.

[0058] As the above embodiment, specifically, the polynomial regression model describes the influence of temperature on angular velocity by fitting the correlation coefficient, thereby capturing the linear and low-order nonlinear relationship between temperature data and angular velocity, and calculating and outputting preliminary angular velocity prediction data, and the calculation process satisfies the following:

[0059] ;

[0060] in, are the coefficients of the polynomial regression model, corresponding to the temperature terms of different powers, where is a constant term, arrive is corresponding to 1 to the power of The coefficient of the power, Indicates temperature of Power and corresponding regression coefficient By fitting the polynomial coefficients, the polynomial regression model can accurately describe the effect of temperature on angular velocity. In particular, the linear and low-order nonlinear relationships enable the model to quickly capture the direct effect of temperature changes on angular velocity, providing reliable basic data for subsequent predictions. At the same time, it not only simplifies the computational complexity, but also improves the real-time performance and accuracy of the system, reflecting the significant progress of this technical solution in terms of accuracy and efficiency.

[0061] More specifically, the neural network model is a recurrent neural network structure, which includes an input layer, a hidden layer and an output layer, wherein the input layer takes the angular velocity output value of the polynomial regression model as input, the hidden layer includes a plurality of feature extraction layers, and the plurality of feature extraction layers are used to extract temperature data and temperature dynamic feature vectors from the collected temperature data, and the output layer inputs the dynamic feature information of the temperature data and the preliminary angular velocity prediction data output by the polynomial regression model based on the activation function, and performs calculation processing to obtain correction data of the preliminary angular velocity prediction data, and the calculation process is as follows:

[0062] ;

[0063] in, , Refers to time The temperature change rate and temperature change acceleration, that is, the dynamic eigenvector, refers to the weight matrix, where is the number of feature extraction layers; is the bias vector, is an activation function, which is specifically a Sigmoid activation function, which is used to introduce nonlinear relationships. It should be noted that the neural network model is particularly suitable for processing time series data, and its structure can effectively capture temporal dependencies and dynamic change characteristics. In the processing of temperature data, the rate of change of temperature and the change acceleration (i.e., the dynamic feature vector) have a crucial impact on the angular velocity. RNN can extract these dynamic features from continuous temperature data through the feature extraction layer in its hidden layer, so as to more accurately predict the change of angular velocity, and the recurrent neural network structure can finely correct the preliminary prediction data of the polynomial regression model through its powerful nonlinear fitting ability and dynamic feature extraction ability to make up for its shortcomings. At the same time, for the recurrent neural network, it has the ability of adaptive adjustment, and can dynamically adjust the weights and biases according to the input data, improve the adaptability and stability of the model, and enable it to provide high-precision and low-error predicted angular velocity data results under different working environments and temperature conditions.

[0064] As a further preferred embodiment of the above, the error correction module 22 obtains the angular velocity data actually output by the gyro body, and compares and processes it with the final predicted angular velocity data output by the data operation and processing module 21 to obtain the current angular velocity error parameter. The operation and processing process is as follows:

[0065] ;

[0066] in, Refers to the time The current angular velocity error parameter obtained is: Refers to the time Angular velocity data actually output by the gyro body 1. Based on the above, the error correction module 22 is further specifically described here: First, the error correction module 22 can accurately capture the error source in the system by obtaining the actual output data of the gyro body 1 in real time (that is, the actual output angular velocity data. Due to the influence of temperature error, the actual output angular velocity data itself has errors) and comparing it with the final predicted angular velocity data, that is, using a simple and effective difference calculation method (subtracting the final predicted angular velocity data from the actual output angular velocity data), so as to quickly obtain the error parameters, ensure the efficient operation and real-time response capability of the system, and adjust and correct the prediction results in real time, which significantly improves the accuracy and reliability of angular velocity measurement; secondly, through the continuous error correction process of the error correction module 22, the system can adaptively optimize its prediction model, further improving the adaptability and robustness of the system in complex working environments.

[0067] It should be noted that the present technical solution is different from the prior art in that after the error correction module 22 obtains the current angular velocity error parameter, the angular velocity error parameter is not directly used, but is calculated and processed through the error evolution model to obtain the final angular velocity compensation parameter. This is because the gyroscope body 1 is a continuous dynamic process when working, and the current angular velocity error parameter is only an instantaneous value at a certain moment, which cannot reflect the dynamic change and cumulative effect of the error over time. Directly using the error parameter may cause the prediction result to be greatly affected by short-term fluctuations, thereby introducing more random noise and uncertainty, and reducing the stability and accuracy of the system. In addition, the instantaneous error parameter does not take into account the time correlation of the error and the influence of historical data, and cannot effectively capture and correct the long-term error trend and system offset, resulting in incomplete and inaccurate error compensation.

[0068] Therefore, as a further optimization of the above embodiment, in order to ensure the high accuracy and robustness of the angular velocity prediction, it is necessary to further dynamically process the current error parameters so as to fully consider the time evolution and cumulative effect of the error, thereby providing a more reliable error correction. After obtaining the current angular velocity error parameters, the error correction module 22 calculates the angular velocity compensation parameters based on the error evolution model, wherein the error evolution model is established based on the historical angular velocity error parameter sequence, and the historical angular velocity error parameters are used to train the model to obtain the law of error evolution over time, and the current angular velocity error parameters are used as model input to calculate and predict the angular velocity error parameters of the next time step, and the angular velocity error parameters of the next time step and the current angular velocity error parameters are used as calculation factors and input into the operation together, thereby obtaining the final angular velocity compensation parameters; the operation process includes:

[0069] Prediction formula: ;

[0070] The compensation formula is: ;

[0071] in, is the predicted angular velocity error parameter for the next time step, is the constant term of the error evolution model, is the order of the evolution model, that is, the number of historical angular velocity error parameters, is the autoregression coefficient, Indicates the delay step number of the historical angular velocity error parameter, is the final angular velocity compensation parameter, is the proportional coefficient. Based on the above, the error evolution model can capture the trend and law of error changes over time by tracking and processing the error parameters in continuous time, thereby providing more accurate and stable error correction. Specifically, the model uses the Kalman filter algorithm to combine the current error parameters with the historical error trend, establish a mathematical model of error evolution, and predict future error changes. Through dynamic adjustment in continuous time, the model can effectively filter out short-term fluctuations and random noise, capture long-term error trends and system offsets, and thus provide more accurate and stable angular velocity compensation parameters. It utilizes the correlation of time series data, considers the cumulative effect and dynamic change law of errors, and significantly improves the accuracy and reliability of error compensation. With this technical means, the system can maintain high-precision angular velocity prediction in complex dynamic environments, enhance the robustness and adaptability of the overall system, and ensure the continuous stability and accuracy of the measurement results.

[0072] As the above embodiment, it is further preferred that the closed-loop control circuit includes a signal conversion circuit and a drive control circuit, the signal conversion circuit includes an MCU chip, a D / A converter and an amplifier, wherein the MCU chip is used to receive the angular velocity compensation parameter and convert it into a correction analog signal output, the D / A converter is used to convert the correction analog signal output by the MCU into a correction electrical signal, the amplifier is used to amplify the analog correction electrical signal, and the drive control circuit is an H-bridge drive circuit, which is used to receive the amplified correction electrical signal and convert it into a control voltage to correct the angular velocity output signal of the gyro body 1. Based on the above technical solution, it realizes the accurate processing of error parameters and dynamic correction of angular velocity. By responding to error changes in real time, the correction system can capture and adjust the angular velocity output, thereby effectively filtering out system errors and environmental interference, significantly improving the accuracy and stability of the output angular velocity measurement of the gyro body 1, so that the correction unit 2 can continuously track and correct errors, ensuring that the angular velocity compensation parameters are more accurate, thereby meeting the continuous stability and accuracy of the angular velocity output of the laser gyro.

[0073] It should be noted that in order to make the technical solution of the present application clearer, the professional terms involved in the present application are explained here;

[0074] In this application, the term "MCU chip" refers to a microcontroller unit, which is usually used to control electronic devices and perform embedded system functions. In this application, it is responsible for receiving the angular velocity compensation parameters from the error correction module and converting them into correction analog signal output;

[0075] In this application, the term "D / A converter" refers to a device that converts a digital signal into an analog signal, receives a digital value and converts it into a continuously changing voltage or current signal. This analog signal can represent the amplitude of the original digital signal. In this application, it is used to convert the digital correction signal output by the MCU chip into an analog voltage signal;

[0076] In this application, the term "amplifier" refers to a device that increases the amplitude of a voltage or current signal, and in this application is used to enhance the analog correction electrical signal output by the D / A converter to ensure that the signal has sufficient amplitude to drive the H-bridge drive circuit;

[0077] In this application, the term "H-bridge drive circuit" refers to a circuit that controls the forward and reverse rotation of a motor or other load. It is composed of four switching devices (such as transistors or relays) arranged in an H shape, and can provide outputs in four quadrants. In this application, it receives the amplified correction electrical signal and converts it into a control voltage for accurately adjusting the voltage of the piezoelectric ceramic driver inside the gyroscope to achieve dynamic correction of the angular velocity output of the gyroscope body; the components referred to by the above-mentioned professional terms are all existing mature technologies, and their specific models are not described in detail here.

[0078] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A temperature-adaptive laser gyro correction system, comprising a gyro body (1) and a correction unit (2), characterized in that: The correction unit (2) comprises: A temperature collection module (20) is arranged on the gyroscope body (1) and is used to collect temperature data of the gyroscope body (1) when it is currently working; A data operation and processing module (21) acquires the collected temperature data, performs operation and processing through a pre-established functional relationship model between the angular velocity and the temperature data, and outputs final predicted angular velocity data; The error correction module (22) obtains the angular velocity data actually output by the gyroscope body (1), performs a comparison operation with the final predicted angular velocity data output by the data operation processing module (21), obtains a current angular velocity error parameter, and calculates an angular velocity compensation parameter based on an error evolution model; A feedback adjustment module (23) outputs the angular velocity compensation parameter to a closed-loop control circuit of the gyroscope body (1) based on the angular velocity compensation parameter, so as to dynamically correct the angular velocity output of the gyroscope body (1); Wherein, in the data operation processing module (21), the functional relationship model between the angular velocity and the temperature data is a hybrid model, the hybrid model includes a polynomial regression model and a neural network model, the polynomial regression model outputs preliminary angular velocity prediction data based on the collected temperature data, the neural network model inputs the preliminary angular velocity prediction data, the temperature data and the temperature dynamic feature vector to obtain correction data of the preliminary angular velocity prediction data, the hybrid model calculates the preliminary angular velocity prediction data and the correction data to obtain the final predicted angular velocity data, and the calculation process satisfies: ； in, It refers to the mixed model in time The final predicted angular velocity data output is Refers to the preliminary angular velocity prediction data output by the polynomial regression model, It refers to the neural network model in time Output correction data for preliminary angular velocity prediction data; The error correction module (22) obtains the angular velocity data actually output by the gyroscope body (1), and compares and processes the angular velocity data with the final predicted angular velocity data output by the data operation processing module (21) to obtain the current angular velocity error parameter. The operation processing process is as follows: ； in, Refers to the time The current angular velocity error parameter obtained is: Refers to the time The angular velocity data actually output by the gyroscope body (1); After obtaining the current angular velocity error parameter, the error correction module (22) calculates the angular velocity compensation parameter based on the error evolution model, wherein the error evolution model is established based on the historical angular velocity error parameter sequence, and the historical angular velocity error parameter is used to train the model to obtain the law of error evolution over time, and the current angular velocity error parameter is used as a model input to calculate and predict the angular velocity error parameter of the next time step, and the angular velocity error parameter of the next time step and the current angular velocity error parameter are used as calculation factors and input into the operation together, so as to obtain the final angular velocity compensation parameter; the operation process includes: ； ； in, is the predicted angular velocity error parameter for the next time step, is the constant term of the error evolution model, is the order of the evolution model, that is, the number of historical angular velocity error parameters, is the autoregression coefficient, Indicates the delay step number of the historical angular velocity error parameter, is the final angular velocity compensation parameter, is the proportionality coefficient.

2. The temperature-adaptive laser gyro correction system according to claim 1, characterized in that: The temperature acquisition module (20) comprises a first temperature sensor (200) and a second temperature sensor (201); the first temperature sensor (200) is located inside the resonant cavity (10) of the gyroscope body (1); the second temperature sensor (201) is located on the outer shell (11) of the gyroscope body (1); the first temperature sensor (200) and the second temperature sensor (201) respectively acquire internal and external temperature data of the gyroscope body (1).

3. The temperature-adaptive laser gyro correction system according to claim 2, characterized in that: The temperature acquisition module (20) further comprises a computing component, through which the internal and external temperature data of the gyroscope body (1) are pre-processed to eliminate the influence of data noise.

4. The temperature-adaptive laser gyro correction system according to claim 3, characterized in that: The preprocessing includes cleaning the collected temperature data, removing abnormal values, and performing weighted average processing to reduce random noise in the data and make the data smoother. The operation process of the weighted average processing is: ； in, It refers to the temperature acquisition module (20) at time Output temperature data, It means in time The internal temperature data collected, It means in time The collected external temperature data, are the weighting coefficients of the internal temperature data and the external temperature data, which satisfy .

5. The temperature-adaptive laser gyro correction system according to claim 1, characterized in that: The polynomial regression model describes the effect of temperature on angular velocity by fitting the correlation coefficient, thereby capturing the linear and low-order nonlinear relationship between temperature data and angular velocity, and calculating and outputting preliminary angular velocity prediction data. The calculation process satisfies the following: ； in, are the coefficients of the polynomial regression model, corresponding to the temperature terms of different powers, where is a constant term, arrive is corresponding to 1 to the power of The coefficient of the power, Indicates temperature of Power and corresponding regression coefficient The product of .

6. The temperature-adaptive laser gyro correction system according to claim 1, characterized in that: The neural network model is a recurrent neural network structure, which includes an input layer, a hidden layer and an output layer, wherein the input layer takes the angular velocity output value of the polynomial regression model as input, the hidden layer includes a plurality of feature extraction layers, and the plurality of feature extraction layers are used to extract temperature data and temperature dynamic feature vectors from the collected temperature data. The output layer inputs the dynamic feature information of the temperature data and the preliminary angular velocity prediction data output by the polynomial regression model based on the activation function, and performs calculation processing to obtain correction data of the preliminary angular velocity prediction data, and the calculation process is as follows: ； in, , Refers to time The temperature change rate and temperature change acceleration, that is, the dynamic eigenvector, refers to the weight matrix, where is the number of feature extraction layers; is the bias vector, Is the activation function, which is specifically the Sigmoid activation function.

7. The temperature-adaptive laser gyro correction system according to claim 1, characterized in that: The closed-loop control circuit includes a signal conversion circuit and a drive control circuit. The signal conversion circuit includes an MCU chip, a D / A converter and an amplifier. The MCU chip is used to receive angular velocity compensation parameters and convert them into correction analog signal outputs. The D / A converter is used to convert the correction analog signal output by the MCU into a correction electrical signal. The amplifier is used to amplify the analog correction electrical signal. The drive control circuit is an H-bridge drive circuit, which is used to receive the amplified correction electrical signal and convert it into a control voltage to correct the angular velocity output signal of the gyroscope body.

Images