IMU dynamic calibration and compensation method, device, vehicle-mounted navigation equipment and storage medium

Abstract

The present invention discloses an IMU dynamic calibration and compensation method, device, vehicle-mounted navigation equipment and storage medium. The IMU dynamic calibration and compensation method includes: obtaining the original data of the IMU, sampling and integrating it with environmental parameters to obtain an initial measurement sequence; periodically decomposing and calculating the influence of environmental factors on the IMU output value to obtain a measurement supplement sequence; performing segmented calibration and iterating to obtain a calibration reference value; analyzing the error variation law, characteristic analysis to obtain a calibration mapping amount; performing environmental factor decomposition to obtain a compensation factor; dynamically testing each compensation factor, and obtaining a correction coefficient through threshold comparison; using the correction coefficient to compensate the original data, and reconstructing the data to obtain a correction sequence; performing stability analysis on the correction sequence, and obtaining a drift parameter according to the change trend; performing difference analysis on the drift parameter and the calibration reference, and obtaining a final calibration result through multi-point correction. The technical solution of the present invention can accurately obtain the original data of the IMU, effectively deal with environmental influences, realize dynamic compensation and ensure long-term stability.

Description

Technical Field

[0001] The present invention relates to the field of vehicle-mounted navigation technology, and in particular to an IMU dynamic calibration and compensation method, device, vehicle-mounted navigation equipment and storage medium. Background Art

[0002] As a key sensor for measuring the motion state and attitude of a vehicle, IMU (Inertial Measurement Unit) is widely used in the fields of vehicle navigation and attitude measurement. However, in actual use, the measurement accuracy of IMU is easily affected by environmental factors such as temperature change, vibration and magnetic field. Especially in the vehicle environment, temperature changes caused by engine heat dissipation can cause zero drift, vibrations generated by vehicle operation can cause measurement fluctuations, and magnetic fields generated by surrounding electrical equipment and infrastructure can interfere with heading angle measurement. When these environmental factors act at the same time, complex coupling effects will also occur, causing the measurement error to increase further.

[0003] Existing IMU calibration methods are mainly performed in static environments, which makes it difficult to effectively handle error changes under dynamic conditions. In addition, the compensation model is too simple to accurately describe the nonlinear error characteristics of the IMU in complex environments. In addition, the acquisition of calibration parameters often requires a professional laboratory environment, the acquisition of environmental parameters and IMU data is poorly synchronized, and the compensation parameters lack an adaptive adjustment mechanism. These all limit the measurement accuracy of the IMU in practical applications.

[0004] Therefore, there is an urgent need for a solution that can accurately obtain IMU raw data, effectively handle environmental impacts, achieve dynamic compensation and ensure long-term stability. Summary of the invention

[0005] The present invention provides an IMU dynamic calibration and compensation method, which aims to accurately obtain IMU raw data, effectively deal with environmental influences, realize dynamic compensation and ensure long-term stability.

[0006] To achieve the above purpose, the IMU dynamic calibration and compensation method proposed in the present invention is applied to a vehicle-mounted navigation device, comprising:

[0007] Obtain the raw data from the IMU, sample the data at the standard sampling frequency, and integrate it with the environmental parameters to obtain the initial measurement sequence;

[0008] Performing periodic decomposition on the initial measurement sequence, calculating the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and obtaining a measurement supplement sequence through parameter association;

[0009] Performing front-end calibration, middle-end calibration and end-end calibration on the measurement supplementary sequence, and obtaining a calibration reference value through parameter iteration;

[0010] According to the calibration reference value, analyzing the error variation law of different working conditions, and obtaining the calibration mapping amount through static and dynamic characteristic analysis;

[0011] Decomposing environmental factors using the calibration mapping amount to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor;

[0012] Dynamically test each compensation factor, combine it with the characteristics of environmental changes, and obtain the correction coefficient through threshold comparison;

[0013] Using the correction coefficient to perform compensation processing on the original data, and obtaining a correction sequence through data reconstruction;

[0014] Performing stability analysis on the correction sequence and obtaining drift parameters according to the change trend;

[0015] The drift parameters are differentially analyzed with the calibration reference, and the final calibration result is obtained through multi-point correction.

[0016] The present invention also proposes an IMU dynamic calibration and compensation device, which is applied to a vehicle-mounted navigation device, comprising:

[0017] The acquisition module is used to obtain the raw data of the IMU, sample the data according to the standard sampling frequency, and integrate it with the environmental parameters to obtain the initial measurement sequence;

[0018] A first decomposition module is used to periodically decompose the initial measurement sequence, calculate the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and obtain a measurement supplement sequence through parameter association;

[0019] A calibration module, used for performing front-end calibration, middle-end calibration and end-end calibration on the measurement supplementary sequence, and obtaining a calibration reference value through parameter iteration;

[0020] A first analysis module is used to analyze the error variation rules of different working conditions according to the calibration reference value, and obtain the calibration mapping amount through static and dynamic feature analysis;

[0021] A second decomposition module is used to decompose environmental factors using the calibration mapping amount to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor;

[0022] The inspection module is used to dynamically inspect each compensation factor and obtain the correction coefficient by comparing the threshold value in combination with the characteristics of environmental changes;

[0023] A correction module, used to use the correction coefficient to perform compensation processing on the original data, and obtain a correction sequence through data reconstruction;

[0024] A second analysis module is used to perform stability analysis on the correction sequence and obtain a drift parameter according to a change trend;

[0025] The output module is used to perform difference analysis between the drift parameter and the calibration reference, and obtain the final calibration result through multi-point correction.

[0026] The present invention also proposes a vehicle-mounted navigation device, including an IMU sensor, the vehicle-mounted navigation device also including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program implements the steps of the above-mentioned IMU dynamic calibration and compensation method when executed by the processor.

[0027] The present invention also proposes a storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned IMU dynamic calibration and compensation method are implemented.

[0028] The technical solution of the IMU dynamic calibration and compensation method of the present invention can obtain key parameters such as temperature gradient field, vibration transfer characteristics and magnetic field spatial distribution in real time by establishing a complete process of IMU raw data acquisition and dynamic calibration, combined with innovative methods of periodic decomposition and environmental parameter monitoring; this data acquisition solution based on multi-sensor collaboration significantly improves the accuracy of environmental factor identification, making subsequent compensation processing more targeted, and effectively overcomes the technical defects of traditional acquisition methods in insufficient identification of environmental factors. In addition, a multi-level compensation mechanism is innovatively proposed, by respectively establishing temperature compensation factors, vibration compensation factors and magnetic field compensation factors, and combining dynamic inspection methods to optimize correction coefficients, a complete error compensation system is formed; this compensation mechanism can independently compensate for the influence characteristics of different environmental factors, greatly improving the measurement accuracy of IMU in complex environments. In addition, in terms of drift parameter extraction and multi-point correction, a feature extraction method based on time scale decomposition is designed. By establishing a combined drift parameter model and a multi-point correction mechanism, the long-term stability of the IMU output is achieved. The technical solution of the present invention fully considers the influence of initial drift, linear drift and periodic drift, and ensures the reliability of the calibration results through cyclic verification under different working conditions, thereby solving the stability problem of the IMU in long-term operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic diagram of a flow chart of an embodiment of an IMU dynamic calibration and compensation method of the present invention;

[0030] Figure 2 for Figure 1 A schematic diagram of a flow chart of an embodiment of step S20 in the embodiment;

[0031] Figure 3 for Figure 1A schematic diagram of a flow chart of an embodiment of step S30 in the embodiment;

[0032] Figure 4 for Figure 1 A schematic diagram of a flow chart of an embodiment of step S40 in the embodiment;

[0033] Figure 5 for Figure 1 A schematic diagram of a flow chart of an embodiment of step S60 in the embodiment;

[0034] Figure 6 for Figure 1 A schematic diagram of a flow chart of an embodiment of step S70 in the embodiment;

[0035] Figure 7 for Figure 1 A schematic diagram of a flow chart of an embodiment of step S80 in the embodiment;

[0036] Figure 8 A schematic diagram of the module structure of an embodiment of an IMU dynamic calibration and compensation device of the present invention;

[0037] Fig. 9 It is a schematic diagram of the structure of an electronic device in a hardware operating environment involved in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] The present invention proposes an IMU dynamic calibration and compensation method, which is applied to a vehicle-mounted navigation device. The implementation terminal of the IMU dynamic calibration and compensation method of the present invention can be a vehicle-mounted navigation device or an electronic device, and the electronic device can be a controller of the vehicle-mounted navigation device, or a computing device such as a single-chip microcomputer, a desktop computer, a notebook, a palm computer, and a server.

[0040] like Figure 1 As shown, Figure 1 It is a flow chart of the first embodiment of the IMU dynamic calibration and compensation method of the present invention.

[0041] In this embodiment, the IMU dynamic calibration and compensation method includes:

[0042] Step S10, obtaining the raw data of the IMU, sampling the data according to the standard sampling frequency, and integrating it with the environmental parameters to obtain an initial measurement sequence.

[0043] The first thing is to obtain the raw data of the IMU (i.e. the raw output signal of the IMU). The process first fixes the IMU in the static area of ​​the test site. The installation position is selected close to the center of gravity of the vehicle, and the IMU measurement axis is aligned with the vehicle coordinate system through a special bracket. The IMU is connected to the data acquisition system through a shielded cable, and the acquisition system is equipped with a complete signal conditioning circuit at the front end. The signal conditioning circuit includes a two-stage amplification and low-pass filtering design. The first stage uses a precision instrumentation amplifier to amplify the weak signal output by the IMU, and the second stage uses an operational amplifier for signal conditioning and level conversion. The low-pass filter circuit is used to suppress high-frequency interference and ensure the effective bandwidth of the signal. The acquisition system records the three-axis angular velocity signal and the three-axis acceleration signal output by the IMU respectively, and ensures that the signal amplitude is within the optimal working range of the sampling system through reasonable signal conditioning. The signal conditioning circuit uses an independent power supply and shielding measures to effectively isolate the influence of external interference.

[0044] The acquisition system samples the IMU output signal at the standard sampling frequency. The system uses a temperature-compensated crystal oscillator to generate a reference clock, and generates the required sampling clock signal through a frequency division circuit. In each sampling cycle, the system synchronously acquires the six measurement axis signals of the IMU through a sample-and-hold circuit. After the sampled data is converted into digital quantities by a high-precision A / D converter, it is stored in the system cache according to a preset format. The system sets up a data cache area with sufficient capacity, uses a circular cache mechanism to organize the data flow, and ensures the continuity of the acquisition process through a data batch transfer mechanism. In order to improve data reliability, the system sets up a multi-level verification mechanism during the data flow process.

[0045] Environmental parameter monitoring uses a special sensor array configuration to form a complete environmental monitoring subsystem. The temperature monitoring module uses two high-precision temperature sensors, one of which is mounted on the surface of the IMU device and fixed by thermal conductive glue to ensure good thermal conductivity, which is used to accurately monitor the operating temperature of the IMU chip. The other temperature sensor is installed at the preset position of the bracket, keeping a certain distance from the IMU, and is specifically used to monitor changes in ambient temperature. The signal of the temperature sensor is processed by a special conditioning circuit, including constant current source excitation, signal amplification and filtering circuit. Considering the characteristics of temperature changes, the temperature signal uses a lower sampling frequency, which effectively reduces the burden of data storage. Vibration monitoring uses a highly sensitive three-axis acceleration sensor. The sensor is fastened to the IMU bracket with special bolts, and the installation position is optimized to ensure the accurate transmission of vibration signals. The signal conditioning circuit of the sensor includes a charge amplifier, signal conditioning and anti-aliasing filter. The acquisition system uses the same sampling frequency as the IMU for the vibration signal, ensuring the synchronization of vibration data and IMU data. Magnetic field monitoring uses a three-axis magnetometer, and the sensor is arranged at the optimal position after magnetic field interference testing to monitor changes in the local magnetic field environment. The signal conditioning of the magnetometer includes fluxgate drive circuit, signal amplification and low-pass filtering. The signal conditioning circuits of all environmental sensors adopt unified design standards, including power isolation, signal shielding and common mode suppression.

[0046] The system integrates all sampled data to form a complete initial measurement sequence. During the entire acquisition process, the system completes a complete data record in each sampling cycle. Each data record contains three parts: a unified timestamp, IMU six-axis measurement data, and environmental parameter data. The timestamp is generated by a high-precision timer to accurately mark the moment of each data acquisition. The IMU data part records the sampling values ​​of the three-axis angular velocity and three-axis acceleration. All data are stored in engineering units after calibration and conversion. The environmental parameter part contains temperature, vibration, and magnetic field data.

[0047] Step S20, periodically decompose the initial measurement sequence, calculate the impact of temperature change, vibration change, and magnetic field change on the IMU output value, and obtain a measurement supplement sequence through parameter association.

[0048] Combination Figure 2 As shown, in some embodiments, step S20 may include: step S21, step S22, step S23, step S24 and step S25. Among them:

[0049] Step S21, setting the analysis window length and sliding step size according to the signal characteristics of the IMU. In each analysis window, performing fast Fourier transform on the angular velocity and acceleration signals to obtain the spectral characteristics of the signal, and identifying the main frequency components in the signal by calculating the power spectral density.

[0050] For example, when the initial measurement sequence is periodically decomposed, the system sets the analysis window length to 10 seconds (of course, it can also be 15 seconds, 20 seconds, etc. in other embodiments) and the sliding step length to 2 seconds (of course, it can also be 1 second, 3 seconds, etc. in other embodiments) according to the IMU signal characteristics. In each analysis window, the angular velocity and acceleration signals are subjected to a fast Fourier transform to obtain the spectral characteristics of the signal. By calculating the power spectral density, the main frequency components in the signal are identified. These frequency components are manifested as low-frequency drift of the angular velocity signal in the range of 0.01-0.1Hz, and resonant response in the range of 20-50Hz. The acceleration signal is mainly concentrated in the mid-frequency band of 50-200Hz, with obvious second harmonics in the 80-120Hz band.

[0051] Step S22, based on the spectrum characteristics and main frequency components, calculate and analyze the correlation between temperature change, vibration change, magnetic field change and IMU output value.

[0052] Based on the results of spectrum analysis, the system further conducts an in-depth analysis of the impact of temperature changes. By comparing the output difference between the IMU internal temperature sensor and the ambient temperature sensor, the temperature gradient field distribution is calculated. When the temperature change rate exceeds 0.5°C / min, the IMU output will produce obvious periodic drift. This drift is manifested as a 180-second periodic change in the angular velocity output, and its amplitude is positively correlated with the temperature change rate. Detailed analysis found that every 1°C change in temperature will cause a change of about 0.02° / s in the angular velocity zero bias, and the acceleration output is manifested as a periodic change in the sensitivity coefficient, with a change amplitude of about 0.1% of the full scale.

[0053] Then the system analyzes the impact of vibration. The wavelet decomposition method is adopted, and the db4 wavelet basis function is used to decompose the signal into 5 layers. By reconstructing the coefficients of each layer, it is clearly shown that the IMU has significant resonance characteristics in the 20-50Hz frequency band, and the amplitude amplification factor is as high as 1.8. Especially in the acceleration output, the second harmonic generated in the 80-120Hz frequency band shows obvious nonlinear characteristics, and its amplitude shows a quadratic function relationship with the vibration intensity. This nonlinear characteristic is coupled with the previously discovered temperature periodicity effect. When the temperature rises, the vibration sensitivity will increase accordingly, which is manifested as a change in the amplification factor.

[0054] While understanding the influence of temperature and vibration, the system analyzes the influence of magnetic field. By calculating the spatial distribution of magnetic field gradient, it is found that when the magnetic field gradient exceeds 0.2Gs / m, the angular velocity output of IMU will produce synchronous fluctuations. This fluctuation is most obvious in heading angle measurement, and its amplitude is not only proportional to the rate of change of magnetic field, but also correlated with the vibration characteristics identified previously. The cross-correlation analysis of the magnetic field change vector and the IMU output shows that there is a time delay of about 5ms in magnetic field interference. This delay feature remains relatively stable under different temperature conditions, which makes real-time compensation possible.

[0055] Step S23, based on the correlation between temperature change, vibration change, magnetic field change and IMU output value, construct a correlation matrix between environmental parameters and IMU output, the row vectors of the correlation matrix represent temperature values, three-axis vibration values, and three-axis magnetic field values, the column vectors of the correlation matrix represent the three-axis angular velocity and three-axis acceleration of the IMU, calculate the influence coefficient of temperature on zero bias, the influence coefficient of temperature change rate, the gain coefficient of conduction function, and the sensitivity coefficient of magnetic field interference in the correlation matrix.

[0056] Based on the analysis results of the above environmental factors (temperature, vibration, magnetic field), the system constructs the correlation matrix between environmental parameters and IMU output. The row vector of the matrix represents the temperature value T, the three-axis vibration value Vx / Vy / Vz, and the three-axis magnetic field value Mx / My / Mz, and the column vector represents the three-axis angular velocity ωx / ωy / ωz and the three-axis acceleration ax / ay / az of the IMU. Each matrix element contains both direct influence coefficients and cross-influence coefficients. The system solves by the least squares method, first obtaining the influence coefficient k1 of temperature on zero bias, which describes the linear relationship between the temperature change and the zero bias change; the influence coefficient k2 of the temperature change rate, which characterizes the degree of influence of the temperature change rate on the zero bias. For the influence of vibration, the gain coefficient kv of the transfer function is solved, which represents the amplification factor of the IMU output under different frequency vibrations and reflects the time delay characteristics in the vibration transmission process. In terms of magnetic field, the sensitivity coefficient km of magnetic field interference is calculated to quantify the degree of influence of magnetic field intensity changes on IMU output.

[0057] Step S24, using each coefficient of the correlation matrix, performing item-by-item calculation on the IMU output to obtain the zero bias change caused by temperature, the output fluctuation caused by vibration, and the influence of magnetic field interference respectively;

[0058] Using these influence coefficients, the system performs item-by-item calculations on the IMU output. In the temperature compensation phase, the temperature change ΔT is substituted into the formula δT = k1·ΔT + k2·dT / dt to calculate the zero bias change δT caused by temperature.

[0059] Vibration compensation adopts the frequency domain calculation method. The vibration time domain signal v(t) is converted to the frequency domain through Fourier transform to obtain V(f), where f represents the vibration frequency. It is then multiplied by the transfer function H(f) = kv·ejφv, where kv is the vibration sensitivity coefficient, which represents the amplification factor of the IMU output under different frequency vibrations, and ejφv is the complex exponential term represented by the Euler formula (e is the base of the natural logarithm, j is the imaginary unit, and φv is the phase angle), which is used to describe the phase delay characteristics in the vibration transmission process, and finally the output fluctuation δv caused by the vibration is obtained.

[0060] Magnetic field compensation is based on the magnetic field gradient And the sensitivity coefficient km calculates the influence of magnetic field interference These sub-item calculation results are compared with the frequency characteristics obtained by period decomposition to verify the correctness of the calculation results. At the same time, the system also takes into account the coupling effect between parameters during the calculation process: temperature changes cause the gain coefficient of the vibration transfer function to change, which is manifested as the correction of kv with temperature; the magnetic field sensitivity coefficient in a strong vibration environment needs to be adjusted according to the vibration intensity, which is reflected in the dynamic correction of km.

[0061] Step S25: generating a supplementary measurement sequence corresponding to the time point in the initial measurement sequence according to the sub-item calculation results, wherein the supplementary measurement sequence records the data at each sampling moment.

[0062] As various analyses and calculations are completed, the system begins to generate a supplementary measurement sequence. This sequence strictly corresponds to the time points in the initial measurement sequence and records the data at each sampling moment. The angular velocity and acceleration signal characteristics obtained through spectrum analysis and wavelet decomposition constitute the basic data of the sequence. Based on the results of temperature change analysis, vibration impact analysis, and magnetic field interference analysis, the system uses the established environmental parameters and IMU output correlation matrix to calculate the impact of various environmental factors on the IMU output.

[0063] Step S30, performing front-end calibration, middle-end calibration and end-end calibration on the measurement supplementary sequence, and obtaining a calibration reference value through parameter iteration.

[0064] The measurement supplement sequence obtained in step S20 is calibrated in sections, and the front section calibration, middle section calibration and final section calibration methods are used to obtain the calibration reference value through actual measurement under different working conditions. This section calibration method is based on the progressive design of the complexity of the test environment, from the analysis of a single environmental factor in a static state, to a stable dynamic environment of uniform motion, and finally to a multi-factor coupling scenario of actual working conditions, realizing a scientific calibration process from simple to complex, from controllable to comprehensive.

[0065] Combination Figure 3As shown, in some embodiments, step S30 may include: step S31, step S32, step S33, step S34 and step S35. Among them:

[0066] Step S31 , collecting front-end calibration data in a stationary area of ​​the test site, collecting middle-end calibration data on a standard loop, and collecting final-end calibration data on a working road.

[0067] First, start the front-end calibration data collection in the static area of ​​the test site. During this stage, the vehicle remains completely stationary, and the system collects the output data of the IMU according to the preset temperature points. The vehicle is fixed on a specially designed test platform, and the IMU is ensured to be in a horizontal position through the leveling system. Starting from the reference temperature of 25°C, the system records the initial zero bias value Z0, and then gradually increases the temperature at a rate of 0.5°C / min through the environmental control equipment. When the temperature reaches 30°C, wait for 10 minutes for the temperature field to stabilize and then record the zero bias value Z1. Continue to heat up to 35°C to record the zero bias value Z2, and finally record the zero bias value Z3 at 40°C. At each temperature point, the system not only records the six-axis output data of the IMU, but also synchronously collects the readings of the IMU internal temperature sensor and the external ambient temperature sensor to ensure the uniformity of the temperature field. These temperature-zero bias pairs constitute a detailed measured data set for establishing the actual impact of temperature on the IMU output. At the same time, the system uses a high-precision magnetometer to record the magnetic field environment parameters in a stationary state, including the output deviation M0 at the magnetic field intensity B0 (range 0.2-1.0Gs), and the background noise level N0 measured under the environmental background vibration V0 (frequency range 0-200Hz) using a high-sensitivity accelerometer.

[0068] The mid-section calibration data collection is carried out on a standard loop. The system uses a circular test track with a radius of 200 meters, and the road surface is paved with asphalt materials to ensure that the road surface flatness meets the test requirements. The calibration test is carried out at a constant speed according to the preset speed level. First, under the low-speed condition of 40km / h, the steady-state output response R1 of the IMU is recorded after the vehicle speed stabilizes, including the output values ​​of each axis of angular velocity and acceleration. The system simultaneously records auxiliary parameters such as engine speed, wheel speed and suspension displacement to evaluate the stability of vehicle motion. Then, the acceleration is slowly increased to the medium-speed condition of 60km / h at an acceleration of 2km / h / s, and the output response R2 is obtained after the speed stabilizes. Finally, it is increased to the high-speed condition of 80km / h in the same way and the output response R3 is recorded. The test at each speed level lasts for 5 minutes, and the system records a complete set of motion parameters including angular velocity ω, acceleration a, and attitude angle θ. At the same time, the vibration sensor installed on the IMU bracket recorded in detail the frequency characteristics of the engine vibration at different speeds (mainly distributed in the 20-100 Hz frequency band) and the output fluctuations of the IMU, and established the speed-vibration-output correlation characteristics.

[0069] The final calibration data collection was transferred to the working road. The test road contains a variety of typical road conditions, and a comprehensive test route with a total length of 2 kilometers was designed. In the acceleration and deceleration section, through precise throttle and brake control, the output value O1 under 0.3g acceleration and the output value O2 under 0.5g acceleration were recorded respectively. Each acceleration condition was tested 5 times to ensure the reliability of the data. In the turning section, standard curves with radii of 20 meters and 50 meters were designed respectively. The system collected the output values ​​O3 and O4 under these two turning radii respectively, and recorded key parameters such as steering angular velocity and lateral acceleration. Finally, in the ramp section, standard ramps with slopes of 5% and 10% were used to record the output values ​​O5 and O6 of the IMU during the climbing process, while monitoring the changes in the vehicle pitch angle and longitudinal acceleration. Under each working condition, the system synchronously recorded environmental parameters such as ambient temperature, vibration intensity and magnetic field intensity, providing complete data support for subsequent comprehensive analysis.

[0070] Step S32, establishing an IMU output error data model based on the front-end calibration data, the middle-end calibration data and the final-end calibration data;

[0071] Step S33, for each sampling point, calculating the error vector between the measured output value and the theoretical calculated value, where the theoretical calculated value is calculated based on the kinematic model;

[0072] Step S34, constructing a measurement matrix, where each row of the measurement matrix corresponds to an observation, and each column corresponds to the coefficients of temperature, vibration, and magnetic field, and using the error vector to perform parameter optimization by the least squares method to calculate the parameter correction amount;

[0073] Step S35, updating the parameters of the error data model according to the parameter correction amount and continuing the iterative optimization until the parameter change amount is less than a preset threshold value, thereby obtaining a calibration reference value.

[0074] Based on the collected multiple sets of measured data, the system established a complete IMU output error mathematical model E = f(T, V, B). E represents the total error vector, which includes the angular velocity error Eω and the acceleration error Ea; T represents the temperature influence term, which takes into account the influence of the absolute value of temperature and the temperature change rate; V represents the vibration influence term, which includes two components: vibration frequency and vibration intensity; B is the magnetic field influence term, which takes into account the combined effect of magnetic field intensity and gradient change. For each sampling point, the system calculates the error vector E between the measured output value and the theoretical calculated value, where the theoretical calculated value is calculated based on the kinematic model. To optimize the model parameters, the system constructs a measurement matrix H, where each row of the matrix corresponds to an observation and each column corresponds to the coefficient of influencing factors such as temperature, vibration, and magnetic field.

[0075] Using the error vector E, the system first constructs the optimization target of the least squares method, that is, the cost function J=ETE. This cost function represents the sum of the squared differences between the actual measured error and the model prediction error. The smaller its value, the more accurate the model prediction. The cost function is optimized by the least squares method, and the parameter correction ΔP = (HTH)-1HTE is calculated, where HT is the transposed matrix of H. Here, HTE reflects the degree of correlation between each influencing factor and the actual measured error, while (HTH)-1HTE gives the optimal parameter correction when the cost function is minimized, so that the error predicted by the model can better fit the actual measured error E. The parameter update adopts an iterative method. The model parameter P_new after each update is equal to the parameter value P_old before the update plus the correction amount ΔP of this time. The iteration process continues until the parameter change is less than the preset threshold (for example, usually 0.1% of the value of the previous iteration).

[0076] After multiple rounds of iterative optimization, the system finally obtains a complete set of calibration reference values. The temperature compensation reference value includes the temperature-zero bias correction coefficient KT0 (in rad / s / ℃, with a typical value of 0.010.05rad / s / ℃) and the temperature change rate correction coefficient KT1 (in rad / s / ℃ / min, with a typical value of 0.0020.01rad / s / ℃ / min). The vibration compensation reference value includes the vibration amplitude correction coefficient KV0 (dimensionless, range 0.81.2) and the vibration frequency response coefficient KV1 (in Hz-1, range 0.0010.005Hz-1). The magnetic field compensation reference value includes the magnetic field strength correction coefficient KB0 (in rad / s / Gs, with a typical value of 0.050.2rad / s / Gs) and the magnetic field gradient correction coefficient KB1 (in rad / s / Gs / m, with a typical value of 0.010.05rad / s / Gs / m). Among them, the coefficients (KT0, KT1, KV0, KV1, KB0, KB1) are the "compensation coefficients" obtained through actual calibration, which are the refinement and expansion of the coefficients of step S20: KT0 corresponds to k1, but is optimized through measured data; KT1 corresponds to k2, which is also optimized; KV0 and KV1 are the refinements of kv, taking into account the effects of vibration amplitude and frequency respectively; KB0 and KB1 are the refinements of km, taking into account the effects of magnetic field strength and gradient respectively.

[0077] The nominal value, measured value, effective range and confidence level of each benchmark value are determined through repeated tests and statistical analysis to ensure the reliability and applicability of the calibration parameters.

[0078] Step S40: Analyze the error variation rules of different working conditions according to the calibration reference value, and obtain the calibration mapping value through static and dynamic characteristic analysis.

[0079] Based on the calibration reference value, through static and dynamic feature analysis, the system establishes a hierarchical calibration mapping volume, which is divided into two levels: static mapping and dynamic mapping based on the characteristics of environmental factors. Static mapping mainly deals with the influence of slow-changing environmental factors such as temperature, while dynamic mapping targets the influence of fast-changing factors such as vibration and motion characteristics.

[0080] Combination Figure 4 As shown, in some embodiments, step S40 may include: step S41, step S42, step S43 and step S44. Among them:

[0081] Step S41, determining the temperature influence characteristics of the temperature change on the IMU output based on the measured temperature-bias correction coefficient and the temperature change rate correction coefficient;

[0082] Step S42, based on the test results of the vibration amplitude correction coefficient and the vibration frequency response coefficient, recording the vibration response characteristics at different vehicle speeds;

[0083] Step S43, based on the test data of the magnetic field intensity correction coefficient and the magnetic field gradient correction coefficient, record the magnetic field interference effect of the magnetic field change on the IMU output.

[0084] The system first conducts an in-depth analysis of the temperature calibration data collected in the static area, and establishes the characteristics of the impact of temperature changes on the IMU output based on the measured temperature-bias correction coefficient KT0 and temperature change rate correction coefficient KT1. For example, in the range of 25-40°C, the zero bias change caused by every 1°C change in temperature recorded by KT0 and the dynamic error caused by the temperature change rate reflected by KT1 form a complete temperature response characteristic curve. These characteristic data are organized into a temperature-error mapping table, which records the zero bias output characteristics of the IMU at different temperature points. At the same time, the system also establishes a temperature change rate-dynamic error mapping relationship to describe the transient response characteristics of the IMU when the temperature changes rapidly.

[0085] The vibration characteristics analysis is carried out on a standard loop track, which uses a standard asphalt road surface and has a radius of 200 meters. The vibration test starts from a stationary state, first under idling conditions, and then under uniform speed tests at different vehicle speeds. The system collects vibration data through a high-precision vibration sensor and records the output response of the IMU under various vibration conditions. Based on the test results of the vibration amplitude correction coefficient KV0 and the vibration frequency response coefficient KV1, the system establishes a complete vibration response characteristic map. Under idling conditions, the engine vibration produces a fundamental frequency vibration of 20-30Hz, and KV0 shows that the output amplification ratio is in the range of 0.8-0.9. When the vehicle speed increases to 60km / h, the vibration frequency expands to a range of 30-40Hz, and KV0 increases to 0.9-1.0. Under high-speed conditions such as 80km / h, the vibration frequency band is further expanded to 40-50Hz, and KV0 reaches a level of 1.0-1.2. Through the measurement results of KV1, the system found that the IMU exhibits obvious frequency selection characteristics under vibrations of different frequencies. In particular, there is a resonance point in the frequency range of 35-45Hz, where the output amplification is most obvious. In addition, the vibration generated by the transmission system is closely related to the vehicle speed. For example, the imbalance of the drive shaft causes mechanical vibrations of 30-50Hz, and the meshing of the transmission gears produces high-frequency vibrations of more than 100Hz. The vibrations generated by road excitation are more complex. Flat roads mainly excite low-frequency vibrations such as 10-30Hz. These vibrations are transmitted to the sensitive elements through the IMU mounting bracket. Due to the inherent frequency characteristics of the MEMS structure, vibrations of different frequencies will cause different degrees of response.

[0086] The system uses a high-precision magnetometer to record magnetic field environment parameters and study the influence of magnetic field changes on IMU output. In a stationary state, the response characteristics of the IMU under different magnetic field strength conditions are recorded. When the magnetic field strength changes in the range of 0.2-1.0Gs, the change characteristics of the IMU output are recorded through the magnetic field strength correction coefficient KB0 (such as 0.05-0.2rad / s / Gs). At the same time, the system also studies the influence of the magnetic field gradient, and describes the influence of the uneven spatial distribution of the magnetic field on the IMU through the magnetic field gradient correction coefficient KB1 (such as 0.01-0.05rad / s / Gs / m). In actual tests, special attention is paid to the response characteristics of the IMU when the magnetic field changes rapidly. When the magnetic field gradient exceeds 0.2Gs / m, the angular velocity output of the IMU will produce synchronous fluctuations. This fluctuation is most obvious in the heading angle measurement, and its amplitude is not only proportional to the magnetic field change rate, but also associated with the vibration characteristics. By analyzing the cross-correlation between the magnetic field change vector and the IMU output, it is found that the magnetic field interference has a time delay of 5ms, and this delay feature remains relatively stable under different temperature conditions.

[0087] Step S44: Based on the temperature influence characteristics, the vibration response characteristics and the magnetic field interference influence, a hierarchical structure including a static mapping layer and a dynamic mapping layer is established to form a complete calibration mapping amount.

[0088] Through the systematic analysis of these characteristics, a hierarchical calibration mapping volume was established. The first layer is the static mapping layer, which mainly contains the temperature mapping relationship. The temperature response characteristic curve is constructed using the temperature-zero bias correction coefficient KT0 and the temperature change rate correction coefficient KT1, and the temperature range of 25°C to 40°C is discretized into 15 temperature intervals, each with 1°C. For example, in the 25-26°C interval, the zero bias change value caused by the temperature increase was recorded as 0.018° / s; in the 35-36°C interval, due to the increase in temperature sensitivity, the recorded zero bias change value increased to 0.022° / s. At the same time, the influence of the temperature change rate is also recorded at each temperature point. For example, when the temperature changes at a rate of 0.5°C / min, an additional dynamic error of 0.002° / s will be generated.

[0089] The second layer is the dynamic mapping layer, which integrates vibration and magnetic field characteristics. For vibration characteristics, the system establishes a two-dimensional mapping table covering frequency ranges such as 20-100Hz and amplitude ranges such as 0.1-2.0g. The table records the output corrections under different working conditions. For example, at a frequency of 40Hz and an amplitude of 0.5g, the output needs to be multiplied by a correction factor such as 0.92, and at a frequency of 80Hz and an amplitude of 1.0g, due to the resonance effect, the correction factor increases to 1.15. When recording vibration characteristics, special attention was paid to the frequency selection characteristics of the IMU under vibrations of different frequencies, especially in the range of 35-45Hz, where there are obvious resonance points, and the output amplification is most significant.

[0090] For magnetic field characteristics, the system constructs a mapping structure for the influence of magnetic field environment. The influence of uniform magnetic field is recorded by magnetic field strength correction coefficient KB0 (such as 0.05-0.2rad / s / Gs), and the influence of magnetic field gradient is described by magnetic field gradient correction coefficient KB1 (such as 0.01-0.05rad / s / Gs / m). In actual tests, when the magnetic field strength changes from 0.2Gs to 1.0Gs, the system observes obvious deviations in heading angle measurement. In particular, when the magnetic field gradient exceeds 0.2Gs / m, the angular velocity output of the IMU will produce synchronous fluctuations, and this fluctuation characteristic is related to vibration. Through the analysis of magnetic field changes and IMU output, it is found that magnetic field interference has a time delay characteristic of 5ms. In terms of parameter coupling, the system finds that temperature changes will affect the vibration transfer characteristics, and the vibration correction coefficient needs to be adjusted when the temperature rises. Similarly, in a strong vibration environment, the magnetic field sensitivity coefficient also needs to be corrected according to the vibration intensity. This parameter coupling processing mechanism improves the compensation accuracy of the system in complex environments.

[0091] The final calibration mapping is a multi-dimensional data set. For any working state, the system can query or interpolate the corresponding error correction value from the mapping table based on the current temperature, vibration, magnetic field and other parameters. This mapping mechanism based on measured data provides a complete error compensation basis for the high-precision application of IMU. In this embodiment, the calibration data is organized in the form of a mapping table, so that the system can respond quickly to environmental changes and provide accurate error compensation values ​​in real time, which effectively improves the measurement accuracy of the IMU under various working conditions.

[0092] Step S50, using the calibration mapping amount to decompose the environmental factors to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor.

[0093] Based on the calibration mapping established in step S40, the system uses the correspondence between the IMU output data and the environmental parameters recorded in the static mapping layer and the dynamic mapping layer for analysis. From the correspondence between the IMU output data and the environmental parameters recorded in detail in the calibration mapping, it can be found that the output error of the IMU is significantly affected by three factors: temperature, vibration, and magnetic field. In the actual vehicle environment, these three factors often exist at the same time. For example, when a vehicle is driving at high speed in hot summer weather, it will face the influence of increased temperature, increased vibration, and magnetic field interference. Therefore, it is necessary to systematically decompose and analyze each influencing factor and establish a corresponding compensation factor model.

[0094] First, the temperature effect is decomposed. Temperature acts on IMU and affects measurement accuracy mainly through two channels. The first channel is that the absolute value change of temperature causes the thermal stress change of MEMS structure, resulting in zero drift. This static temperature effect behaves differently under different working conditions: when the vehicle is started for the first time after being exposed to the sun in summer, the temperature of the IMU shell may be as high as 50°C or more, causing significant thermal deformation of the internal silicon structure and the maximum zero drift; in cold winter, low temperature will also affect the measurement accuracy by changing the material properties; in addition, the operation of the air-conditioning system will also cause periodic fluctuations in the temperature around the IMU, causing slow drift of the measured value. The second channel is that the rate of change of temperature produces a temperature gradient inside the device, causing dynamic errors. This dynamic temperature effect will occur in many scenarios: when the vehicle enters the hot outdoor environment from the air-conditioned environment, the IMU may experience a temperature change rate of 5-10°C / min; the ambient temperature difference when entering and leaving the underground garage will cause transient temperature distribution inside the IMU; even local temperature fluctuations caused by engine start and stop will affect the stability of the measurement.

[0095] The static mapping layer records the zero bias characteristics of the IMU at different temperatures in detail, while the dynamic mapping layer records the transient response characteristics corresponding to the temperature change rate. Based on these mapping data, a temperature compensation factor model is established: CT = α1·T + α2·dT / dt, where T is the real-time temperature value (℃), reflecting the thermal stress effect caused by the absolute value change of temperature, such as the IMU structure deformation caused by exposure and the zero drift caused by high temperature; dT / dt is the temperature change rate (℃ / min), describing the transient response caused by rapid temperature changes, such as the temperature gradient generated inside the IMU when the vehicle enters and exits the underground garage; α1 and α2 are compensation coefficients, corresponding to the static and dynamic effects of temperature, respectively. This model structure reflects the two main characteristics of temperature effects: the absolute value of temperature mainly compensates for the steady-state error caused by the thermal deformation of the MEMS structure; the temperature change rate term handles the transient error caused by drastic temperature fluctuations.

[0096] Then the influence characteristics of vibration are analyzed. There are many vibration sources in the vehicle environment, and each vibration source has its specific frequency characteristics and transmission rules. The engine vibration mainly comes from the reciprocating motion of the cylinder when it is working. It produces a fundamental frequency vibration of 20-30Hz at idle speed, which can reach 50-60Hz as the speed increases, accompanied by harmonic components. The transmission system will produce vibration related to the vehicle speed when it is working, such as the imbalance of the transmission shaft causing mechanical vibration of 30-50Hz, and the high-frequency vibration generated by the meshing of the transmission gear can reach more than 100Hz. Road excitation produces a more complex vibration spectrum. The flat road surface mainly excites low-frequency vibration of 10-30Hz, while passing over speed bumps or manhole covers will produce transient impact vibrations with a frequency range of up to 200Hz. These vibrations are transmitted to the sensitive element through the IMU mounting bracket. Due to the inherent frequency characteristics of the MEMS structure, vibrations of different frequencies will cause different degrees of response. In particular, when the vibration frequency is close to the natural frequency of the IMU sensitive structure (usually in the range of 50-100Hz), resonance will occur, and even a small vibration amplitude may cause significant measurement errors.

[0097] The dynamic mapping layer records the influence of vibrations in each frequency band on the IMU output, including amplitude amplification characteristics and frequency response characteristics. Based on these mapping relationships, the vibration compensation factor model CV = β1·V(f) + β2·A is established, where V(f) represents the frequency characteristics of vibration (Hz), reflecting the frequency distribution of different vibration sources from the engine, transmission system to road excitation, especially the frequency band where the IMU sensitive structure is prone to resonance (such as 50-100Hz); A is the vibration amplitude (m / s²), which describes the change in vibration intensity, covering from weak vibration at idle to impact vibration when passing over speed bumps; β1 and β2 are compensation coefficients, corresponding to frequency response and intensity effects, respectively. This model structure comprehensively considers the two core characteristics of vehicle-mounted vibration: the frequency component V(f) mainly deals with the frequency selectivity of various vibration sources, such as engine fundamental frequency vibration and road excitation; the amplitude characteristic A reflects the direct impact of vibration intensity on measurement.

[0098] The analysis of the influence of magnetic field focuses on its spatial distribution characteristics. The sources of magnetic field in the vehicle operating environment are very complex, mainly including three categories: one is the working magnetic field generated by the vehicle's own electrical system, such as the starter, generator, high-voltage line, etc.; the second is the fixed magnetic field generated by road infrastructure, such as reinforced concrete bridges, underground pipelines, substation facilities, etc.; the third is the mobile magnetic field generated by other vehicles, especially large trucks, construction machinery, etc. The superposition of these magnetic field sources forms a complex spatial distribution, which is not only manifested as changes in local magnetic field intensity, but more importantly, it produces obvious spatial gradients. For example, when a vehicle passes through a steel structure overpass, the local magnetic field intensity may suddenly change from 0.2Gs to 1.0Gs, and a gradient of up to 0.5Gs / m is formed at the edge of the bridge; at the entrance of the underground garage, the superposition of the magnetic field of the steel mesh and underground facilities will produce a complex spatial distribution; even when waiting for the traffic light, the movement of adjacent large vehicles will cause changes in the local magnetic field environment. The impact of these magnetic field changes on the IMU is mainly reflected in the heading angle measurement, especially when the vehicle turns, the coupling of lateral acceleration and magnetic field changes will lead to greater measurement errors.

[0099] The static mapping layer records the basic influencing features under a uniform magnetic field environment, and the dynamic mapping layer reflects the influencing rules of magnetic field gradient changes. Based on these mapping features, a magnetic field compensation factor model is established: , where B is the magnetic field intensity (Gs), which represents the basic influence in a uniform magnetic field environment, and mainly reflects the systematic deviation of the heading angle measurement caused by the stable magnetic field around the vehicle; is the magnetic field gradient (Gs / m), reflecting the spatial rate of change of the magnetic field intensity, and describing the disturbance effect of the uneven local magnetic field distribution on the measurement; γ1 and γ2 are compensation coefficients, corresponding to the uniform field and gradient field effects, respectively. This model structure simultaneously considers two key features of the magnetic field environment: the static uniform magnetic field mainly affects the absolute measurement accuracy of the heading angle, such as the stable magnetic field generated by fixed metal facilities in the garage; the spatial gradient field causes dynamic deviations in the heading angle, such as the sudden change of the magnetic field when passing through an overpass or an underground passage.

[0100] These three compensation factors constitute a complete system of IMU error compensation. The temperature compensation factor CT comprehensively describes the impact of thermal effects on MEMS devices through the combination of static and dynamic terms; the vibration compensation factor CV introduces frequency response characteristics to accurately reflect the impact of different vibration sources in the vehicle environment; the magnetic field compensation factor CM effectively handles the interference of complex magnetic field environments by considering spatial distribution.

[0101] Step S60 , dynamically checking each compensation factor, combining the environmental change characteristics, and obtaining the correction coefficient through threshold comparison.

[0102] For the temperature compensation factor CT, vibration compensation factor CV and magnetic field compensation factor CM obtained in step S50, the system designs a progressive verification scheme from static environment to complex working conditions to determine the coefficients in each compensation factor.

[0103] Combination Figure 5 As shown, in some embodiments, step S60 may include: step S61, step S62, step S63, step S64 and step S65. Among them:

[0104] Step S61, performing a temperature compensation verification test in a static area of ​​the test site. During the verification test, the temperature compensation factor is applied in real time to perform compensation processing, and it is determined whether the coefficient of the temperature compensation factor needs to be adjusted according to a preset threshold requirement;

[0105] First, the temperature compensation verification is carried out in the static area of ​​the test site. In this stage, the temperature control system is used to create a stable and controllable temperature environment with a temperature control accuracy better than ±0.5℃. The system first stabilizes the ambient temperature within the range of 25±2℃ for 2 hours as the reference temperature segment, records a set of IMU output data every 1 minute, and obtains the zero bias reference value at the reference temperature. Then, the temperature is gradually increased to 40±2℃ at a steady rate of 0.5℃ / min and maintained for 1 hour, then cooled to 10±2℃ at the same rate and maintained for 1 hour, and finally returned to 25±2℃ to complete the entire temperature cycle verification. During the entire verification process, the system continuously collects the output data of the IMU, and the sampling frequency is maintained at 200Hz. At the same time, the surface temperature of the IMU chip and the ambient temperature are monitored by the temperature sensor. The system applies the temperature compensation factor CT =α1·T +α2·dT / dt in real time for compensation processing, where T is the real-time temperature value and dT / dt is the temperature change rate.

[0106] The system sets two sets of key performance indicators: static temperature compensation indicators and dynamic temperature compensation indicators. In terms of static temperature compensation, the system sets the static threshold TS of zero bias drift to ±0.02° / s. When the compensated zero bias drift exceeds TS for more than 1 minute, the system will mark the temperature point and need to adjust the compensation coefficient. In terms of dynamic temperature compensation, the system sets the dynamic error threshold TD to ±0.05° / s. When the compensated error during the temperature change process exceeds TD, the system will also record the operating point for subsequent correction.

[0107] Step S62, performing a vibration compensation verification test on a standard loop. During the verification test, the vibration compensation factor is applied in real time to process the IMU output data, and the real-time vibration characteristics are recorded. It is determined whether the coefficient of the vibration compensation factor needs to be adjusted according to a preset threshold requirement.

[0108] Then the vibration compensation verification is carried out on the standard loop. The loop adopts standard asphalt pavement, the road surface flatness meets the test requirements, and the radius of the loop is 200 meters. The vibration compensation verification starts from the stationary state of the vehicle, first maintaining the idle condition for 10 minutes, and then completing the uniform speed verification of 40km / h, 60km / h and 80km / h in turn. Each speed level is maintained for 15 minutes, and a slow acceleration and deceleration of 0.5m / s² is used between adjacent speed levels to achieve a smooth transition. The system applies the vibration compensation factor CV =β1·V(f) +β2·A in real time to process the IMU output data, and records the real-time vibration characteristics at the same time. The system sets the frequency threshold VF and the amplitude threshold VA as evaluation indicators. When the power spectrum density of the compensated signal in the 20-50Hz frequency band exceeds VF (40dB), the frequency response coefficient needs to be adjusted; when the acceleration residual after compensation exceeds VA (±5% of the nominal value), the amplitude coefficient needs to be corrected.

[0109] Step S63, performing a magnetic field compensation verification test on a working road. During the verification test, when passing through a preset magnetic field interference area, a magnetic field compensation factor is applied for real-time compensation, and it is determined whether the coefficient of the magnetic field compensation factor needs to be adjusted according to a preset threshold requirement;

[0110] Then, the magnetic field compensation verification is carried out on a specially designed working road. The verification route passes through several typical magnetic field interference areas, including reinforced concrete overpasses, underground garage entrances, substation areas, etc. When the vehicle passes through these areas, the system applies the magnetic field compensation factor. Real-time compensation is performed. The system sets the magnetic field strength threshold MS to ±0.5° and the gradient threshold MG to ±0.2° / s. When the magnetic field strength changes from 0.2Gs to 1.0Gs, if the compensated heading angle deviation exceeds MS, the magnetic field strength coefficient needs to be adjusted; when passing through an area with a 0.5Gs / m gradient, if the angular velocity output fluctuation exceeds MG, the gradient coefficient needs to be corrected.

[0111] Step S64, when it is determined that the coefficient of the temperature compensation factor needs to be adjusted, the temperature compensation correction coefficient is calculated using the temperature compensation correction formula; when it is determined that the coefficient of the vibration compensation factor needs to be adjusted, the vibration compensation correction coefficient is calculated using the vibration compensation correction formula; when it is determined that the coefficient of the magnetic field compensation factor needs to be adjusted, the magnetic field compensation correction coefficient is calculated using the magnetic field compensation correction formula;

[0112] Step S65, after each round of correction coefficient calculation is completed, the temperature compensation verification test, vibration compensation verification test and magnetic field compensation verification test are re-executed until each compensation effect meets the preset threshold requirements.

[0113] Based on the results of the verification test, the system uses the following formula to calculate the correction coefficient: Temperature compensation correction formula: α1_new =α1×(1±ΔT / TS), α2_new =α2×(1±ΔD / TD); Vibration compensation correction formula: β1_new =β1×(1±ΔF / VF), β2_new =β2×(1±ΔV / VA); Magnetic field compensation correction formula: γ1_new =γ1×(1±ΔS / MS), γ2_new =γ2×(1±ΔG / MG); wherein each Δ value represents the deviation of the corresponding parameter beyond the threshold. After each round of correction coefficient calculation is completed, the system must re-execute the complete verification test process until all compensation effects meet the preset threshold requirements. In this embodiment, through this rigorous dynamic verification and correction process, the correction coefficient finally determined can ensure that the IMU maintains stable and reliable measurement accuracy under various actual working conditions.

[0114] Step S70: Use the correction coefficient to perform compensation processing on the original data, and obtain a correction sequence through data reconstruction.

[0115] First, the system starts from the starting point of the time series, reads the original angular velocity and acceleration data at each sampling moment in turn, and synchronously obtains the corresponding environmental parameters such as temperature, vibration, and magnetic field. The environmental parameters correspond to the IMU data one by one, accurately describing the real-time working status of the IMU. This process fully reflects the principles of time synchronization and spatial correspondence. The system uses timestamps to accurately align IMU data with environmental parameters, ensuring that each IMU sampling point can find a unique corresponding temperature, vibration, and magnetic field measurement value. At the same time, the collection of environmental parameters is consistent with the spatial orientation of the IMU sensitive axis, that is, the sensitive axes of the thermometer, accelerometer, and magnetometer are parallel to the x, y, and z axes of the IMU. This spatiotemporal alignment strategy ensures that the effects of environmental factors can be accurately mapped to the various components of the IMU output.

[0116] Combination Figure 6 As shown, in some embodiments, step S70 may include: step S71, step S72, step S73 and step S74. Among them:

[0117] Step S71, for each data point of the original data, according to the measured temperature value and the temperature change rate, substitute into the calculation formula of the temperature compensation factor to calculate the temperature compensation amount at the corresponding moment of the current data point;

[0118] For each data point of the acquired IMU raw data, the system first substitutes the measured temperature value T and the temperature change rate dT / dt into the temperature compensation factor calculation formula CT = α1_new·T + α2_new·dT / dt to obtain the temperature compensation at that moment. α1_new represents the corrected static temperature compensation coefficient, which reflects the sensitive characteristics of the IMU zero bias to temperature; α2_new represents the corrected dynamic temperature compensation coefficient, which is used to process the transient response caused by temperature changes. For example, when the temperature rises from 25°C to 30°C, the system uses α1_new to calculate the zero bias change; when the temperature changes at a rate of 0.5°C / min, α2_new is used to calculate the additional dynamic error.

[0119] Step S72, calculating the vibration compensation amount in combination with the vibration data, substituting the frequency characteristics and amplitude characteristics of the vibration data into the vibration compensation model to obtain the vibration compensation amount;

[0120] Next, the system calculates the vibration compensation amount based on the vibration data. The vibration data contains frequency characteristics V(f) and amplitude characteristics A, which respectively reflect the spectral distribution and intensity of vibration excitations from different sources. The system substitutes V(f) and A into the vibration compensation model CV =β1_new·V(f) +β2_new·A to obtain the vibration compensation amount. β1_new represents the corrected frequency response coefficient, which is used to process the response characteristics of the IMU to vibrations of different frequencies; β2_new represents the corrected amplitude correction coefficient, which is used to compensate for output fluctuations caused by vibration intensity. This step pays special attention to the compensation effects of the two main frequency bands: engine vibration (20-50Hz) and road excitation (10-30Hz).

[0121] Step S73, reading the magnetic field sensor measurement value corresponding to the current data point, extracting the magnetic field strength, and subtracting the magnetic field measurement values ​​of adjacent data points to obtain the magnetic field gradient, substituting the magnetic field strength and the magnetic field gradient into the magnetic field compensation model to obtain the magnetic field compensation amount;

[0122] The calculation of magnetic field compensation requires the use of magnetic field strength B and magnetic field gradient The system reads the magnetic field sensor measurement value corresponding to the current IMU data, extracts the magnetic field strength B, and subtracts the magnetic field measurement values ​​of adjacent data points to obtain the magnetic field gradient. . Substitute into the magnetic field compensation model , calculate the magnetic field compensation amount. γ1_new represents the corrected magnetic field strength compensation coefficient, and γ2_new represents the corrected gradient compensation coefficient. These two parameters are used to deal with the influence of uniform magnetic field and gradient magnetic field respectively.

[0123] Step S74, performing angular velocity data correction and acceleration data correction on each data point in the original data according to the temperature compensation amount, the vibration compensation amount and the magnetic field compensation amount, so as to generate a correction sequence of the same length as the original data.

[0124] After obtaining the temperature compensation, vibration compensation and magnetic field compensation, the system starts to correct the original data of the IMU. For the angular velocity data, the corrected angular velocity value ω_comp can be expressed as the original angular velocity ω_orig minus the temperature compensation CT and magnetic field compensation CM, that is, ω_comp = ω_orig - CT - CM. This process shows in a physical sense that the drift error caused by temperature and magnetic field is effectively suppressed, and the corrected angular velocity is closer to the actual rotation state of the IMU.

[0125] The correction of acceleration data needs to take into account the influence of vibration compensation. After deducting the temperature compensation amount CT, the system divides the corrected acceleration value by the vibration amplification factor (1 + CV) to offset the amplitude error caused by vibration. The corrected acceleration a_comp can be expressed as a_comp = (a_orig - CT) / (1 + CV). This process reflects the nonlinear response characteristics of the IMU output to vibration. When the vibration compensation coefficient CV is large, the correction amplitude of the acceleration reading also increases accordingly.

[0126] By performing the above compensation calculations on each data point in the sampling sequence, the system finally generates a correction sequence of the same length as the original data. The corrected angular velocity sequence and acceleration sequence inherit the time tag of the original data while eliminating the interference of various environmental factors, so they have higher stability and reliability. From the perspective of time characteristics, the corrected IMU data can accurately reflect the continuous motion state of the carrier and eliminate the random walk caused by time-varying errors. From the perspective of spatial characteristics, calibration makes the output of each axis of the IMU return to the orthogonal ideal coordinate system, overcomes the installation error and the original system error, and realizes high-precision attitude measurement.

[0127] Step S80, performing stability analysis on the correction sequence and obtaining a drift parameter according to the change trend.

[0128] Combination Figure 7 As shown, in some embodiments, step S80 may include: step S81, step S82, step S83 and step S84. Among them:

[0129] Step S81, decomposing the correction sequence in time scale, setting three time windows of short-term, medium-term and long-term respectively, and extracting features from the IMU data in each time window using different statistical methods;

[0130] The system decomposes the correction sequence into time scales and sets three observation windows of 1 minute, 10 minutes and 1 hour. For the IMU data (three-axis angular velocity ωx / ωy / ωz and three-axis acceleration ax / ay / az) in each time window, the system uses different statistical methods to extract features.

[0131] In a 1-minute window, data is collected every 0.005s (sampling frequency 200Hz), and the mean square error and kurtosis of 12,000 sampling points are calculated to obtain the short-term fluctuation characteristics of the IMU. The mean square error reflects the degree of discreteness of the data, and the kurtosis characterizes the morphological characteristics of the data distribution. These two parameters together describe the short-term stability of the IMU output.

[0132] Within a 10-minute window, the system calculates the average value once a minute to obtain 10 data points. These points are fitted by the least squares method to obtain the slope of the data change, which reflects the rate of change of the IMU output. At the same time, the curvature characteristics of the data are calculated to characterize the degree of nonlinearity of the IMU output. These medium-term characteristic parameters can reflect the stability change trend of the IMU over a long period of time. The system pays special attention to the change law of the slope because it directly reflects the drift characteristics of the IMU zero bias. For example, in a stationary state, a sudden change in the slope often indicates a decrease in zero bias stability.

[0133] Within a 1-hour window, the system calculates the average value every 10 minutes to obtain 6 data points. The mean and drift rate are calculated from these data points, where the mean reflects the central trend of the IMU output and the drift rate represents the amount of change per unit time. Long-term analysis pays special attention to the long-term stability characteristics of the IMU output, which is crucial for evaluating the performance of the IMU in practical applications. For example, in navigation applications, long-term stability directly affects the cumulative error of the position solution.

[0134] Step S82, establishing a data screening standard, the data screening standard includes: for a short-term time window, a variance threshold automatically adjusted according to the working state of the IMU is used for screening; for a medium-term time window, a reasonable trend threshold is determined by a statistical method; for a long-term time window, a drift threshold is determined based on Allan equation analysis;

[0135] Based on the analysis results of the above three time windows, the system has established strict data screening standards to ensure the accuracy of feature parameters. The short-term window uses a dynamically adjusted variance threshold for screening, which is automatically adjusted according to the working state of the IMU. When the IMU is in a stationary state, the variance threshold of the angular velocity output is set to 0.01 (° / s)², and the variance threshold of the acceleration output is 0.001 (m / s²)²; when the motion state is detected, these thresholds are relaxed to 0.05 (° / s)² and 0.005 (m / s²)² respectively. The medium-term window focuses on the trend changes of the data, and a reasonable trend threshold is determined by statistical methods. The data segment exceeding this threshold is marked as abnormal. The long-term window determines the drift threshold based on Allan variance analysis. This method takes into account the cumulative effect of random noise and can effectively identify long-term drift anomalies.

[0136] The filtered data extracts characteristic parameters in three time windows. In the short-term window, the system extracts the fluctuation intensity characteristics: calculates the standard deviation of angular velocity and acceleration output (reflecting the signal fluctuation amplitude), kurtosis (reflecting the signal mutation characteristics) and signal power spectrum density (reflecting the frequency characteristics). These parameters together characterize the instantaneous stability of the IMU, for example, the typical value of the standard deviation of the angular velocity output should be less than 0.02° / s in a stationary state. In the medium-term window, the system extracts the change trend characteristics: calculates the signal change rate (characterizing the drift speed) by linear fitting, calculates the signal curvature (reflecting the degree of nonlinearity) by polynomial fitting, and analyzes the autocorrelation characteristics of the signal (reflecting the time correlation of the data). These characteristic parameters describe the gradual characteristics of the IMU output, for example, the rate of change of the angular velocity under normal working conditions should be less than 0.1° / s / h. In the long-term window, the system extracts the drift characteristics: calculates the mean shift (reflecting the zero bias stability), the Allan variance coefficient (characterizing the random walk characteristics) and the trend component (reflecting the deterministic drift). These parameters comprehensively characterize the long-term stability of the IMU, for example, the typical value of the zero bias drift of the acceleration output should be less than 1mg / h.

[0137] Step S83, extracting characteristic parameters in the three time windows respectively according to the data screening criteria, and establishing a combined model including initial drift, linear drift and periodic drift based on the extracted characteristic parameters;

[0138] Based on the characteristic parameters extracted in three time windows, the system established a combined model including initial drift, linear drift and periodic drift. The drift parameter model is expressed as: D(t) = A·exp(-t / τ) + B·t + C·sin(ωt). The model contains three key parts: the exponential decay term A·exp(-t / τ) describes the stabilization process of the IMU after it is turned on, where A is the initial drift amplitude (unit: ° / s or m / s²), and τ is the decay time constant (unit: s); the linear term B·t reflects the long-term drift trend of the IMU, where B is the linear drift rate (unit: ° / s² or m / s³); the periodic term C·sin(ωt) is used to describe the periodic drift caused by environmental factors, where C is the periodic drift amplitude (unit: ° / s or m / s²), ω is the angular frequency (unit: rad / s), and t is the time variable (unit: s).

[0139] Step S84, using the data in the three time windows and the data in the correction sequence, calculate and determine the drift parameters of the combined model.

[0140] The system can determine the values ​​of each parameter in the model through the following steps:

[0141] 1 Determine the initial drift amplitude A: First, use the data in the 1-minute window to intercept the initial output sequence x(t) after the IMU is turned on. Calculate the mean μ0 and standard deviation σ0 of the output sequence, then the absolute value of the initial drift |A|=μ0 + 2σ0. This method takes into account the central value and fluctuation range of the output at the beginning of the startup. The sign of A is determined by the previous and next trends of the sequence x(t): if the sequence shows an upward trend, A is positive; if it shows a downward trend, A is negative.

[0142] 2 Determine the decay time constant τ: Use the data of the 10-minute window to fit the IMU output sequence y(t) with an exponential function: y(t) = A·exp(-t / τ) + c, where c is the steady-state value. Use the least squares method to estimate the parameters: minΣ[y(ti) - (A·exp(-ti / τ) + c)]², and find the τ value that minimizes the residual sum of squares through iterative calculation. In order to verify the fitting effect, calculate the coefficient of determination R², and the fitting is considered valid when R²>0.95.

[0143] 3 Determine the linear drift rate B: Segment the data of the 1-hour window into 10-minute segments, and obtain 6 data segments. Perform linear fitting on each segment of data to obtain 6 local slopes {b1, b2, ..., b6}. The reference drift rate B0 is the median of these slopes to reduce the influence of outliers. Then calculate the correction coefficient: k = 1 + (cmax - cmin) / cref, where cmax and cmin are the maximum and minimum curvatures of each segment of data, respectively, and cref is the reference curvature value (empirical value 0.001). The final linear drift rate B = B0·k.

[0144] 4 Determine the periodic term parameters C and ω: Perform FFT transformation on the 8-hour data sequence to obtain the spectrum F(ω). In order to avoid the influence of high-frequency noise, only the spectrum within the range of 0.0001Hz-0.01Hz is analyzed. In this range, the frequency component ω1 with the largest amplitude is found, and its corresponding amplitude is C1. The second largest frequency component ω2 and its amplitude C2 are used as supplements to describe minor periodic effects. These two frequencies usually correspond to the influence of ambient temperature changes (about 1800s period) and magnetic field changes (about 3600s period).

[0145] The drift parameters finally obtained fully describe the performance characteristics of IMU under different working conditions: the initial drift amplitude reflects the startup characteristics and initial accuracy level of the device, the decay time constant reflects the response ability and adjustment process of IMU to changes in working state, the linear drift rate characterizes the stability and cumulative error characteristics of long-term work, and the periodic term parameters reflect the influence of environmental factors. This complete parameter system not only reveals the working characteristics of IMU, but also provides a reliable basis for subsequent accuracy improvement and performance optimization.

[0146] Step S90, performing difference analysis between the drift parameter and the calibration reference, and obtaining the final calibration result through multi-point correction.

[0147] In order to verify the accuracy of the drift parameters obtained in step S80, the system establishes a complete calibration benchmark system, including a static benchmark and a dynamic benchmark. The final calibration result is determined by comparing and analyzing the predicted drift parameter value with the calibration benchmark value.

[0148] The calibration benchmark of IMU includes two parts: static benchmark and dynamic benchmark. The static benchmark is provided by a high-precision turntable to provide standard attitude angle and gravity acceleration benchmark, where the turntable's angular resolution is better than 0.001°, the repeatability is better than ±1", the attitude stability is better than ±0.2", and the controlled temperature environment is 25±0.5℃; the dynamic benchmark provides the benchmark value in motion through the standard acceleration platform and angular velocity platform, the acceleration measurement range is ±10g, the accuracy is better than 0.1mg, the angular velocity measurement range is ±300° / s, and the accuracy is better than 0.01° / s. These high-precision benchmark devices provide reliable reference standards for the calibration of IMU.

[0149] The system sets up multiple calibration points in the test field, including: 8 angle points (0°, 45°, 90°, 135°, 180°, 225°, 270°, 315°) in the static area to evaluate the measurement characteristics of the IMU in different postures; 8 speed points (20km / h, 30km / h, 40km / h, 50km / h, 60km / h, 70km / h, 80km / h, 90km / h) in the loop to analyze the dynamic response of the IMU at different speeds; 12 compound motion points are designed for the working road, including 5 acceleration points (0.1g, 0.2g, 0.3g, 0.4g, 0.5g), 4 turning radius points (10m, 20m, 30m, 40m, 50m) and 3 slope points (3%, 6%, 12%). The selection of these points is based on the actual application scenarios of IMU, where the 0° angle point faces the geographic north, the turning radius is selected taking into account the steering limit of the vehicle, and the slope setting refers to the standard road design specifications. Each point is tested three times, with a 30-minute interval between adjacent tests to eliminate the influence of residual effects.

[0150] First, a benchmark test is performed in a static area. The IMU is rotated to a preset angle using a precision turntable, and stays at each angle point for 10 minutes to record the three-axis angular velocity and acceleration output of the IMU. The system calculates the static error distribution by comparing the IMU output value with the standard attitude angle provided by the turntable. The purpose of the benchmark test is to determine the basic performance indicators of the IMU under ideal static conditions. By analyzing the 8 angle points, the system establishes a preliminary static error mapping relationship to provide a basic reference for subsequent dynamic tests.

[0151] After the static test is completed, the system starts the loop test based on the established error mapping relationship. On the standard loop, the vehicle passes through 8 speed points in sequence using a precision speed control system. Each speed point maintains a constant speed for 3 laps to ensure the stability of the data. The characteristics of the loop test are that the motion state is controllable and stable. The theoretical centripetal acceleration value can be accurately calculated by the vehicle speed and the loop radius, and it is used as a dynamic benchmark. By comparing the acceleration value measured by the IMU with the theoretical benchmark value, the system obtains the dynamic error distribution under different speed conditions. After obtaining the static and steady-state dynamic error characteristics, the system enters the working condition road test phase. 12 typical composite motion conditions are designed for the working condition road section, including: linear motion with acceleration from 0.1g to 0.5g (one point every 0.1g, a total of 5 points), turning motion with a turning radius from 10m to 50m (one point every 10m, a total of 4 points), and ramp motion with three slopes of 3%, 6%, and 12%. These working condition points are not isolated, but form a continuous test route through reasonable combination. For example, after completing a 0.3g acceleration, a 30m radius turn is performed immediately, and then a transition is made to a 6% slope climb. This combination of working conditions is closer to the actual application scenario.

[0152] During the working condition test, the system simultaneously collected data from high-precision benchmark sensors (including fiber optic gyroscopes, precision accelerometers, and RTK positioning systems) as a reference. By comparing the IMU output with the benchmark data, the error characteristics under complex dynamic conditions were obtained. The tests at this stage showed that the IMU error is not only related to a single motion parameter, but also affected by the motion state switching process. For example, when switching from uniform motion to turning acceleration, there will be a short-term error fluctuation.

[0153] Based on the results of the difference analysis, the three compensation factors of temperature, vibration and magnetic field established in step 5 were finally optimized to form a complete calibration system. This optimization method based on dual benchmarks (drift parameter prediction and measured benchmark) realizes comprehensive calibration from static to dynamic working conditions. Based on this, the system establishes compensation models for static components and dynamic components respectively. For the static component, the system first uses the test data of 8 angle points to calculate the error values ​​of the three axes of the IMU at each angle point. These error values ​​are sorted according to the angle size, and continuous angle-error curves are obtained through cubic spline interpolation to establish the mapping relationship between attitude angle and measurement error. For the dynamic component, the system first uses the data of 8 speed points in the loop test to establish the corresponding relationship between speed and centripetal acceleration error. Then, the data of 12 composite motion points in the working condition test are used to calculate the relationship between each motion parameter (such as acceleration, angular velocity, slope) and the measurement error. These error data are fitted by least squares to obtain the error coefficients of each motion parameter. Finally, these error coefficients are combined to form a complete dynamic compensation model. The coefficients of the compensation model are optimized through iterative calculations, and the optimization process is completed when the residual error is less than the set threshold (angular velocity 0.1° / s, acceleration 0.01g). This two-step compensation model can handle static errors and dynamic errors separately, improving the pertinence and effectiveness of calibration.

[0154] The final IMU dynamic calibration and compensation method includes two key parts: one is static component compensation, which solves the problem of basic measurement error by modeling and compensating the error of IMU at different attitude angles; the other is dynamic component compensation, which realizes error compensation under dynamic conditions by establishing a mapping relationship between motion parameters (speed, acceleration, angular velocity, slope, etc.) and measurement errors. This method realizes continuous calibration from static to dynamic conditions, and provides an effective technical solution for IMU dynamic error compensation.

[0155] The technical solution of the IMU dynamic calibration and compensation method of the present invention, by establishing a complete process of IMU raw data acquisition and dynamic calibration, combined with innovative methods of periodic decomposition and environmental parameter monitoring, can obtain key parameters such as temperature gradient field, vibration transfer characteristics and magnetic field spatial distribution in real time; this data acquisition solution based on multi-sensor collaboration significantly improves the accuracy of environmental factor identification, making subsequent compensation processing more targeted, and effectively overcomes the technical defects of traditional acquisition methods in insufficient environmental factor identification. In addition, a multi-level compensation mechanism is innovatively proposed, by respectively establishing temperature compensation factors, vibration compensation factors and magnetic field compensation factors, and combining dynamic inspection methods to optimize correction coefficients, a complete error compensation system is formed; this compensation mechanism can independently compensate for the influence characteristics of different environmental factors, greatly improving the measurement accuracy of IMU in complex environments. In addition, in terms of drift parameter extraction and multi-point correction, a feature extraction method based on time scale decomposition is designed. By establishing a combined drift parameter model and a multi-point correction mechanism, the long-term stability of the IMU output is achieved. The technical solution of the present invention fully considers the influence of initial drift, linear drift and periodic drift, and ensures the reliability of the calibration results through cyclic verification under different working conditions, thereby solving the stability problem of the IMU in long-term operation.

[0156] It should be noted that all the above-mentioned embodiments of the IMU dynamic calibration and compensation method of the present invention can be arbitrarily combined or combined to form a new embodiment when there is no contradiction or conflict between them.

[0157] Reference Figure 8 The present invention also proposes an IMU dynamic calibration and compensation device 100, which is applied to a vehicle-mounted navigation device.

[0158] In this embodiment, the data fusion processing device 100 includes:

[0159] The acquisition module 101 is used to acquire the raw data of the IMU, sample the data according to the standard sampling frequency, and integrate it with the environmental parameters to obtain an initial measurement sequence;

[0160] The first decomposition module 102 is used to periodically decompose the initial measurement sequence, calculate the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and obtain a measurement supplement sequence through parameter association;

[0161] The calibration module 103 is used to perform front-end calibration, middle-end calibration and end-end calibration on the measurement supplement sequence, and obtain the calibration reference value through parameter iteration;

[0162] The first analysis module 104 is used to analyze the error variation law of different working conditions according to the calibration reference value, and obtain the calibration mapping amount through static and dynamic feature analysis;

[0163] A second decomposition module 105 is used to decompose environmental factors using the calibration mapping amount to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor;

[0164] The inspection module 106 is used to dynamically inspect each compensation factor, and obtain a correction coefficient by comparing the threshold value in combination with the environmental change characteristics;

[0165] A correction module 107 is used to perform compensation processing on the original data using the correction coefficient and obtain a correction sequence by data reconstruction;

[0166] The second analysis module 108 is used to perform stability analysis on the correction sequence and obtain a drift parameter according to the change trend;

[0167] The output module 109 is used to perform difference analysis between the drift parameter and the calibration reference, and obtain the final calibration result through multi-point calibration.

[0168] The IMU dynamic calibration and compensation device 100 of the present invention can implement the steps of the above-mentioned IMU dynamic calibration and compensation method, and therefore at least has all the beneficial effects brought by the technical solutions of the above-mentioned IMU dynamic calibration and compensation method embodiments, which will not be described one by one here.

[0169] The present invention also provides a vehicle-mounted navigation device, which includes an IMU sensor and an electronic device, referring to Fig. 9 , Fig. 9 It is a schematic diagram of the structure of an electronic device in a hardware operating environment involved in an embodiment of the present invention.

[0170] The electronic device of the embodiment of the present invention may be a controller of a vehicle-mounted navigation device, or a computing device such as a single-chip microcomputer, a notebook, a palm computer, and a server. Fig. 9 As shown, the electronic device may include: a processor 1001 (such as a CPU), a network interface 1004, a user interface 1003, a memory 1005 and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit, such as a keyboard (Keyboard), and the user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0171] Those skilled in the art will understand that Fig. 9 The electronic device structure shown in the figure does not constitute a limitation of the electronic device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0172] like Fig. 9 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module and a computer program.

[0173] exist Fig. 9 In the electronic device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client (user end) and communicate data with the client; and the processor 1001 can be used to call the computer program stored in the memory 1005. When the computer program is called and executed by the processor 1001, the steps of the above-mentioned IMU dynamic calibration and compensation method are implemented.

[0174] Based on the computer program proposed in the aforementioned embodiment, the present invention further proposes a storage medium, which stores the computer program. When the computer program is executed by a controller, the IMU dynamic calibration and compensation method recorded in the aforementioned embodiment is implemented.

[0175] Since the vehicle-mounted navigation device and storage medium of the present invention can implement the steps of the above-mentioned IMU dynamic calibration and compensation method, they at least have all the beneficial effects brought by the technical solutions of the above-mentioned IMU dynamic calibration and compensation method embodiments, which will not be described one by one here.

[0176] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0177] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0178] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0179] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0180] The above description is only a partial or preferred embodiment of the present invention. Neither the text nor the drawings can limit the scope of protection of the present invention. All equivalent structural changes made by using the contents of the present invention specification and drawings under the overall concept of the present invention, or direct / indirect application in other related technical fields are included in the scope of protection of the present invention.

Claims

1. An IMU dynamic calibration and compensation method, applied to a vehicle-mounted navigation device, characterized in that: include: Obtain the raw data from the IMU, sample the data at the standard sampling frequency, and integrate it with the environmental parameters to obtain the initial measurement sequence; Performing periodic decomposition on the initial measurement sequence, calculating the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and obtaining a measurement supplement sequence through parameter association; Performing front-end calibration, middle-end calibration and end-end calibration on the measurement supplementary sequence, and obtaining a calibration reference value through parameter iteration; According to the calibration reference value, analyzing the error variation law of different working conditions, and obtaining the calibration mapping amount through static and dynamic characteristic analysis; Decomposing environmental factors using the calibration mapping amount to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor; Dynamically test each compensation factor, combine it with the characteristics of environmental changes, and obtain the correction coefficient through threshold comparison; Using the correction coefficient to perform compensation processing on the original data, and obtaining a correction sequence through data reconstruction; Performing stability analysis on the correction sequence and obtaining drift parameters according to the change trend; The drift parameters are differentially analyzed with the calibration reference, and the final calibration result is obtained through multi-point correction.

2. The IMU dynamic calibration and compensation method according to claim 1, characterized in that: The periodic decomposition of the initial measurement sequence, the calculation of the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and the acquisition of the measurement supplement sequence by parameter association include: The analysis window length and sliding step length are set according to the signal characteristics of the IMU, and within each analysis window, the angular velocity and acceleration signals are subjected to fast Fourier transformation to obtain the frequency spectrum characteristics of the signal, and the main frequency components in the signal are identified by calculating the power spectrum density; Based on the frequency spectrum characteristics and the main frequency components, the correlation between the temperature change, the vibration change, the magnetic field change and the IMU output value is analyzed by calculation; Based on the correlation between temperature change, vibration change, magnetic field change and the IMU output value, a correlation matrix between environmental parameters and IMU output is constructed, wherein the row vectors of the correlation matrix represent temperature values, three-axis vibration values ​​and three-axis magnetic field values, and the column vectors of the correlation matrix represent the three-axis angular velocity and three-axis acceleration of the IMU, and the influence coefficient of temperature on zero bias, the influence coefficient of temperature change rate, the gain coefficient of conduction function and the sensitivity coefficient of magnetic field interference in the correlation matrix are calculated; Using each coefficient of the correlation matrix, the IMU output is calculated item by item to obtain the zero bias change caused by temperature, the output fluctuation caused by vibration, and the influence of magnetic field interference; According to the item-by-item calculation results, a measurement supplement sequence is generated whose time points correspond to the time points in the initial measurement sequence, and the measurement supplement sequence records the data at each sampling moment.

3. The IMU dynamic calibration and compensation method according to claim 1, characterized in that: The measurement supplementary sequence is subjected to front-end calibration, middle-end calibration and end-end calibration, and a calibration reference value is obtained through parameter iteration, including: Collect the front calibration data in the static area of ​​the test site, collect the middle calibration data on the standard loop, and collect the final calibration data on the working road; Based on the front-end calibration data, the middle-end calibration data and the final-end calibration data, an IMU output error data model is established; For each sampling point, calculating the error vector between the measured output value and the theoretical calculated value, wherein the theoretical calculated value is calculated based on the kinematic model; Constructing a measurement matrix, wherein each row of the measurement matrix corresponds to an observation, and each column corresponds to a coefficient of temperature, vibration, and magnetic field, and using the error vector to perform parameter optimization by the least squares method to calculate the parameter correction amount; The parameters of the error data model are updated according to the parameter correction amount and the iterative optimization is continued until the parameter change amount is less than a preset threshold value, thereby obtaining the calibration reference value; Among them, the calibration reference value includes a temperature compensation reference value, a vibration compensation reference value and a magnetic field compensation reference value. The temperature compensation reference value includes a temperature-zero bias correction coefficient and a temperature change rate correction coefficient. The vibration compensation reference value includes a vibration amplitude correction coefficient and a vibration frequency response coefficient. The magnetic field compensation reference value includes a magnetic field intensity correction coefficient and a magnetic field gradient correction coefficient.

4. The IMU dynamic calibration and compensation method according to claim 3, characterized in that: The step of analyzing the error variation rules of different working conditions according to the calibration reference value and obtaining the calibration mapping amount through static and dynamic feature analysis includes: Determine the temperature effect characteristics of the temperature change on the IMU output based on the measured temperature-bias correction coefficient and the temperature change rate correction coefficient; Based on the test results of the vibration amplitude correction coefficient and the vibration frequency response coefficient, recording the vibration response characteristics at different vehicle speeds; Based on the test data of the magnetic field intensity correction coefficient and the magnetic field gradient correction coefficient, record the magnetic field interference effect of the magnetic field change on the IMU output; Based on the above-mentioned temperature influence characteristics, vibration response characteristics and magnetic field interference influence, a hierarchical structure including a static mapping layer and a dynamic mapping layer is established to form the complete calibration mapping amount.

5. The IMU dynamic calibration and compensation method according to claim 1, characterized in that: The dynamic inspection of each compensation factor, combined with the characteristics of environmental changes, and the correction coefficient obtained by threshold comparison include: Performing a temperature compensation verification test in a static area of ​​a test site, during which the temperature compensation factor is applied in real time to perform compensation processing, and determining whether a coefficient of the temperature compensation factor needs to be adjusted according to a preset threshold requirement; Conducting a vibration compensation verification test on a standard loop. During the verification test, the vibration compensation factor is applied in real time to process the IMU output data, and the real-time vibration characteristics are recorded. It is determined whether the coefficient of the vibration compensation factor needs to be adjusted according to the preset threshold requirements. A magnetic field compensation verification test is performed on a working road. During the verification test, when passing through a preset magnetic field interference area, the magnetic field compensation factor is applied for real-time compensation, and it is determined whether the coefficient of the magnetic field compensation factor needs to be adjusted according to a preset threshold requirement; When it is determined that the coefficient of the temperature compensation factor needs to be adjusted, the temperature compensation correction coefficient is calculated using the temperature compensation correction formula; when it is determined that the coefficient of the vibration compensation factor needs to be adjusted, the vibration compensation correction coefficient is calculated using the vibration compensation correction formula; when it is determined that the coefficient of the magnetic field compensation factor needs to be adjusted, the magnetic field compensation correction coefficient is calculated using the magnetic field compensation correction formula; After each round of correction coefficient calculation is completed, the temperature compensation verification test, vibration compensation verification test and magnetic field compensation verification test are re-executed until each compensation effect meets the preset threshold requirements.

6. The IMU dynamic calibration and compensation method according to claim 1, characterized in that: The using the correction coefficient to perform compensation processing on the original data and obtaining a correction sequence by data reconstruction includes: For each data point of the original data, according to the measured temperature value and the temperature change rate, the temperature compensation factor is substituted into the calculation formula to calculate the temperature compensation amount at the time corresponding to the current data point; Calculate the vibration compensation amount based on the vibration data, substitute the frequency characteristics and amplitude characteristics of the vibration data into the vibration compensation model, and calculate the vibration compensation amount; Read the magnetic field sensor measurement value corresponding to the current data point, extract the magnetic field strength, and obtain the magnetic field gradient by subtracting the magnetic field measurement values ​​of adjacent data points, and substitute the magnetic field strength and the magnetic field gradient into the magnetic field compensation model to obtain the magnetic field compensation amount; According to the temperature compensation amount, the vibration compensation amount and the magnetic field compensation amount, angular velocity data correction and acceleration data correction are performed on each data point in the original data to generate a correction sequence with the same length as the original data.

7. The IMU dynamic calibration and compensation method according to any one of claims 1 to 6, characterized in that: The stability analysis of the correction sequence is performed to obtain the drift parameter according to the change trend, including: Decomposing the correction sequence in time scale, setting three time windows of short-term, medium-term and long-term respectively, and extracting features from the IMU data in each time window by using different statistical methods; Establishing data screening criteria, the data screening criteria include: for short-term time windows, using a variance threshold that is automatically adjusted according to the working state of the IMU for screening; for medium-term time windows, determining a reasonable trend threshold by statistical methods; for long-term time windows, determining a drift threshold based on Allan equation analysis; According to the data screening criteria, characteristic parameters are extracted in the three time windows respectively, and based on the extracted characteristic parameters, a combined model including initial drift, linear drift and periodic drift is established; The drift parameters of the combined model are calculated and determined using the data in the three time windows and the data in the correction sequence.

8. An IMU dynamic calibration and compensation device, applied to a vehicle-mounted navigation device, characterized in that: include: The acquisition module is used to obtain the raw data of the IMU, sample the data according to the standard sampling frequency, and integrate it with the environmental parameters to obtain the initial measurement sequence; A first decomposition module is used to periodically decompose the initial measurement sequence, calculate the influence of temperature change, vibration change, and magnetic field change on the IMU output value, and obtain a measurement supplement sequence through parameter association; A calibration module, used for performing front-end calibration, middle-end calibration and end-end calibration on the measurement supplementary sequence, and obtaining a calibration reference value through parameter iteration; A first analysis module is used to analyze the error variation rules of different working conditions according to the calibration reference value, and obtain the calibration mapping amount through static and dynamic feature analysis; A second decomposition module is used to decompose environmental factors using the calibration mapping amount to obtain a temperature compensation factor, a vibration compensation factor and a magnetic field compensation factor; The inspection module is used to dynamically inspect each compensation factor and obtain the correction coefficient by comparing the threshold value in combination with the characteristics of environmental changes; A correction module, used to use the correction coefficient to perform compensation processing on the original data, and obtain a correction sequence through data reconstruction; A second analysis module is used to perform stability analysis on the correction sequence and obtain a drift parameter according to a change trend; The output module is used to perform difference analysis between the drift parameter and the calibration reference, and obtain the final calibration result through multi-point correction.

9. A vehicle-mounted navigation device, comprising an IMU sensor, characterized in that: The vehicle-mounted navigation device also includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of the IMU dynamic calibration and compensation method as described in any one of claims 1 to 7 are implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of the IMU dynamic calibration and compensation method according to any one of claims 1 to 7.

Images