A method and apparatus for closed loop decoding of a circular shift register synchronizer

Abstract

The application discloses a closed-loop decoding method of a circular inductive synchro, which comprises the following steps: applying a double-carrier in-phase excitation signal to the circular inductive synchro, wherein the double-carrier in-phase excitation signal comprises a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference; synchronously sampling inductive signals output by two orthogonal inductive coils; performing phase-sensitive demodulation and low-pass filtering processing on the first carrier frequency signal and the second carrier frequency signal through the phase reference, and obtaining baseband quadrature components; constructing a joint observation quantity based on the baseband quadrature components; iteratively estimating gain mismatch and phase deviation between two orthogonal inductive signal channels by using the joint observation quantity, and decoupling and correcting the baseband quadrature components according to the estimation result, so as to obtain corrected quadrature signals; and performing angle calculation and error compensation according to the corrected quadrature signals, so as to obtain an angle measurement value of the circular inductive synchro.

Description

Technical Field

[0001] This invention belongs to the field of position sensing and signal processing technology, and particularly relates to a closed-loop decoding method and apparatus for a circular induction synchronizer. Background Technology

[0002] Circular inductive synchronizers form angle-related coupling signals through an excitation coil and two orthogonal inductive coils. Traditional decoding often employs single-carrier excitation combined with single-path arctangent (atan2) or PLL tracking to calculate the angle. While this approach can meet general angle measurement requirements under static or low-to-medium dynamic conditions, it is prone to problems such as decreased linearity, transient jumps, pole pair misjudgment, and difficulty in simultaneously addressing end-to-end delay when faced with external narrowband interference, orthogonal channel gain and phase mismatch, second and third harmonic distortion, front-end local saturation, and temperature drift. Especially in high-dynamic servo scenarios, simply increasing filtering, lookup table nodes, or phase-locked loop bandwidth cannot simultaneously achieve low delay, anti-interference capability, and full-temperature consistency. Therefore, a circular inductive synchronizer decoding scheme that can form a closed-loop collaborative mechanism encompassing multi-source error joint observation, online correction, harmonic suppression, nonlinear compensation, parallel angle calculation, and parameter write-back updates is urgently needed. Summary of the Invention

[0003] To address the aforementioned issues, this invention proposes a closed-loop decoding method and apparatus for circular inductive synchronizers. This invention employs a circular inductive synchronizer decoding scheme that combines dual-carrier observation, online amplitude-phase joint correction, harmonic sequence suppression, and dual-path adaptive fusion. This invention enhances the observability of online joint estimation by using a shared phase reference for dual carriers, reduces the burden of subsequent compensation through one-time linear decoupling, and simultaneously drives path fusion and online parameter updates by constructing a set of health indicators. This forms a closed-loop decoding mechanism encompassing joint observation, online correction, harmonic suppression, spline compensation, parallel angle calculation, health fusion, and parameter updates. Dual-carrier in-phase excitation enables redundant observation information for the main and auxiliary dual channels, providing a data source for subsequent error estimation and correction, thus enhancing the system's anti-interference capability and reliability. Constructing joint observation measures allows for real-time and adaptive compensation of inherent signal channel distortion, thereby improving the angle decoding accuracy and system robustness of the circular inductive synchronizer.

[0004] The first aspect of the present invention provides a closed-loop decoding method for a circular inductive synchronizer, comprising: A dual-carrier in-phase excitation signal is applied to the circular inductive synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The induced signals output from the two orthogonal induction coils are synchronously sampled, and the first carrier frequency signal and the second carrier frequency signal are phase-sensitively demodulated and low-pass filtered through the phase reference to obtain the baseband quadrature components. Based on the baseband quadrature components, a joint observation is constructed. The joint observation is used to iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels. Based on the estimation results, the baseband quadrature components are decoupled and corrected to obtain the corrected quadrature signal. Angle calculation and error compensation are performed based on the corrected orthogonal signal to obtain the angle measurement value of the circular induction synchronizer.

[0005] Preferred options also include: Based on the phase change trend of the angle measurements and the zero-crossing count results, construct an absolute angle index and an incremental angle index; Based on the aforementioned angle and a pre-built set of health indicators, the parameters of the new decoupling matrix and spline interpolation compensation are updated online.

[0006] Preferably, the step of constructing the absolute angle index and the incremental index further includes: Based on the fusion perspective, the initial discrimination result of the number of pole pairs at the current moment is obtained by analyzing the phase change trend, counting the number of zero crossings, and calculating the phase circulation. A voting window is set up to conduct a majority vote on the initial results of the extreme logarithms, and the results are filtered by combining the residence time threshold to obtain the extreme logarithm discrimination results. When the pole-log number discrimination result is obtained by identifying two consecutive opposite decision results, the consistency between the phase circulation and the zero-crossing count is re-evaluated to obtain the evaluation result, and the pole-log number discrimination result is updated according to the evaluation result.

[0007] Preferably, the step of re-evaluating the consistency between the phase loop quantity and the zero-crossing count further includes: When two consecutive opposite extreme logarithm determination results are obtained, the abnormal reconfirmation mechanism is triggered, pausing the output of the extreme logarithm determination result at the current moment and maintaining the extreme logarithm determination result at the previous moment; Recalculate the phase loop quantity and the zero-crossing count, and perform a consistency check. If the verification passes, the updated extreme logarithm discrimination result is output; if the verification fails, the extreme logarithm discrimination result of the previous time step is maintained, and the abnormal status bit of the health indicator is marked.

[0008] Preferably, the step of constructing the joint observations further includes: The baseband quadrature component corresponding to the first carrier frequency is set as the main state tracking channel, and the baseband quadrature component corresponding to the second carrier frequency is set as the auxiliary observation channel. A joint error vector is calculated based on the data from the auxiliary observation channel. The joint error vector includes a static bias component and a dynamic drift component. The joint error vector is used as the objective function constraint term of the recursive least squares algorithm to perform joint iterative estimation of the gain mismatch and phase deviation of the master state tracking channel.

[0009] Preferably, the step of obtaining orthogonal signals further includes: Set the leakage factor; Construct the initial covariance matrix, the calculation expression is:

[0010] In the formula, α is the uncertainty scaling factor. It is the identity matrix; Based on the initial covariance matrix and the leakage factor, the gain vector, covariance matrix, channel gain ratio, and phase deviation are calculated through recursive least squares iteration. Construct the decoupling matrix and calculate the orthogonal signals. The calculation expression is as follows:

[0011]

[0012] In the formula, For orthogonal components, The orthogonal signal, For decoupling matrix, , These are the channel gain ratio and the phase deviation, respectively.

[0013] Preferably, the step of obtaining the angle of the circular induction synchronizer further includes: Based on the orthogonal signals, a compensated orthogonal signal is obtained through harmonic-spline composite processing; Based on the orthogonal signals, the first angle and the second angle are obtained by executing an angle calculation path; Within a preset statistical time window, calculate the noise spectral density, harmonic amplitude ratio, saturation duty cycle and temperature drift of the programmable gain amplifier, and construct health indicators. Based on the health indicators, a state machine with hysteresis is used to select the optimal mode, which includes CORDIC mode, PI-PLL mode and fusion mode, and obtain the corresponding angle.

[0014] Preferably, the step of obtaining the compensated orthogonal signal through harmonic-spline composite processing further includes: The filtered signal is obtained by adaptive notch filtering of the second and third harmonics based on the orthogonal signals; Based on the filtered signal, nonlinear compensation is performed using finite node spline search, linear or cubic spline interpolation, to obtain the compensated orthogonal signal.

[0015] Preferably, the step of obtaining the first angle and the second angle further includes: The first angle is calculated using the CORDIC module based on the first path signal in the orthogonal signals. The calculation expression is as follows:

[0016] In the formula, , These are the orthogonal components along the y-axis and the orthogonal components along the x-axis in the baseband orthogonal components, respectively. The second angle is calculated using the PI-PLL module based on the second path signal in the orthogonal signals.

[0017] A second aspect of the present invention provides a closed-loop decoding device for a circular induction synchronizer, comprising: An excitation generation module is used to apply a dual-carrier in-phase excitation signal to the circular inductive synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The sampling and demodulation module is used to synchronously sample the induced signals output by the two orthogonal induction coils, and to perform phase-sensitive demodulation and low-pass filtering on the first carrier frequency signal and the second carrier frequency signal through the phase reference to obtain the baseband quadrature components. The joint correction module is used to construct a joint observation based on the baseband quadrature components, use the joint observation to iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels, and decouple and correct the baseband quadrature components according to the estimation results to obtain the corrected quadrature signal. The calculation and compensation module performs angle calculation and error compensation based on the corrected orthogonal signal to obtain the angle measurement value of the circular induction synchronizer.

[0018] Because the present invention adopts the above technical solution, it has the following advantages and positive effects compared with the prior art: (1) By constructing joint observations using dual carriers, the recursive least squares (RLS) algorithm is used to estimate and correct the gain mismatch and phase deviation of the signal channel in real time and adaptively, thereby eliminating system errors. An adaptive notch filter is used to suppress harmonic interference, and spline interpolation is combined for nonlinear compensation, which effectively improves the signal-to-noise ratio and linearity of the signal, and further ensures the accuracy of angle calculation.

[0019] (2) By using dual-carrier in-phase excitation, redundant observation of the main-auxiliary dual channels is realized, providing richer data sources for error estimation and correction, and enhancing the system's anti-interference capability and fault tolerance in complex electromagnetic environments.

[0020] (3) The system adopts two solution paths, CORDIC and PI-PLL, in parallel and constructs a set of health indicators. It uses a state machine with hysteresis to perform optimal fusion, which ensures that the system can still output stable and reliable angle values ​​when a single path fails or its performance degrades.

[0021] (4) A robust integrated absolute / incremental angle index was constructed through extreme logarithm discrimination, majority voting and anomaly reconfirmation mechanism, which effectively avoids extreme logarithm misjudgment caused by noise or interference and ensures the continuity and correctness of angle information.

[0022] (5) Based on the set of health indicators, the system can dynamically update the decoupling matrix and spline compensation parameters, realizing adaptive compensation for long-term changes such as device aging and temperature drift.

[0023] (6) From excitation, sampling, correction to solution and compensation, a complete closed-loop feedback system is formed, which enables the system to continuously optimize its performance and adapt to different working states. Attached Figure Description

[0024] The specific embodiments of the present invention will be further described in detail below with reference to the accompanying drawings, wherein: Figure 1 This is a main flowchart of a closed-loop decoding method for a circular induction synchronizer according to the present invention; Figure 2 This is a frame diagram of a closed-loop decoding device for a circular induction synchronizer according to the present invention; Figure 3 This is a diagram illustrating an embodiment of the multi-frequency in-phase excitation and phase-sensitive demodulation timing in this invention; Figure 4 This is a diagram illustrating one embodiment of obtaining orthogonal signals in this invention; Figure 5 This is a diagram illustrating one embodiment of the solution and compensation calculation in this invention; Figure 6 This is a diagram illustrating an embodiment of constructing an integrated absolute / incremental angle index in this invention; Figure 7 This is a diagram illustrating an embodiment of updating the decoupling matrix and spline interpolation compensation parameters in this invention. Figure 8 This is a circuit diagram corresponding to a closed-loop decoding method for a circular induction synchronizer in this invention. Detailed Implementation

[0025] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer from the following description and claims. It should be noted that the drawings are all in a very simplified form and use non-precise ratios, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0026] It should be noted that all directional indicators (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicator will also change accordingly.

[0027] First Embodiment See Figures 1-4 and Figure 8 The first aspect of the present invention provides a closed-loop decoding method for a circular inductive synchronizer, comprising: A dual-carrier in-phase excitation signal is applied to the circular inductive synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The induced signals output from the two orthogonal induction coils are synchronously sampled, and the first and second carrier frequency signals are phase-sensitively demodulated and low-pass filtered by a phase reference to obtain the baseband quadrature components. Based on the baseband quadrature components, a joint observation is constructed. The joint observation is used to iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels. Based on the estimation results, the baseband quadrature components are decoupled and corrected to obtain the corrected quadrature signal. Angle calculation and error compensation are performed based on the corrected orthogonal signals to obtain the angle measurement value of the circular induction synchronizer. This application establishes a closed-loop decoding mechanism encompassing joint observation, online correction, harmonic suppression, spline compensation, parallel angle calculation, healthy fusion, and parameter write-back. The main process is as follows: S100: Multi-frequency in-phase excitation, specifically, superimposing a second carrier frequency that is in phase with and shares a phase reference on the first carrier frequency. The second carrier frequency is an integer nearest neighbor carrier frequency or located outside a preset external narrowband interference band, forming a dual-carrier excitation for online joint observation. The first and second carrier signals are restricted: the frequency range of the first carrier signal is set to [5, 50] kHz; the ratio f2 / f1 of the second and first carrier signals ranges from [1.8, 2.2] or is located outside a preset external narrowband interference band; the first and second carrier signals share the same phase-locked phase reference; and the phase difference between the first and second carrier signals ranges from [-5°, 5°]. The sampling frequency of the induced signals output from the two orthogonal induction coils meets the requirements; the two induced signals are sampled synchronously and aligned with the local oscillator phase in the digital-analog clock domain. The purpose of using dual carriers is not simply to increase frequency components, but to provide multi-source observation information under a unified phase reference for subsequent online joint estimation. The two induction signals are synchronously sampled after passing through a PGA and an anti-mixing filter; the demodulation local oscillator is generated by a phase-locked reference and is phase-sensitively demodulated for the first and second carrier frequency signals respectively to ensure phase alignment in the digital-to-analog clock domains.

[0028] S200: Synchronous sampling and phase-sensitive demodulation. The induced signals output from the two quadrature induction coils are simultaneously sampled after passing through a PGA and anti-aliasing filter. The first and second carrier frequency signals are then phase-sensitively demodulated using the phase reference to obtain the corresponding baseband quadrature components I / Q (x, y). The induced signals output from the two quadrature induction coils are synchronously sampled and aligned with the local oscillator phase in the digital-to-analog clock domain. The amplitude-phase mismatch of the I / Q baseband vectors is modeled as a coupling error resulting from unequal equivalent gain and non-orthogonal phase. An RLS with a leakage factor is used to jointly estimate the channel gain ratio and phase deviation online. Based on the joint estimation results, a decoupling matrix is ​​constructed to perform a one-time linear decoupling of the baseband quadrature components. Compared to separate estimation or offline static correction, this step maintains better online adaptability under temperature drift and device deviation changes, and reduces the coupling error faced by subsequent compensation steps. The amplitude-phase mismatch of the I / Q baseband vectors is modeled as a coupling error resulting from unequal equivalent gain and non-orthogonal phase. An online joint estimation of channel gain ratio and phase deviation is performed using RLS with leakage factor. Based on the joint estimation results, a decoupling matrix is ​​constructed to perform a one-time linear decoupling of the baseband quadrature components. Compared with separate estimation or offline static correction, this step can maintain better online adaptability under temperature drift and device deviation changes, and reduce the coupling error faced by subsequent compensation steps.

[0029] S300: Amplitude and phase joint online correction, constructing observations based on the baseband orthogonal components corresponding to dual carriers, using a recursive least squares algorithm with leakage factor to jointly estimate the channel gain ratio and phase deviation online, and using the decoupling matrix to perform one-time linear decoupling to obtain orthogonal signals; S400: First, the angle of the circular inductor synchronizer is obtained through harmonic-spline sequential composite compensation, parallel calculation, and adaptive fusion. The end-to-end decoding delay is no more than 50µs at a 100MHz system clock.

[0030] This application also includes S500: pole-logarithm discrimination and indexing, which uses the phase circulation and zero-crossing count of the angle, combined with majority voting and dwell thresholds, to perform anomaly reconfirmation in case of conflict, thereby reducing the probability of index misjudgment under high dynamic conditions. S600: output and online update, outputting the fused angle and health indicators, and using the health indicators to update the decoupling matrix parameters and spline lookup table parameters online. S100~S600 are deployed on any platform of FPGA, SoC, or MCU combined with DSP, wherein the platform clock setting frequency is 100MHz, and the end-to-end delay is no greater than 50µs. This method is applied to a host computer or servo driver, wherein the host computer is configured to receive the fused angle and health indicators and perform closed-loop control and self-diagnosis.

[0031] This application uses the following parameters: in this embodiment, the frequencies of the first and second carrier signals are 10kHz and 19kHz respectively, the phase difference between the first and second carrier signals is 0°±3°, the sampling frequency is 200kHz, the PGA gain is 6dB, and the frequency used for anti-aliasing filtering is 35kHz; the leakage factor in the least squares algorithm is set to 0.995, and the initial covariance matrix is... The step size for adaptive notch filtering: The finite node spline search node is located at 5° intervals, using cubic spline interpolation; the PI-PLL module has a bandwidth of 80Hz; the CORDIC module has 12 iteration levels; the fusion hysteresis bandwidth of the state machine with hysteresis is set to 3dB, and the voting window is set to 20ms; in temperature zone management, K=3, the threshold is set to -20°C / +25°C / +85°C, the cross-zone linear transition is 200ms, the end-to-end delay is less than or equal to 45us, and a 100MHz clock can be used. This embodiment can also employ the following: under external 20kHz narrowband interference injection, the frequencies of the first and second carrier frequencies are 9kHz and 18.5kHz, respectively. When the harmonic ratio or PGA saturation duty cycle exceeds the threshold, the fusion unit biases the weights towards the PLL and increases the bandwidth to 150Hz; if two consecutive pole pair discriminations are mutually exclusive, the pole pair output is frozen for 10ms and the circulation and zero-crossing consistency are re-evaluated; if the CRC check fails, it reverts to the last successful parameter set and marks the QI abnormal bit.

[0032] See Figure 6 and Figure 7 Preferably, it also includes: An integrated absolute / incremental angle index is constructed by using pole-log discrimination, majority voting, and dwell time thresholds to determine phase circumference and zero-crossing counts. The parameters of the decoupling matrix and spline interpolation compensation are updated based on the set of angles and health indicators.

[0033] Preferred options also include: Based on the phase change trend of the angle measurement values ​​and the zero-crossing count results, construct the absolute angle index and the incremental index; Based on the aforementioned angle and a pre-built set of health indicators, the parameters of the new decoupling matrix and spline interpolation compensation are updated online.

[0034] The temperature range is divided into multiple temperature zones, each maintaining joint correction parameters and spline lookup table parameters. Smooth transitions between zones are achieved using linear or piecewise polynomial interpolation. The system not only outputs a set of health indicators but also uses this set to update the decoupling matrix parameters and spline lookup table parameters online, forming a closed-loop link from joint observation to parameter write-back. Parameter storage is CRC-checked, and in case of anomalies, it rolls back to the most recent valid version. Under a 100MHz system clock, the end-to-end decoding latency is no greater than 50μs. This invention achieves real-time compensation for signal channel gain mismatch and phase deviation through dual-carrier observation and online amplitude-phase joint correction, significantly improving the accuracy and robustness of angle decoding. The system constructs an adaptive closed-loop mechanism based on the health indicator set, capable of updating parameters online according to temperature changes and device aging, and possessing parameter verification and rollback functions, ensuring long-term stability and reliability. Through multi-path solution and intelligent optimization fusion, and by optimizing the end-to-end processing flow, low-latency decoding below 50μs is achieved, meeting the stringent requirements of high real-time application scenarios. By combining multiple information sources such as phase circulation, zero-crossing count, and majority voting, and introducing dwell time as a threshold, noise and transient interference are effectively filtered out, avoiding pole-logarithmic misjudgments caused by signal jitter and ensuring the accuracy of the angle index. When the system detects two consecutive opposite pole-logarithmic judgment results, an anomaly reconfirmation mechanism is triggered. During this period, the system freezes its output and reassesses the consistency of key parameters, thereby self-correcting when potential errors occur, greatly improving the system's fault tolerance and reliability. This method is not just a simple discrimination logic, but a complete closed-loop process including initial discrimination, voting filtering, and anomaly confirmation. The judgment strategy is dynamically adjusted based on historical judgment results to ensure stable and accurate output of the absolute / incremental angle index even under complex operating conditions.

[0035] Preferably, the steps of constructing the absolute index and the incremental index further include: Based on the fusion perspective, the initial discrimination result of the number of pole pairs at the current moment is obtained by analyzing the phase change trend, counting the number of zero crossings, and calculating the phase circulation. Set up a voting window to conduct a majority vote on the initial extreme logarithmic results, and combine the joint threshold of residence time to filter and obtain the extreme logarithmic discrimination results; When the pole-log discrimination result is obtained by identifying two consecutive opposite decision results, the consistency between the phase circulation and the zero-crossing count is re-evaluated to obtain the evaluation result, and the pole-log discrimination result is updated according to the evaluation result.

[0036] The pole-log discrimination method utilizes the phase circulation and zero-crossing count of the fusion angle, combined with the majority voting window and dwell threshold; if two consecutive opposite decisions occur, an abnormal reconfirmation is triggered to avoid misjudgment. The pole-log discrimination method employs... majority vote and length of stay The system employs a joint threshold of milliseconds (ms) and performs anomaly reconfirmation upon two consecutive opposing decisions. The temperature range is divided into multiple temperature zones, each maintaining its own joint correction parameters and spline lookup table parameters. Smooth transitions between zones utilize linear or piecewise polynomial interpolation. The system not only outputs a set of health indicators but also uses this set to update the decoupling matrix parameters and spline lookup table parameters online, forming a closed-loop link from joint observation to parameter write-back. The parameter storage is CRC-checked, and in case of anomalies, it reverts to the most recent valid version. Under a 100MHz system clock, the end-to-end decoding latency is no greater than 50μs.

[0037] Preferably, the step of re-evaluating the consistency between the phase loop quantity and the zero-crossing count further includes: When two consecutive opposite extreme logarithm determination results are obtained, the abnormal reconfirmation mechanism is triggered, pausing the output of the extreme logarithm determination result at the current moment and maintaining the extreme logarithm determination result at the previous moment; Recalculate the phase circulation and zero-crossing counts and perform a consistency check; If the verification passes, the updated extreme logarithm discrimination result is output; if the verification fails, the extreme logarithm discrimination result of the previous time step is maintained, and the abnormal status bit of the health indicator is marked.

[0038] Path A uses the CORDIC module for fast angle calculation, while path B uses the PI-PLL module for phase tracking angle calculation. Running the two paths in parallel is not simply redundant, but rather leverages the complementary characteristics of the two algorithms under different operating conditions: when noise and harmonics are low, the CORDIC module offers advantages in implementation simplicity and low latency; when harmonics are enhanced, local saturation is present, or temperature drift is significant, the PI-PLL module provides smoother dynamic tracking output. The health index QI is defined as a fixed set of indicators, including: noise spectral density (I / Q channel baseband noise estimation); harmonic amplitude ratio (2f / 3f to fundamental amplitude ratio); PGA saturation duty cycle; and temperature drift rate (parameter change rate). The fusion unit performs adaptive weighted fusion or selective switching based on the set of health indicators: when the harmonic amplitude ratio or PGA saturation duty cycle increases, the PLL path weight is increased; otherwise, the CORDIC path weight is increased. Hysteresis is used to avoid back-and-forth jitter, and the state and threshold are recorded during switching. The pole-pair discrimination utilizes the phase circulation of the fused angle and the zero-crossing count, combined with majority voting and a dwell threshold; if two consecutive opposite decisions occur, an anomaly reconfirmation is triggered to avoid misjudgment. The iteration level of the CORDIC module is not less than 10, and the bandwidth range of the PI-PLL module is... Hz.

[0039] Preferably, the step of constructing the joint observations further includes: The baseband quadrature component of the first carrier frequency is set as the main state tracking channel, and the baseband quadrature component of the second carrier frequency is set as the auxiliary observation channel. The baseband orthogonal component calculation based on the auxiliary observation channel includes the joint error vector containing static bias and dynamic drift; By incorporating the joint error vector into the objective function of the recursive least squares algorithm, online joint constraints and iterations are performed on the gain mismatch and phase deviation of the master state tracking channel.

[0040] By dividing the dual-carrier signal into a main tracking channel and an auxiliary observation channel, and using the joint error vector calculated from the auxiliary channel, this method can simultaneously and in real-time estimate and correct the two most critical error sources in the main channel: gain mismatch and phase deviation, fundamentally improving the fidelity of orthogonal signals. This method not only considers the static deviation of the signal channel but also, through the introduction of dynamic drift, can adapt to and compensate for slowly varying errors caused by factors such as temperature changes and device aging, making the correction algorithm more adaptable and robust. By directly incorporating the error vector into the objective function of the recursive least squares (RLS) algorithm, online joint constraints and iterations of the main channel parameters are achieved.

[0041] Preferably, the step of obtaining orthogonal signals further includes: Set the leakage factor; Construct the initial covariance matrix, the calculation expression is:

[0042] In the formula, For uncertainty scaling factor, It is the identity matrix; Based on the initial covariance matrix and leakage factor, the gain vector, covariance matrix, channel gain ratio, and phase deviation are calculated through recursive least squares iteration. Construct the decoupling matrix and calculate the orthogonal signals. The calculation expression is as follows:

[0043]

[0044] In the formula, For orthogonal components, Orthogonal signals For decoupling matrix, , These are the channel gain ratio and phase deviation, respectively.

[0045] By introducing a leakage factor into the Recursive Least Squares (RLS) algorithm, the covariance matrix is ​​prevented from approaching zero after long-term operation. This allows the system to maintain sensitivity to parameter changes and ensures long-term effective error tracking under complex conditions. By setting an initial covariance matrix that includes an uncertainty scaling factor, the gain in the initial stage of the algorithm can be flexibly controlled. This avoids oscillations caused by blind parameter jumps at startup and optimizes the convergence speed and stability of the algorithm based on the prediction of initial errors. The decoupling matrix is ​​directly constructed using the channel gain ratio and phase deviation obtained through real-time iteration, and the comprehensive correction of the original orthogonal signals can be completed in a single matrix multiplication operation. This method of transforming complex nonlinear amplitude and phase errors into standard linear algebra operations greatly reduces the compensation burden of the subsequent angle calculation module and improves the real-time performance of the system. The range of the uncertainty scaling factor is... Leakage factor .

[0046] Figure 4 and Figure 5 Preferably, the step of obtaining the angle of the circular induction synchronizer further includes: Compensated orthogonal signals are obtained by harmonic-spline composite processing based on orthogonal signals; The first and second angles are obtained by executing an angle calculation path based on orthogonal signals; Within a preset statistical time window, calculate the noise spectral density, harmonic amplitude ratio, saturation duty cycle and temperature drift of the programmable gain amplifier, and construct health indicators. Based on health indicators, a state machine with hysteresis is used to select the optimal mode, which includes CORDIC mode, PI-PLL mode and fusion mode, and obtain the corresponding perspective.

[0047] A composite processing mechanism combining harmonic suppression and spline interpolation is employed to eliminate high-order harmonic errors and spatial nonlinear errors in the induced signal, improving the quality of the orthogonal signal fed into the solution module and laying the foundation for angle acquisition. Breaking through the limitations of traditional methods that rely solely on a single signal for evaluation, a comprehensive set of health indicators is constructed by comprehensively extracting multi-dimensional features such as noise spectral density, harmonic amplitude ratio, PGA saturation duty cycle, and temperature drift rate within a time window. This enables a deep and quantitative perception of the system's underlying hardware status and signal quality. A state machine with hysteresis characteristics is introduced to dynamically and smoothly switch between CORDIC, PI-PLL, and fusion modes based on the health indicators. This not only effectively avoids the risk of single solution path failure but also prevents frequent jitter at mode switching boundaries through the hysteresis mechanism, ensuring the optimality and stability of the angle output under all operating conditions. The statistical time window for the health indicator (QI) is set within a preset ms. The health indicator includes noise spectral density, harmonic amplitude ratio, saturation duty cycle of the programmable gain amplifier, and temperature drift. When the harmonic amplitude ratio or the saturation duty cycle of the programmable gain amplifier increases, the weight of the PI-PLL module is increased; otherwise, the weight of the CORDIC path is increased. Hysteresis is implemented during weight switching to avoid jitter. Weight parameters are updated or selectively switched based on the health indicator, with hysteresis to prevent jitter.

[0048] Preferably, the step of obtaining the compensated orthogonal signal through harmonic-spline composite processing further includes: The filtered signal is obtained by adaptive notch filtering of the second and third harmonics based on orthogonal signals; Based on the filtered signal, nonlinear compensation is performed through finite node spline search, linear or cubic spline interpolation, and the compensated orthogonal signal is obtained.

[0049] An adaptive notch filter is used to suppress the second and third harmonics of the equivalent fundamental wave, with a step size set to... The notch filters are sequentially cascaded before the spline lookup table (LUT). The node spacing of the LUT is preset, and the interpolation method is linear or cubic spline, maintained separately for each temperature range. For the most prominent second and third spatial harmonic errors in the circular inductive synchronizer, adaptive notch filters are used for fixed-point elimination. Compared to conventional low-pass filtering, the notch filter can retain the full bandwidth of the baseband signal while filtering out specific harmonics, and by setting a specific step size, adaptive convergence performance is ensured, avoiding phase lag of the useful signal. Sequentially cascading the notch filters before the LUT follows the optimal signal processing logic of first removing harmonics and smoothing, then performing nonlinear fitting. This prevents residual harmonic components from contaminating the spline interpolation nodes, avoids harmonic aliasing or overfitting that may be caused by spline compensation, and improves the accuracy of nonlinear compensation. By employing linear or cubic spline interpolation with finite node intervals, this invention accurately approximates the complex high-order spatial nonlinearity errors of the inductive synchro while ensuring high real-time performance (replacing complex real-time calculations with table lookup). Maintaining spline tables separately for each temperature range overcomes the effects of thermal expansion and contraction of mechanical structures and changes in electrical parameters caused by temperature drift, ensuring compensation fidelity across the entire temperature range. Furthermore, this invention places second and third harmonic adaptive notch filtering before finite-node spline compensation, ensuring that the spline lookup surface corresponds to the residual nonlinearity after harmonic interference has been weakened. This reduces the size of the lookup table nodes and the fitting burden, while improving compensation consistency across different temperature zones.

[0050] Preferably, the step of obtaining the first angle and the second angle further includes: The first angle is calculated using the CORDIC module based on the first path signal in the orthogonal signals. The calculation expression is as follows:

[0051] In the formula, , These are the orthogonal components along the y-axis and the orthogonal components along the x-axis in the baseband orthogonal components, respectively. The second angle is calculated using the PI-PLL module based on the second path signal in the orthogonal signals.

[0052] Employing the CORDIC algorithm, the arctangent angle can be directly calculated through simple shifting and addition iterations, eliminating the need for complex multiplication and division operations. This parallel pipeline structure results in extremely low computational latency, enabling rapid, lag-free tracking of high-frequency dynamic changes. While retaining the fast response path of CORDIC, a parallel PI-PLL (Proportional-Integral Phase-Locked Loop) path is introduced. The inherent loop filtering characteristics of the PLL provide strong narrowband anti-interference capabilities, effectively smoothing high-frequency noise. This dual-path design provides a foundation for subsequent mode selection or fusion based on health indicators, balancing the system's dynamic response speed and steady-state accuracy.

[0053] Second Embodiment See Figure 8 A second aspect of the present invention provides a closed-loop decoding device for a circular induction synchronizer, comprising: The excitation generation module is used to apply a dual-carrier in-phase excitation signal to the circular induction synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The sampling and demodulation module is used to synchronously sample the induced signals output by the two orthogonal induction coils, and to obtain the baseband quadrature components by performing phase-sensitive demodulation and low-pass filtering on the first and second carrier frequency signals through a phase reference. The joint correction module is used to construct joint observations based on the baseband quadrature components, iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels using the joint observations, and decouple and correct the baseband quadrature components based on the estimation results to obtain the corrected quadrature signal. The calculation and compensation module performs angle calculation and error compensation based on the corrected orthogonal signal to obtain the angle measurement value of the circular induction synchronizer.

[0054] This device eliminates time delay and phase asynchrony errors between channels at the physical layer source by using dual-carrier in-phase excitation and synchronous sampling and demodulation based on the same phase reference, ensuring high-fidelity acquisition of the original signal. By constructing joint observations, it estimates and corrects gain mismatch and phase deviation between the two channels in real time online, fundamentally eliminating the inherent defects of orthogonal signals and significantly improving the original accuracy of angle calculation. The tightly integrated functional modules form a complete closed-loop architecture of excitation, sampling, correction, and calculation, enabling the system to continuously self-correct errors under complex operating conditions, achieving highly robust and reliable absolute angle decoding.

[0055] See details Figure 8 In the circuit diagram of this device, the LDO powers the SIP6117 processor. The 2-axis angle conversion chip generates an excitation signal, which is amplified by an instrumentation amplifier and then applied to the 7-circular induction synchronizer to generate excitation and a feedback signal. The signal is then filtered and amplified by a preamplifier module, a precision operational amplifier module, and a clamping circuit module. The processed signal is then sent to the axis angle conversion chip for axis angle conversion, outputting a digital code value. Finally, the digital code value is sent to the SIP6117 processor for algorithm correction.

[0056] In the description of this application, it should be noted that the terms "inner" and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product is in use. They are used only for the convenience of describing this application and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0057] It should also be noted that, unless otherwise explicitly specified and limited, the terms "setup" and "connection" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0058] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific identification content executed by the system and device described above can be referred to the corresponding process in the foregoing method embodiments.

[0059] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they shall still fall within the protection scope of the present invention.

Claims

1. A closed-loop decoding method for a circular inductive synchro, characterized in that, include: A dual-carrier in-phase excitation signal is applied to the circular inductive synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The induced signals output from the two orthogonal induction coils are synchronously sampled, and the first carrier frequency signal and the second carrier frequency signal are phase-sensitively demodulated and low-pass filtered through the phase reference to obtain the baseband quadrature components. Based on the baseband quadrature components, a joint observation is constructed. The joint observation is used to iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels. Based on the estimation results, the baseband quadrature components are decoupled and corrected to obtain the corrected quadrature signal. Angle calculation and error compensation are performed based on the corrected orthogonal signal to obtain the angle measurement value of the circular induction synchronizer.

2. The closed-loop decoding method for a circular inductive synchronizer according to claim 1, characterized in that, Also includes: Based on the phase change trend of the angle measurements and the zero-crossing count results, construct an absolute angle index and an incremental angle index; Based on the aforementioned angle and a pre-built set of health indicators, the parameters of the new decoupling matrix and spline interpolation compensation are updated online.

3. The closed-loop decoding method for a circular inductive synchro according to claim 2, characterized in that, The steps of constructing the absolute angle index and the incremental index further include: Based on the fusion perspective, the initial discrimination result of the number of pole pairs at the current moment is obtained by analyzing the phase change trend, counting the number of zero crossings, and calculating the phase circulation. A voting window is set up to conduct a majority vote on the initial results of the extreme logarithms, and the results are filtered by combining the residence time threshold to obtain the extreme logarithm discrimination results. When the pole-log number discrimination result is obtained by identifying two consecutive opposite decision results, the consistency between the phase circulation and the zero-crossing count is re-evaluated to obtain the evaluation result, and the pole-log number discrimination result is updated according to the evaluation result.

4. The closed-loop decoding method for the circular induction synchronizer according to claim 3, characterized in that, The step of re-evaluating the consistency between the phase loop quantity and the zero-crossing count further includes: When two consecutive opposite extreme logarithm determination results are obtained, the abnormal reconfirmation mechanism is triggered, pausing the output of the extreme logarithm determination result at the current moment and maintaining the extreme logarithm determination result at the previous moment; Recalculate the phase loop quantity and the zero-crossing count, and perform a consistency check. If the verification passes, the updated extreme logarithm discrimination result is output; if the verification fails, the extreme logarithm discrimination result of the previous time step is maintained, and the abnormal status bit of the health indicator is marked.

5. The closed-loop decoding method for a circular inductive synchro according to claim 1, characterized in that, The steps for constructing a joint observation measurement further include: The baseband quadrature component corresponding to the first carrier frequency is set as the main state tracking channel, and the baseband quadrature component corresponding to the second carrier frequency is set as the auxiliary observation channel. A joint error vector is calculated based on the data from the auxiliary observation channel. The joint error vector includes a static bias component and a dynamic drift component. The joint error vector is used as the objective function constraint term of the recursive least squares algorithm to perform joint iterative estimation of the gain mismatch and phase deviation of the master state tracking channel.

6. The closed-loop decoding method for a circular inductive synchro according to claim 5, characterized in that, The steps for obtaining orthogonal signals further include: Set the leakage factor; Construct the initial covariance matrix, the calculation expression is: ； In the formula, α is the uncertainty scaling factor. I It is the identity matrix; Based on the initial covariance matrix and the leakage factor, the gain vector, covariance matrix, channel gain ratio, and phase deviation are calculated through recursive least squares iteration. Construct the decoupling matrix and calculate the orthogonal signals. The calculation expression is as follows: ； ； In the formula, For orthogonal components, The orthogonal signal, For decoupling matrix, , These are the channel gain ratio and the phase deviation, respectively.

7. The closed-loop decoding method for a circular inductive synchro according to claim 1, characterized in that, The steps for obtaining the angle of the circular induction synchronizer further include: Based on the orthogonal signals, a compensated orthogonal signal is obtained through harmonic-spline composite processing; Based on the orthogonal signals, the first angle and the second angle are obtained by executing an angle calculation path; Within a preset statistical time window, calculate the noise spectral density, harmonic amplitude ratio, saturation duty cycle and temperature drift of the programmable gain amplifier, and construct health indicators. Based on the health indicators, a state machine with hysteresis is used to select the optimal mode, which includes CORDIC mode, PI-PLL mode and fusion mode, and obtain the corresponding angle.

8. The closed-loop decoding method for a circular inductive synchro according to claim 7, characterized in that, The step of obtaining the compensated orthogonal signal through harmonic-spline composite processing further includes: The filtered signal is obtained by adaptive notch filtering of the second and third harmonics based on the orthogonal signals; Based on the filtered signal, nonlinear compensation is performed using finite node spline search, linear or cubic spline interpolation, to obtain the compensated orthogonal signal.

9. The closed-loop decoding method for a circular inductive synchro according to claim 7, characterized in that, The steps for obtaining the first angle and the second angle further include: The first angle is calculated using the CORDIC module based on the first path signal in the orthogonal signals. The calculation expression is as follows: ； In the formula, , These are the orthogonal components along the y-axis and the orthogonal components along the x-axis in the baseband orthogonal components, respectively. The second angle is calculated using the PI-PLL module based on the second path signal in the orthogonal signals.

10. A closed-loop decoding device for a circular inductive synchro, characterized in that, include: An excitation generation module is used to apply a dual-carrier in-phase excitation signal to the circular inductive synchronizer. The dual-carrier in-phase excitation signal includes a first carrier frequency signal and a second carrier frequency signal, and the first carrier frequency signal and the second carrier frequency signal have the same phase reference. The sampling and demodulation module is used to synchronously sample the induced signals output by the two orthogonal induction coils, and to perform phase-sensitive demodulation and low-pass filtering on the first carrier frequency signal and the second carrier frequency signal through the phase reference to obtain the baseband quadrature components. The joint correction module is used to construct a joint observation based on the baseband quadrature components, use the joint observation to iteratively estimate the gain mismatch and phase deviation between the two quadrature induction signal channels, and decouple and correct the baseband quadrature components according to the estimation results to obtain the corrected quadrature signal. The calculation and compensation module performs angle calculation and error compensation based on the corrected orthogonal signal to obtain the angle measurement value of the circular induction synchronizer.

Images