An interference signal phase extraction method and system combining DPLL with Goertzel

Abstract

The application discloses an interference signal phase extraction method and system combined with DPLL and Goertzel, and belongs to the technical field of precision angle measurement. The method takes real-time frequency tracking of DPLL as a dynamic reference, and uses Goertzel frequency domain solution to optimize the precision of a traditional digital phase-locked loop, so that fast tracking of a variable frequency signal and high-precision phase extraction of a steady-state signal are realized in the same processing architecture. Through a frequency change rate adaptive switching mechanism, the tracking continuity of a fast variable frequency scene and the solution precision of a slow variable frequency and steady-state scene are taken into account. While retaining the advantages of simple structure and easy hardware implementation of DPLL, the method significantly improves the interference signal phase extraction precision and the full-working-condition adaptation capability, effectively solves the technical contradiction that the traditional DPLL cannot simultaneously consider dynamic tracking performance and steady-state measurement precision, and can be widely applied to the precise demodulation scene of interference signals of ultra-precision machine tools, nanometer three-coordinate measuring machines and other equipment.

Description

Technical Field

[0001] This invention relates to a method and system for phase extraction of interferometric signals combining DPLL and Goertzel, belonging to the field of precision angle measurement technology. Background Technology

[0002] With the rapid development of high-tech fields such as aerospace, high-end equipment manufacturing, semiconductor lithography, and precision optical inspection, the phase extraction accuracy of interference signals directly determines the core performance of ultra-precision measurement systems, precision vibration monitoring systems, and high-speed communication demodulation systems. In typical application scenarios such as ultra-precision machine tool spindle error compensation and nanometer coordinate measuring machine calibration, the measured interference signals often exhibit complex dynamic characteristics such as fast frequency conversion, slow frequency conversion, and steady-state single frequency, which places dual demands on the dynamic tracking capability and steady-state measurement accuracy of phase extraction technology.

[0003] In existing technologies, digital phase-locked loops (DPLLs), with their closed-loop negative feedback architecture, possess excellent fast frequency conversion signal tracking capabilities and dynamic response characteristics, enabling real-time locking of the input signal's frequency and phase. However, the phase output performance of DPLLs is affected by multiple factors such as loop filter bandwidth, PI controller parameters, and loop noise accumulation. This makes it impossible to improve phase extraction accuracy in steady-state scenarios without sacrificing dynamic response. In slow frequency conversion and steady-state single-frequency signal measurement scenarios, phase jitter is significant, creating a technical contradiction where "dynamic tracking" and "steady-state accuracy" are mutually exclusive (Nguyen TVH, Pham CK. An Overview of Phase-Locked Loop: From Fundamentals to the Frontier. Sensors(Basel). 2025 Sep 9;25(18):5623.), limiting its application in precision measurement across all scenarios.

[0004] In summary, existing DPLL phase extraction technology suffers from a core technical contradiction where dynamic tracking and steady-state accuracy cannot be simultaneously achieved. Existing optimization schemes cannot fundamentally overcome the inherent defects of closed-loop systems. Therefore, developing an interferometric signal phase extraction method that can significantly improve steady-state phase extraction accuracy while retaining the advantages of DPLL dynamic tracking, and achieving high-performance phase measurement in all scenarios, is of significant practical importance and engineering value for breaking through the technical bottlenecks in the field of ultra-precision measurement and promoting the upgrading of traditional DPLL technology. Summary of the Invention

[0005] This invention aims to solve the problem that traditional DPLLs cannot simultaneously achieve both dynamic tracking performance and steady-state measurement accuracy, and proposes an interferometric signal phase extraction method and system that combines DPLL and Goertzel.

[0006] The technical solution of the present invention:

[0007] A method for phase extraction of interferometric signals combining DPLL and Goertzel includes the following steps:

[0008] S1. Acquire the input analog interference signal, perform analog-to-digital conversion and preprocessing on the analog interference signal to obtain a digital interference signal;

[0009] S2. Input the digital interference signal into the DPLL frequency tracking module, and the DPLL frequency tracking module performs closed-loop frequency tracking and phase tracking on the digital interference signal, and outputs the real-time tracking frequency and DPLL tracking phase.

[0010] S3. Input the digital interference signal into the Goertzel phase calculation module, and use the real-time tracking frequency as the target frequency reference of the Goertzel phase calculation module. The Goertzel phase calculation module performs sliding window iterative calculation on the digital interference signal, extracts the frequency domain complex amplitude corresponding to the target frequency point, and calculates the Goertzel calculated phase based on the frequency domain complex amplitude.

[0011] S4. Calculate the real-time frequency change of the interference signal based on the real-time tracking frequency, compare the real-time frequency change with a preset scene judgment threshold, and generate a switching control signal based on the comparison result.

[0012] S5. Based on the switching control signal, select one of the DPLL tracking phase and the Goertzel solution phase as the final interference signal phase extraction result output.

[0013] Specifically, in step S1, the preprocessing includes bandpass noise reduction and amplitude normalization.

[0014] Specifically, in step S2, the DPLL frequency tracking module adopts a fully digital phase-locked loop architecture and sequentially performs quadrature phase detection, low-pass filtering, PI control, and numerically controlled oscillation processing to achieve closed-loop frequency tracking and phase tracking of the digital interference signal.

[0015] The orthogonal phase detector is used to extract the phase difference of the digital interference signal, the low-pass filter is used to filter out high-frequency noise components, the PI control is used to complete closed-loop error adjustment, and the numerically controlled oscillation is used to complete real-time updates of frequency and phase.

[0016] Specifically, the DPLL frequency tracking module calculates the phase difference at the current sampling moment based on the in-phase and quadrature components after low-pass filtering, and updates the real-time tracking frequency and DPLL tracking phase of the numerically controlled oscillator based on the phase difference via PI control. .

[0017] Specifically, in step S3, the Goertzel phase calculation module uses a sliding window of length N to update the digital interference signal point by point, and performs Goertzel iterative calculation based on the target digital angular frequency corresponding to the real-time tracking frequency.

[0018] The target digital angular frequency is:

[0019] ,

[0020] in, The real-time tracking frequency at the current sampling moment. This represents the signal sampling frequency.

[0021] Specifically, the Goertzel phase calculation module performs sliding window iterative calculations using the following recursive relationship:

[0022] ,

[0023] in, Let m be the signal sequence within the sliding window, where m = 1, 2, ..., N, and N is the length of the sliding window. , , These are intermediate variables in the Goertzel iteration process.

[0024] Specifically, the frequency domain complex amplitude and the original Goertzel phase are as follows:

[0025] ,

[0026] Where Y is the frequency domain complex amplitude corresponding to the target frequency point; This represents the argument of Y.

[0027] Specifically, the sliding window delay phase compensation and the Goertzel solution phase are as follows:

[0028] ,

[0029] in, The fixed phase delay introduced by the sliding window. For Goertzel's original phase, The phase is calculated for the compensated Goertzel solution.

[0030] Specifically, in steps S4 and S5, the real-time frequency change is calculated based on the real-time tracking frequency difference between adjacent sampling times, and the final interference signal phase extraction result is selected based on the comparison result of the real-time frequency change with a preset threshold.

[0031]

[0032] ,

[0033] in, This represents the real-time frequency change between adjacent sampling times. A threshold is preset for the rate of change of frequency. This is the phase extraction result of the interference signal finally output by the device.

[0034] An interferometric signal phase extraction system combining DPLL and Goertzel includes:

[0035] The signal acquisition and preprocessing module is used to acquire the input analog interference signal, perform analog-to-digital conversion and preprocessing on the analog interference signal to obtain a digital interference signal;

[0036] The DPLL frequency tracking module is used to perform closed-loop frequency tracking and phase tracking on the digital interference signal, and outputs the real-time tracking frequency and DPLL tracking phase.

[0037] The Goertzel phase calculation module is used to receive the digital interference signal, and with the real-time tracking frequency as the target frequency reference, perform sliding window iterative calculation on the digital interference signal to extract the frequency domain complex amplitude corresponding to the target frequency point, and calculate the Goertzel calculated phase based on the frequency domain complex amplitude.

[0038] An adaptive switching module is used to compare the real-time frequency change with a preset threshold and generate a switching control signal based on the comparison result.

[0039] The phase output module is used to select one of the DPLL tracking phase and the Goertzel solution phase as the final interference signal phase extraction result output according to the switching control signal;

[0040] Specifically, when the real-time frequency change is greater than or equal to the preset threshold, the phase output module selects the DPLL tracking phase as the final interference signal phase extraction result; when the real-time frequency change is less than the preset threshold, the phase output module selects the Goertzel solved phase as the final interference signal phase extraction result.

[0041] The beneficial effects of this invention are:

[0042] 1. The interference signal phase extraction method combining DPLL and Goertzel of the present invention is based on the principles of digital phase-locked loop closed-loop tracking, Goertzel frequency domain phase extraction, and scene adaptive switching. It can simultaneously adapt to all scenarios of fast frequency conversion, slow frequency conversion, and steady-state single frequency, and complete the real-time frequency tracking and high-precision phase extraction of interference signals.

[0043] 2. The method of this invention is based on the collaborative mechanism of DPLL providing a dynamic frequency reference for Goertzel, which enables closed-loop frequency tracking and high-precision phase calculation in the frequency domain to be realized simultaneously in the same processing architecture. The architecture is simple and compact, and has strong hardware feasibility.

[0044] 3. Based on the precise phase extraction of interference signals with different frequency conversion characteristics, the method of the present invention further eliminates the scene limitations of a single algorithm through a scene adaptive switching mechanism, which significantly improves the phase calculation accuracy under all working conditions.

[0045] 4. The method of the present invention achieves high-precision phase extraction in all scenarios while maintaining the advantages of simple module architecture and strong hardware portability, and can be widely used in engineering scenarios such as laser interferometry, communication signal demodulation, and precision vibration monitoring. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the method of the present invention;

[0047] Figure 2 Simulation results of phase output for DPLL and Goertzel algorithms under steady-state single-frequency conditions;

[0048] Figure 3 Simulation results of phase output for DPLL and Goertzel algorithms at a frequency conversion rate of 50kHz / s;

[0049] Figure 4 Simulation results of phase output for DPLL and Goertzel algorithms at a frequency conversion rate of 500 kHz / s;

[0050] In the diagram, 1-signal acquisition and preprocessing module, 2-DPLL frequency tracking module, 3-adaptive switching module, 4-phase output module, and 5-Goertzel phase calculation module. Detailed Implementation

[0051] Example 1:

[0052] This embodiment provides a method for phase extraction of interferometric signals combining DPLL and Goertzel, including the following steps:

[0053] Signal acquisition and preprocessing module 1 acquires and preprocesses the interference signal, outputting a digital interference signal that meets the algorithm requirements. This signal is then transmitted to DPLL frequency tracking module 2 and Goertzel phase calculation module 5, respectively. After receiving the digital interference signal, DPLL frequency tracking module 2 performs real-time frequency and phase tracking using a fully digital closed-loop phase-locked architecture. The output real-time tracking frequency is transmitted to Goertzel phase calculation module 5 and adaptive switching module 3, while the output real-time tracking phase is transmitted to phase output module 4. Goertzel phase calculation module 5 uses the real-time tracking frequency output by DPLL frequency tracking module 2 as a dynamic reference. After receiving the digital interference signal, it performs high-precision phase calculation in the frequency domain and outputs the Goertzel-calculated phase, which is then transmitted to phase output module 4. Adaptive switching module 3 receives the real-time tracking frequency output by DPLL frequency tracking module 2, calculates the rate of change of frequency, and determines the scene. It then outputs a switching control signal to phase output module 4. Phase output module 4 selects the corresponding phase source based on the switching control signal and outputs the final interference signal phase extraction result. The differences in phase measurement accuracy between the DPLL and Goertzel algorithms under different frequency conversion conditions are as follows: Figure 2 , Figure 3 and Figure 4 As shown, under fast frequency conversion conditions, the measurement accuracy of the DPLL algorithm is better than that of the Goertzel algorithm, while under slow frequency conversion and stable frequency conditions, the measurement accuracy of the Goertzel algorithm is better than that of the DPLL algorithm.

[0054] like Figure 1 As shown, when the input interference signal undergoes a frequency change, the digital interference signal output by the signal acquisition and preprocessing module 1... The DPLL frequency tracking module 2 and the Goertzel phase calculation module 5 are simultaneously entered. The DPLL frequency tracking module 2, based on the digital phase-locked loop principle, completes closed-loop frequency and phase tracking, and obtains the real-time tracking frequency. DPLL tracking phase ;

[0055]

[0056]

[0057]

[0058] in, The phase difference at the current sampling time; , These are the in-phase and quadrature components after low-pass filtering, respectively; , These are the proportional and integral coefficients of the PI controller, respectively. This is the output control quantity of the PI controller; This refers to the real-time output frequency of the numerically controlled oscillator. This represents the signal sampling frequency.

[0059] The Goertzel phase resolution module 5 uses the real-time tracking frequency output by the DPLL frequency tracking module 2. As a dynamic reference, frequency domain phase extraction is completed through sliding window iteration, and the original Goertzel phase is obtained respectively. The compensated Goertzel solution phase ;

[0060]

[0061]

[0062]

[0063] in, The digital angular frequency corresponding to the target frequency; Let m be the signal sequence within the sliding window, where m = 1, 2, …, N, and N is the length of the sliding window; , , is an intermediate variable in the Goertzel iteration process; Y is the frequency domain complex amplitude corresponding to the target frequency point; Indicates the argument of Y; The fixed phase delay introduced by the sliding window.

[0064] The adaptive switching module 3 determines the scene based on the frequency change rate and generates a switching control signal. The phase output module 4 selects the optimal phase output based on the switching control signal, first calculating the real-time change rate of the interference signal. Finally, scene determination and phase output switching are completed;

[0065]

[0066]

[0067] in, A preset threshold for the rate of change of frequency can be set, which can be adjusted according to the actual application scenario and parameters; This is the phase extraction result of the interference signal finally output by the device.

[0068] Example 2:

[0069] This embodiment provides an interferometric signal phase extraction system combining DPLL and Goertzel, such as Figure 1 As shown, it includes:

[0070] The module consists of a signal acquisition and preprocessing module 1, a DPLL frequency tracking module 2, a Goertzel phase calculation module 5, an adaptive switching module 3, and a phase output module 4.

[0071] The signal acquisition and preprocessing module 1 provides the basic signal source for the architecture. It has a built-in analog-to-digital conversion unit and a preprocessing unit. The implementation method is as follows: the analog-to-digital conversion unit performs synchronous sampling and digital conversion on the input analog interference signal, and then the preprocessing unit completes bandpass noise reduction and amplitude normalization processing. Finally, it outputs a digital interference signal that meets the requirements of subsequent algorithm processing and transmits it to the DPLL frequency tracking module.

[0072] The DPLL frequency tracking module 2 provides a dynamic frequency reference and closed-loop phase tracking for the architecture. It adopts a fully digital phase-locked loop architecture and is implemented as follows: it receives the digital interference signal output by the signal acquisition module, and sequentially passes it through an orthogonal phase detector unit to extract the signal phase difference, a low-pass filter unit to filter out high-frequency noise components, a PI control unit to adjust the closed-loop error, and a digitally controlled oscillator to update the frequency and phase in real time. Finally, it outputs the real-time tracking frequency and tracking phase. The tracking frequency is synchronously transmitted to the Goertzel phase calculation module and the adaptive switching module, and the tracking phase is synchronously transmitted to the phase output module.

[0073] The Goertzel phase calculation module 5 provides high-precision frequency domain phase calculation for the architecture. When the frequency reference is accurate, the Goertzel algorithm can effectively suppress noise and extract clean phase through sliding window frequency domain filtering. The implementation method is as follows: receiving the digital interference signal output by the signal acquisition module, and using the real-time tracking frequency output by the DPLL frequency tracking module as the target frequency reference, the signal is updated and iterated point by point through sliding window buffering. The frequency domain complex amplitude of the signal is extracted and the original phase is calculated. Then, the fixed phase delay caused by the sliding window is eliminated through the phase compensation unit. Finally, the high-precision Goertzel phase is output and synchronously transmitted to the phase output module.

[0074] The adaptive switching module 3 provides the architecture with scene-adaptive phase output selection capability. The implementation method is as follows: it receives the real-time tracking frequency output by the DPLL frequency tracking module, calculates the real-time frequency change rate of the interference signal, compares the frequency change rate with the preset scene judgment threshold, and outputs a switching control signal according to the comparison result: when it is determined to be a fast frequency conversion scene, the control selects the DPLL tracking phase as the output source; when it is determined to be a slow frequency conversion or steady-state single-frequency scene, the control selects the Goertzel solved phase as the output source, and the switching control signal is synchronously transmitted to the phase output module.

[0075] The phase output module 4 is the final result output unit of the architecture. It is implemented by receiving the switching control signal output by the adaptive switching module, selecting the DPLL tracking phase or the Goertzel phase calculation accordingly, and outputting the final interference signal phase extraction result after completing the standardized conversion of the output format. The module and the front-end units form a collaborative working mechanism. The DPLL frequency tracking module provides dynamic frequency support for the Goertzel phase calculation module, and the Goertzel phase calculation module supplements the accuracy shortcomings of DPLL in slow frequency conversion and steady-state scenarios. The two form a complementary and collaborative working mode.

Claims

1. A method for phase extraction of interferometric signals combining DPLL and Goertzel, characterized in that, Includes the following steps: S1. Acquire the input analog interference signal, perform analog-to-digital conversion and preprocessing on the analog interference signal to obtain a digital interference signal; S2. Input the digital interference signal into the DPLL frequency tracking module, and the DPLL frequency tracking module performs closed-loop frequency tracking and phase tracking on the digital interference signal, and outputs the real-time tracking frequency and DPLL tracking phase. S3. Input the digital interference signal into the Goertzel phase calculation module, and use the real-time tracking frequency as the target frequency reference of the Goertzel phase calculation module. The Goertzel phase calculation module performs sliding window iterative calculation on the digital interference signal, extracts the frequency domain complex amplitude corresponding to the target frequency point, and calculates the Goertzel calculated phase based on the frequency domain complex amplitude. S4. Calculate the real-time frequency change of the interference signal based on the real-time tracking frequency, compare the real-time frequency change with a preset scene judgment threshold, and generate a switching control signal based on the comparison result. S5. Based on the switching control signal, select one of the DPLL tracking phase and the Goertzel solution phase as the final interference signal phase extraction result output.

2. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 1, characterized in that: In step S1, the preprocessing includes bandpass noise reduction and amplitude normalization.

3. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 1, characterized in that: In step S2, the DPLL frequency tracking module adopts a fully digital phase-locked loop architecture and sequentially performs quadrature phase detection, low-pass filtering, PI control, and numerically controlled oscillation processing to achieve closed-loop frequency tracking and phase tracking of the digital interference signal. The orthogonal phase detector is used to extract the phase difference of the digital interference signal, the low-pass filter is used to filter out high-frequency noise components, the PI control is used to complete closed-loop error adjustment, and the numerically controlled oscillation is used to complete real-time updates of frequency and phase.

4. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 3, characterized in that: The DPLL frequency tracking module calculates the phase difference at the current sampling time based on the in-phase and quadrature components after low-pass filtering, and updates the real-time tracking frequency and DPLL tracking phase of the numerically controlled oscillator based on the phase difference via PI control. .

5. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 4, characterized in that: In step S3, the Goertzel phase calculation module uses a sliding window of length N to update the digital interference signal point by point, and performs Goertzel iterative calculation based on the target digital angular frequency corresponding to the real-time tracking frequency. The target digital angular frequency is: , in, The real-time tracking frequency at the current sampling moment. This is the signal sampling frequency.

6. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 5, characterized in that: The Goertzel phase calculation module performs sliding window iterative calculations using the following recursive relationship: , in, The signal sequence within the sliding window. m= 1,2,…, N , N The length of the sliding window; , , The first m After the first iteration, the three intermediate variables in Goertzel's recursion process are... , , This is the initial value for the iteration.

7. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 6, characterized in that: The frequency domain complex amplitude and the Goertzel original phase are respectively: ， Where Y is the frequency domain complex amplitude corresponding to the target frequency point; This represents the argument of Y.

8. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 7, characterized in that: The sliding window delay phase compensation and the Goertzel solution phase are as follows: ， in, The fixed phase delay introduced by the sliding window. For Goertzel's original phase, The phase is calculated for the compensated Goertzel solution.

9. The method for phase extraction of interferometric signals combining DPLL and Goertzel as described in claim 8, characterized in that: In steps S4 and S5, the real-time frequency change is calculated based on the real-time tracking frequency difference between adjacent sampling times, and the final interference signal phase extraction result is selected based on the comparison result of the real-time frequency change with a preset threshold. ， in, This represents the real-time frequency change between adjacent sampling times. A threshold is preset for the rate of change of frequency. This is the phase extraction result of the interference signal finally output by the device.

10. A phase extraction system for interferometric signals combining DPLL and Goertzel, characterized in that, include: The signal acquisition and preprocessing module is used to acquire the input analog interference signal, perform analog-to-digital conversion and preprocessing on the analog interference signal to obtain a digital interference signal; The DPLL frequency tracking module is used to perform closed-loop frequency tracking and phase tracking on the digital interference signal, and outputs the real-time frequency change and DPLL tracking phase. The Goertzel phase calculation module is used to receive the digital interference signal, and use the real-time frequency change as the target frequency reference to perform sliding window iterative calculation on the digital interference signal, extract the frequency domain complex amplitude corresponding to the target frequency point, and calculate the Goertzel calculated phase based on the frequency domain complex amplitude. An adaptive switching module is used to compare the real-time frequency change with a preset threshold and generate a switching control signal based on the comparison result. The phase output module is used to select one of the DPLL tracking phase and the Goertzel solution phase as the final interference signal phase extraction result output according to the switching control signal; Specifically, when the real-time frequency change is greater than or equal to the preset threshold, the phase output module selects the DPLL tracking phase as the final interference signal phase extraction result; when the real-time frequency change is less than the preset threshold, the phase output module selects the Goertzel solved phase as the final interference signal phase extraction result.

Images