A laser wavelength stable output method based on polarization shift interference demodulation

Abstract

The application discloses a laser wavelength stable output method based on polarization phase shift interference demodulation, which comprises the following steps: generating four light paths with phase of 0, 90, 180 and 270 degrees through polarization phase shift interference from the output wavelength of a laser; pre-processing data, filtering high-frequency signals by using a FIR filter, and then using waveform calibration to make the mean value and amplitude of the four light paths consistent; calculating wrapped phase by using an arctangent function based on a Cordic algorithm; restoring the wrapped phase into continuous real phase by using a threshold judgment phase unwrapping algorithm; inputting the current real phase obtained by calculation and the phase setting value of the system into a PID controller to calculate the final control amount and feedback control the laser. The application has low cost, integrates laser wavelength monitoring and wavelength tuning, and realizes laser wavelength monitoring, wavelength stable control and smooth phase shift.

Description

Technical Field

[0001] This invention belongs to the field of laser technology, specifically relating to a laser wavelength stabilization output method based on polarization phase-shifting interference demodulation. Background Technology

[0002] Laser wavelength measurement plays a crucial role in many applications involving laser sources, such as optical communication, laser spectroscopy, and optical sensing. Wavelength phase-shifting interferometers are essential instruments for measuring wavefront aberrations in optical components and systems. By controlling the output wavelength of a tunable laser, the interferometer can achieve phase-shift modulation of the interferogram and obtain wavefront aberration information through phase-shifting algorithms.

[0003] For phase-shifting interferometers, the accuracy of the phase shift affects the measurement accuracy, which is related to the stability of the laser wavelength and the stability of the interferometer cavity. In high-precision wavelength phase shifting, it is necessary not only to measure the laser wavelength in real time, but also to keep the laser wavelength stable during the phase shifting process.

[0004] Typically, monitoring the wavelength of a laser is separate from phase shifting. During phase shifting, the real-time wavelength of the laser can be detected using a wavelength meter. Most wavelength meters are also based on the principle of interference, such as the Michelson interferometer and the Fizeau interferometer. However, the resolution of a wavelength meter does not match the laser tuning resolution of a wavelength phase-shifting interferometer, making wavelength measurement during phase shifting difficult.

[0005] Pound-Drever-Hall (PDH) based on an ultra-stable optical reference cavity is a highly effective solution due to its high frequency discrimination sensitivity, allowing for better control of the laser's frequency (wavelength). However, systems implemented using this solution are relatively complex and employ active electro-optic modulators. In the industrial laser field, the price of electro-optic modulators is roughly equivalent to that of single-frequency lasers, severely limiting the market acceptance of this solution in industrial laser applications. Summary of the Invention

[0006] The purpose of this invention is to provide a low-cost laser wavelength stabilization output method based on polarization phase-shifting interference demodulation, which integrates laser wavelength monitoring and wavelength tuning to achieve the functions of laser wavelength monitoring, wavelength stabilization control, and smooth phase shifting.

[0007] The technical solution to achieve the purpose of this invention is: a laser wavelength stabilization output method based on polarization phase-shifting interference demodulation, comprising the following steps:

[0008] Step S1: The output wavelength of the laser is used to generate four optical paths with phases of 0, 90, 180, and 270 degrees respectively through polarization phase-shifting interference.

[0009] Step S2: Data preprocessing, using an FIR filter to filter high-frequency signals, and then using waveform calibration to calibrate the mean and amplitude of the four optical paths to be consistent.

[0010] Step S3: Calculate the wrap phase using the arctangent function based on the Cordic algorithm;

[0011] Step S4: Use a threshold-based phase unpacking algorithm to restore the wrapped phase to a continuous true phase;

[0012] Step S5: Input the calculated current real phase and the system phase setpoint to the PID controller to calculate the final control quantity and provide feedback control to the laser.

[0013] Compared with the prior art, the significant advantages of this invention are: (1) The phase-shifting interferometric demodulation algorithm adopts a fast implementation architecture based on FPGA, including data preprocessing with filtering and calibration, calculation of arctangent function based on Cordic algorithm, phase unpacking and discrete digital PID closed-loop control, which realizes long-term stable output of laser wavelength and achieves high-precision, low-cost, fast and stable wavelength tuning; (2) No active devices such as electro-optic modulators are required in the system, and only passive original devices such as fiber polarization devices are used, which greatly reduces the system cost and is conducive to its application in industrial lasers; (3) It has good versatility and is suitable for various lasers. Attached Figure Description

[0014] Figure 1 This is a flowchart of the laser wavelength stabilization output method based on polarization phase-shifting interference demodulation according to the present invention.

[0015] Figure 2 This is a schematic diagram of the optical path principle of polarization phase-shifting interference of the present invention.

[0016] Figure 3 This is a phase change diagram after closed-loop control is activated.

[0017] Figure 4 This is a diagram showing the phase change of a laser after it is subjected to external interference.

[0018] Figure 5 It is a phase change diagram during the phase shifting process. Detailed Implementation

[0019] This invention discloses a laser wavelength stabilization output method based on polarization phase-shifting interference demodulation, comprising the following steps:

[0020] Step S1: The output wavelength of the laser is used to generate four optical paths with phases of 0, 90, 180, and 270 degrees respectively through polarization phase-shifting interference.

[0021] Step S2: Data preprocessing, using an FIR filter to filter high-frequency signals, and then using waveform calibration to calibrate the mean and amplitude of the four optical paths to be consistent.

[0022] Step S3: Calculate the wrap phase using the arctangent function based on the Cordic algorithm;

[0023] Step S4: Use a threshold-based phase unpacking algorithm to restore the wrapped phase to a continuous true phase;

[0024] Step S5: Input the calculated current real phase and the system phase setpoint to the PID controller to calculate the final control quantity and provide feedback control to the laser.

[0025] As a specific example, in step S1, the four optical paths are generated based on the polarization phase-shifting interferometry method of Mach-Zehnder interferometry, and the waveform formula is as follows:

[0026]

[0027] Among them, I n A represents the light intensity level of the nth light source. n B is the mean of the nth signal. n Let n be the amplitude of the nth signal. This represents the phase information contained in the nth signal.

[0028] As a specific example, in step S2, the FIR filter uses a set number of tap coefficients h(t) to convolve with the input signal x(t) in the time domain to filter high-frequency noise;

[0029] The convolution process is represented as:

[0030]

[0031] Where x(t) is the input signal, h(t) is the tap coefficient, y(t) is the filtered output signal, and n is the filter order; the equiripple method is chosen to calculate the FIR coefficients, and the minimum order tap coefficient is selected.

[0032] As a specific example, in step S2, waveform calibration uses the 0th optical path as a reference and performs linear scaling and translation on all optical paths to ensure that the mean and amplitude of all optical paths are consistent. The calibrated waveform is as follows:

[0033]

[0034] Among them, I n B0 represents the light intensity level of the nth light source after calibration, and B0 is the amplitude of the 0th light source. This refers to the phase information contained in the signal.

[0035] As a specific example, in step S3, the method for calculating the wrap phase is based on the arctangent function of the Cordic algorithm, as follows:

[0036] For a given coordinate point (x, y), using displacement and addition / subtraction operations, the iterative formula is:

[0037] Rotate clockwise, i.e., y i-1 >0

[0038] x i <=x i-1 +(y i-1 >>i)

[0039] y i <=y i-1 -(x i-1 >>i)

[0040] z i <=z i-1 +θ i

[0041] Rotate clockwise, i.e., y i-1 <0 time

[0042] x i <=x i-1 -(y i-1 >>i)

[0043] y i <=y i-1 +(x i-1 >>i)

[0044] z i <=z i-1 -θ i

[0045] Where i represents the i-th iteration, x i-1 and y i-1 Let x represent the x and y coordinates before the i-th iteration. i and y i Let θ represent the x and y coordinates after the i-th iteration. i z represents the change in phase during the i-th iteration. i-1 and z i This represents the phase values ​​before and after the i-th iteration;

[0046] After the iteration is completed, the final z i The value is the angle corresponding to the initial point (x, y).

[0047] As a specific example, in step S4, the method for restoring the wrapped phase to a continuous true phase is based on a threshold judgment algorithm, and the calculation formula is as follows:

[0048]

[0049] in, Given the input package phase sequence, The output is the true phase sequence;

[0050] k is an integer, if the wrapping phase If the gradient is greater than π, then k is decreased by 1; if the phase is wrapped If the gradient is less than -π, then k is incremented by 1; otherwise, the value of k remains unchanged.

[0051] As a specific example, the PID controller in step S5 is a discrete PID controller based on FPGA. The relationship between the input error signal e[n] and the output u[n] is as follows:

[0052]

[0053] Where K p This is called the proportionality coefficient, K. i Called the integral coefficient, K d It is called the differential coefficient; e[n] is the error value at time n, which is obtained by subtracting the set value from the current true value, and u[n] is the output control quantity of the PID system.

[0054] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0055] Example

[0056] Combination Figure 1 Real-time laser wavelength stabilization based on polarization phase-shifting interference demodulation includes:

[0057] Step S1 involves using polarization phase-shifting interferometry to convert the laser's wavelength information into phase information. Combined with... Figure 2 The laser beam first passes through a collimating lens, then is split into two paths by a polarization beam splitter (PBS). One path passes through an adjustable attenuator. The other path passes through a 3-meter delay fiber interferometer, producing Mach-Zehnder interference. The two beams converge to a polarization beam combiner (PBC), and then are focused by a lens onto a quarter-wave plate (QWP). Four small holes are made in the aperture, each fitted with polarizers at 0, 45, 90, and 135 degrees for reception in the four quadrants of the detector. Thus, four optical paths with phases of 0, 90, 180, and 270 degrees are obtained through the four holes, and finally converted into electrical signals by a photodetector. The waveforms of the four signals are shown below:

[0058]

[0059] Where A n B is the mean of the signal, representing the background light intensity value. n The amplitude of the signal represents the intensity of the modulated light.

[0060] Step S2: First, all optical paths are passed through an FIR filter to filter out high-frequency noise signals. The FIR filter uses a finite number of tap coefficients h, and in the time domain, it is convolved with the input signal x(t). The convolution process is expressed as:

[0061]

[0062] The method chosen for calculating the FIR coefficients is the equiripple method, and the tap coefficients are selected as the minimum order.

[0063] Due to the A of 4-way light n With B n They may not be the same. We take the mean value A0 and amplitude B0 of the 0th optical path (I0) as a reference, and calibrate the mean and amplitude of all optical paths to be consistent. The calibration formula is:

[0064]

[0065] After waveform calibration, the waveforms of all optical paths are represented as follows:

[0066]

[0067] Step S3: After data preprocessing, the phase angle is further calculated.

[0068]

[0069] Express it using x and y:

[0070] x = I0 - I2

[0071] y = I3 - I1

[0072] This is transformed into finding the arctangent angle for a given coordinate (x, y). The Cordic (Coordinate Rotation Digital Computer) algorithm is used to solve for the arctangent angle, obtaining the wrap-around phase. The iterative formula for the Cordic algorithm is:

[0073] Rotate clockwise, i.e., y i-1 >0

[0074] x i <=x i-1 +(y i-1 >>i)

[0075] y i <=y i-1 -(x i-1 >>i)

[0076] z i <=z i-1 +θ i

[0077] Rotate clockwise, i.e., y i-1 <0 time

[0078] x i <=x i-1 -(y i-1 >>i)

[0079] y i <=y i-1 +(x i-1 >>i)

[0080] z i <=z i-1 -θ i

[0081] Different initial phase angles z0 are set based on the position of the initial point, and the calculation is performed iteratively a certain number of times until the final z... n The value is the angle corresponding to the initial point (x, y).

[0082] Step S4: Use a threshold judgment algorithm to recover the continuous true phase from the discontinuous wrapped phase. The principle is that when the phase gradient exceeds π, the current phase is added to or subtracted from an integer multiple of 2π to compensate for the phase discontinuity caused by the arctangent calculation. The calculation formula is as follows:

[0083]

[0084] in Given the input package phase sequence, This is the output true phase sequence. k is an integer, representing the phase sequence. If the gradient is greater than π, then k is decreased by one; if the phase is wrapped If the gradient is less than -π, then k is incremented by one; otherwise, the value of k remains unchanged.

[0085] Step S5: Use PID (Proportional, Integral, Differential) to continuously control the laser through feedback. For discrete PID control, let e[n] = s[n] - x[n], where s[n] is the system setpoint and x[n] is the current measured value of the system. Then, in the discrete case, the output value of the PID is...

[0086]

[0087] Among them, K p This is called the proportionality coefficient, K. i Called the integral coefficient, K d These are called differential coefficients.

[0088] Figure 3 This is the phase change diagram after closed-loop control is activated. Within a 500s timeframe, the phase exhibits a small jitter, with an amplitude of 0.0997 rad, corresponding to a wavelength change of 0.00142 pm. Simultaneously, the phase remains stable around 0. Within 500s, the wavelength instability under closed-loop control is 2.2 × 10⁻⁶. -10 .

[0089] Figure 4 This diagram shows the phase change of the laser after it is subjected to external interference, demonstrating the laser's rapid recovery. Thanks to FPGA-based real-time phase calculation and feedback control, the disturbance recovery time can be controlled within 5ms.

[0090] Figure 5 It is a phase change diagram during the phase shifting process. The phase of the laser can smoothly and quickly follow the set value.

[0091] In summary, this invention achieves long-term stable output of laser wavelength, realizing high-precision, low-cost, fast, and stable wavelength tuning. Furthermore, this solution is applicable to various lasers, exhibiting good versatility.

Claims

1. A laser wavelength stabilization output method based on polarization phase-shifting interference demodulation, characterized in that, The steps are as follows: Step S1: The output wavelength of the laser is used to generate four optical paths with phases of 0, 90, 180, and 270 degrees respectively through polarization phase-shifting interference. Step S2: Data preprocessing, using an FIR filter to filter high-frequency signals, and then using waveform calibration to calibrate the mean and amplitude of the four optical paths to be consistent. Step S3: Calculate the wrap phase using the arctangent function based on the Cordic algorithm; Step S4: Use a threshold-based phase unpacking algorithm to restore the wrapped phase to a continuous true phase; Step S5: Input the calculated current real phase and the system phase setpoint to the PID controller to calculate the final control quantity and feed it back to control the laser. In step S1, the four optical paths are generated based on the polarization phase-shifting interferometry method of Mach-Zehnder interferometry, and the waveform formula is as follows: Among them, I n A represents the light intensity level of the nth light source. n B is the mean of the nth signal. n Let n be the amplitude of the nth signal. The phase information contained in the nth signal; In step S2, the FIR filter uses a set number of tap coefficients h(t) to convolve with the input signal x(t) in the time domain to filter high-frequency noise. The convolution process is represented as: Where x(t) is the input signal, h(t) is the tap coefficient, y(t) is the filtered output signal, and n is the filter order; the equiripple method is chosen to calculate the FIR coefficients, and the minimum order tap coefficient is selected; In step S2, waveform calibration uses the 0th optical path as a reference and performs linear scaling and translation on all optical paths to ensure that the mean and amplitude of all optical paths are consistent. The calibrated waveform is as follows: Among them, I n This represents the light intensity level of the nth light source after calibration, where A0 is the mean value of the 0th signal and B0 is the amplitude of the 0th light source. This refers to the phase information contained in the signal; In step S3, the method for calculating the wrap phase is based on the arctangent function of the Cordic algorithm, as follows: For a given coordinate point (x, y), using displacement and addition / subtraction operations, the iterative formula is: Rotate clockwise, i.e., y i-1 >0 x i <=x i-1 +(y i-1 >>i) and i <=and i-1 -(x i-1 >>i) With i <=z i-1 +θ i Rotate clockwise, i.e., y i-1 <0 o'clock x i <=x i-1 -(y i-1 >>i) and i <=and i-1 +(x i-1 >>i) With i <=z i-1 -θ i Where i represents the i-th iteration, x i-1 and y i-1 Let x represent the x and y coordinates before the i-th iteration. i and y i Let θ represent the x and y coordinates after the i-th iteration. i z represents the change in phase during the i-th iteration. i-1 and z i This represents the phase values ​​before and after the i-th iteration; After the iteration is completed, the final z i The value is the angle corresponding to the initial point (x, y); In step S4, the method for restoring the wrapped phase to a continuous true phase is based on a threshold judgment algorithm, and the calculation formula is as follows: in, Given the input package phase sequence, The output is the true phase sequence; k is an integer, if the wrapping phase If the gradient is greater than π, then k is decreased by 1; if the phase is wrapped If the gradient is less than -π, then k is incremented by 1; otherwise, the value of k remains unchanged. The PID controller in step S5 is a discrete PID controller based on FPGA. The relationship between the input error signal e[n] and the output u[n] is as follows: Where K p This is called the proportionality coefficient, K. i The integral coefficient, K d It is called the differential coefficient; e[n] is the error value at time n, which is obtained by subtracting the set value from the current true value, and u[n] is the output control quantity of the PID system.

Images