CN114216575B - Ultrashort pulse reconstruction system and method based on BP neural network - Google Patents
Ultrashort pulse reconstruction system and method based on BP neural network Download PDFInfo
- Publication number
- CN114216575B CN114216575B CN202111383297.0A CN202111383297A CN114216575B CN 114216575 B CN114216575 B CN 114216575B CN 202111383297 A CN202111383297 A CN 202111383297A CN 114216575 B CN114216575 B CN 114216575B
- Authority
- CN
- China
- Prior art keywords
- pulse
- neural network
- line trace
- scanning spectrum
- detected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 33
- 238000001228 spectrum Methods 0.000 claims abstract description 47
- 239000013078 crystal Substances 0.000 claims abstract description 30
- 230000003287 optical effect Effects 0.000 claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000003062 neural network model Methods 0.000 claims abstract description 15
- 230000003111 delayed effect Effects 0.000 claims abstract description 4
- 238000010586 diagram Methods 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 6
- QBLDFAIABQKINO-UHFFFAOYSA-N barium borate Chemical group [Ba+2].[O-]B=O.[O-]B=O QBLDFAIABQKINO-UHFFFAOYSA-N 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 238000002474 experimental method Methods 0.000 abstract description 5
- 238000012512 characterization method Methods 0.000 abstract description 3
- 238000013135 deep learning Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 6
- 210000002569 neuron Anatomy 0.000 description 5
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- BQCADISMDOOEFD-UHFFFAOYSA-N Silver Chemical compound [Ag] BQCADISMDOOEFD-UHFFFAOYSA-N 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000005305 interferometry Methods 0.000 description 2
- 238000007747 plating Methods 0.000 description 2
- 229910052709 silver Inorganic materials 0.000 description 2
- 239000004332 silver Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000009407 construction method and process Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000009022 nonlinear effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000000523 sample Substances 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J9/00—Measuring optical phase difference; Determining degree of coherence; Measuring optical wavelength
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J11/00—Measuring the characteristics of individual optical pulses or of optical pulse trains
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/02—Details
- G01J3/0205—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
- G01J3/0208—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using focussing or collimating elements, e.g. lenses or mirrors; performing aberration correction
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/02—Details
- G01J3/0205—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows
- G01J3/021—Optical elements not provided otherwise, e.g. optical manifolds, diffusers, windows using plane or convex mirrors, parallel phase plates, or particular reflectors
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/42—Absorption spectrometry; Double beam spectrometry; Flicker spectrometry; Reflection spectrometry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/42—Absorption spectrometry; Double beam spectrometry; Flicker spectrometry; Reflection spectrometry
- G01J2003/423—Spectral arrangements using lasers, e.g. tunable
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Optical Modulation, Optical Deflection, Nonlinear Optics, Optical Demodulation, Optical Logic Elements (AREA)
Abstract
The invention belongs to the technical field of ultrashort pulse characterization, and relates to an ultrashort pulse reconstruction system and method based on a deep learning BP neural network. An ultrashort pulse reconstruction system based on a BP neural network comprises a double-beam autocorrelation optical path unit and a main control unit; the double-beam autocorrelation optical path unit is used for acquiring a delayed scanning spectrum line trace of the ultra-short pulse; the main control unit is used for controlling the system and processing data and comprises a control module and a data processing module; the data processing module comprises a BP neural network model; and the BP neural network model predicts an input scanning spectrum line trace graph and outputs the amplitude and the phase of the ultrashort pulse to be detected. According to the invention, the cutting angle of the BBO crystal is calculated according to different pulse wavelengths to be detected, so that the applicability of the system and the success rate of experiments are improved. And the BP neural network is combined to reconstruct pulse information to be detected, so that the ultra-short pulse reconstruction efficiency is greatly improved.
Description
Technical Field
The invention belongs to the technical field of ultrashort pulse characterization, and relates to an ultrashort pulse reconstruction system and method based on a deep learning BP neural network
Background
Since the advent of ultra-short pulse laser, the ultra-short pulse laser has irreplaceable important applications in a plurality of fields such as microscopic imaging, laser processing, medical treatment and the like, and has brought about wide attention and great interest to society, and has very critical significance in research in the quantum field.
The laser pulse width of the ultrashort pulse laser is very short, generally in the picosecond or even in the femtosecond order, when the pulse width is in the femtosecond order, the electronic response limit is exceeded, and the ultrashort pulse laser cannot be characterized by a general detector, so that it becomes very critical to find a method capable of accurately and efficiently representing the ultrashort pulse. The problem of ultrashort pulse characterization was solved to some extent by the proposal of autocorrelation method in the sixties of the twentieth century, and the following decades, the self-reference spectral interferometry, the frequency-resolved optical switching method, and the like, have been proposed successively. Compared with the former two methods, the frequency resolution optical switching method, especially the second harmonic frequency resolution optical switching method, has the characteristics of high resolution, high accuracy, simple optical path and the like, so that the method becomes the main stream method for representing the ultrashort pulse at present, but the method needs longer time to calculate the data result required by the reduction from the spectrum trace graph, cannot detect in real time, and has poor performance of low signal to noise ratio, so that the method is widely subject to the problem.
Disclosure of Invention
Aiming at the defects of low resolution and accuracy (an autocorrelation method), complex light path (a self-reference spectrum interferometry), low efficiency (a frequency resolution optical switching method), poor performance (all) of low signal to noise ratio and the like existing in the prior art, the invention provides a system for representing ultrashort pulses by using a double-reflection second harmonic frequency resolution optical switching method based on a BP neural network, wherein the optical part of the system is preferably formed by the double-reflection autocorrelation light path and a second harmonic nonlinear crystal, and the BP neural network is used for replacing a traditional inversion algorithm (such as PCGPA) to calculate and obtain amplitude phase data.
The technical scheme adopted for solving the technical problems is as follows: the ultra-short pulse reconstruction system based on the BP neural network is characterized in that: the system comprises a double-beam autocorrelation optical path unit and a main control unit; the double-beam autocorrelation optical path unit is used for acquiring a scanning spectrum line trace of the ultrashort pulse; the main control unit is used for controlling the system and processing data and comprises a control module and a data processing module; the data processing module comprises a BP neural network model; and the BP neural network model predicts an input scanning spectrum line trace graph and outputs the amplitude and the phase of the ultrashort pulse to be detected.
Preferably, the training set of the BP neural network model is composed of simulated scanning spectrum line trace diagrams, and the same group of noiseless pulse amplitude and phase correspond to the simulated scanning spectrum line trace diagrams with n different signal-to-noise ratios; n is more than or equal to 1 and less than or equal to 30.
Preferably, the simulated scanning spectrum line trace is obtained by the following steps:
(1) Collecting pulse amplitude data, and sampling N points of each group of collected pulse amplitude data, wherein N is more than or equal to 100 and less than or equal to 500;
(2) N data are selected from N sampling points of the pulse amplitude data, and random numeralization processing is carried out to different degrees;
(3) And (3) generating pulse amplitude data processed in the step (2) and pulse phase data generated by a function according to a formula:
in the middle ofThe method is characterized in that the method is a scanning spectrum of a second harmonic, E (t) is a pulse to be detected, E (t-tau) is a time-shifted copy of the pulse to be detected, namely a gate function, omega is the pulse frequency to be detected, and tau is a delay amount;
and calculating to obtain simulated scanning spectrum line trace data, and generating n simulated scanning spectrum line trace diagrams with different signal to noise ratios.
Preferably, the different degrees of randomization are: respectively taking different random numbers between 0 and 1 for the n selected pulse amplitude data; and the selected random number is different from the corresponding original value.
Preferably, the control module is configured to control a delay system of the dual-beam autocorrelation optical path unit to generate a delay.
Preferably, the control module is further used for controlling the spectrometer to continuously scan the frequency multiplication pulse spectrum while the delay system generates delay, and generating a delayed scanning spectrum linetrace.
Preferably, the nonlinear crystal in the double-beam autocorrelation optical path unit is barium metaborate crystal; the cutting angle of the barium metaborate crystal selects a main section collinear phase matching frequency multiplication angle, and the calculation method is as follows: let two pulses to be measured be k 1 And k 2 The frequency multiplication pulse is k 3 The phase matching condition under the main section condition is that wave vector superposition is satisfied, namely
Decomposing the vector to obtain a formula:
because the three pulses satisfy the vector superposition
Square adding the formulas (2) and (3), substitutingThe following steps are obtained:
from the frequency multiplication, ω 3 ＝2ω 1 ＝2ω 2 Therefore, it is
Wherein n is p (θ 3 ) For the included angle theta between the frequency doubling light and the main axis 3 Refractive index of time-doubled light; n is n p (θ 1 )、n p (θ 2 ) For the included angle between the pulse to be measured and the main shaft to be theta 1 、θ 2 The refractive index of the pulse to be measured;
the refractive index of the frequency doubling light e is calculated according to the formula:
wherein n is p (θ 3 ) Namely, isCombining the formulas (5) and (6), and obtaining theta by matlab plotting 3 Along with theta 1 Is a change map of the image vertex corresponding to θ 3 The value is the desired cutting angle.
The invention also provides an ultrashort pulse reconstruction method based on the BP neural network, which comprises the following steps:
(1) Acquiring an ultrashort pulse slow scanning spectrum line trace diagram to be detected;
(2) And inputting the scanning spectrum line trace graph into a pre-trained BP neural network model, and outputting the amplitude and the phase of the ultrashort pulse to be detected.
The ultra-short pulse reconstruction system and method based on the BP neural network provided by the invention have the beneficial effects that: the double reflection improvement is carried out on the autocorrelation optical paths on the basis of measuring ultrashort pulses by the traditional FROG method, so that the mutual interference between the optical paths is reduced. The cutting angle calculation of the BBO crystal is carried out according to different pulse wavelengths to be detected, and the applicability of the system and the success rate of experiments are improved. The BP neural network is combined to replace the traditional PCGPA algorithm to reconstruct the pulse information to be detected from the FROG trace, and the training sets presenting different signal-to-noise ratios are used to train the neural network, so that the ultra-short pulse reconstruction efficiency is greatly improved, and meanwhile, the accuracy of reconstructing the ultra-short pulse under the condition of low signal-to-noise ratio is improved.
Drawings
Fig. 1 is a schematic structural diagram of an ultrashort pulse reconstruction system based on a BP neural network according to an embodiment of the present invention;
FIG. 2 shows the result of BBO crystal collineation phase matching calculation at a pulse wavelength of 800nm to be measured;
FIG. 3 is a schematic diagram of a BP neural network structure;
FIG. 4 is a schematic diagram of a randomization process;
fig. 5 is a flowchart of an ultrashort pulse reconstruction method based on a BP neural network according to an embodiment of the present invention;
in fig. 1: 1. a femtosecond laser; 2-1, a first silver-plated reflecting mirror; 2-2 a second silvered mirror; 2-3 a third silvered mirror; 3. a femtosecond beam splitter; 4. a converging mirror; 5. a nonlinear crystal; 6. a diaphragm; 7. a spectrometer; 8. a stepping motor; 9. a computer; 10 an external electrical pulse generator.
Detailed Description
In order that the invention may be readily understood, a more particular description thereof will be rendered by reference to specific embodiments that are illustrated in the appended drawings. Preferred embodiments of the present invention are shown in the drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.
Embodiment 1 the ultra-short pulse reconstruction system based on the BP neural network provided in this embodiment mainly comprises a hardware system and a software system, wherein the hardware system is a double-beam autocorrelation optical path system using a BBO crystal as a nonlinear crystal, and the software system comprises a BP neural network model and a control module.
The structure of the dual-beam autocorrelation optical path system using BBO crystal as nonlinear crystal is shown in fig. 1, and includes: the system comprises a femtosecond laser 1, 5 silver plating reflectors, a femtosecond beam splitter 3, a converging reflector 4, a nonlinear crystal 5, a data acquisition system, a delay system and a delay system controller.
The data acquisition system consists of a spectrometer 7 and a computer 9. The delay system consists of two second silver-plated reflectors 2-2 and a stepping motor 8, and the delay system controller consists of an external electric pulse generator 10 and a control module arranged in a computer 9 for controlling the delay system to generate displacement, and a BP neural network model is arranged in the computer 9.
The fixed light path consists of two first silvered mirrors 2-1. The outgoing light of the fixed light path enters the converging mirror 4. The emergent light of the delay system enters the converging mirror 4 after passing through the third silver plating mirror.
The nonlinear crystal 5 is used for generating a frequency-multiplied pulse signal, and the diaphragm 6 is used for blocking other pulses except the frequency-multiplied pulse.
For ease of movement, the optical path portion of this example is placed on a 25 x 25 hole aluminum honeycomb panel, with stepper motor 8 connected to external pulse generator 10 via data lines, external pulse generator 10 connected to computer 9 via an RS232 data bus, spectrometer 7 connected to computer 9 via instrument control lines, and all instruments in the system controlled by computer 9.
The femtosecond laser 1 outputs ultrashort pulses to be detected, and the femtosecond beam splitter 3 is 1: and 1, splitting the ultrashort pulse to be detected into two beams of input delay systems and a fixed light path. The fixed optical path and the delay system separate the incident beam from the outgoing beam by two reflections.
The data acquisition system acquires a scanning spectrum by a spectrometer 7 and generates a scanning spectrum line trace by a computer 9. The delay system generates optical path delay through the stepping motor 8, the delay system controller consists of a control module in the computer 9 and an external electric pulse generator 10, the control module in the computer 9 uses labview to write a program to control the delay amount, and the external electric pulse generator 10 is controlled to output pulse electric signals to the stepping motor 8.
In this embodiment, the nonlinear crystal 5 is a barium metaborate crystal (BBO crystal) having a large phase matching angle. The cutting angle of the BBO crystal is calculated according to the wavelength to be measured, and the calculation method comprises the following steps:
let two pulses to be measured be k 1 And k 2 The frequency multiplication pulse is k 3 Phase matching under main section conditionsProvided that the wave vector is satisfied by the superposition of vectors, i.e
Decomposing the vector to obtain a formula:
because the three pulses satisfy the vector superposition
Square adding the formulas (2) and (3), substitutingThe following steps are obtained:
from the frequency multiplication, ω 3 ＝2ω 1 ＝2ω 2 Therefore, it is
Wherein n is p (θ 3 ) For the included angle theta between the frequency doubling light and the main axis 3 Refractive index of time-doubled light; n is n p (θ 1 )、n p (θ 2 ) For the included angle between the pulse to be measured and the main shaft to be theta 1 、θ 2 The refractive index of the pulse to be measured; in the embodiment, a type of phase matching ooe is adopted, the optical refractive index of the pulse o to be detected does not change along with the included angle of the main shaft, and the optical refractive index of the frequency doubling light e is calculated according to a formulaThe calculation results are that:
wherein n is p (θ 3 ) Namely, isCombining the formulas (5) and (6), and obtaining theta by matlab plotting 3 Along with theta 1 Is a change map of the image vertex corresponding to θ 3 The value is the desired cutting angle.
Compared with the prior art, the nonlinear crystal cutting angle calculating method provided by the embodiment is more visual, and can calculate the phase matching types, such as ooe, eoe and the like. In a real experiment, the angle of the pulse to be detected entering the nonlinear crystal is difficult to control, and the calculation is based on the fact that the pulse to be detected entering the nonlinear crystal at any angle can generate frequency multiplication pulse, so that the feasibility of the experiment is improved.
In this embodiment, the wavelength of the femtosecond laser is 800nm, the nonlinear crystal cutting angle selects the collinear phase matching frequency multiplication angle of the main section, the calculated angle is 29.3 degrees, and the calculated result is shown in fig. 2.
In this embodiment, the working principle of the hardware system is as follows:
the femtosecond pulse to be measured is divided into light intensity of 1 after passing through the femtosecond beam splitter: 1, wherein one beam of the same pulse is incident to a converging reflecting mirror after passing through a fixed light path, the other beam of the same pulse is reflected to the converging reflecting mirror after passing through a delay system and then is converged to the center of a BBO crystal by the converging reflecting mirror, and frequency doubling pulse is generated between the two pulses after passing through the nonlinear effect of the BBO crystal, and other pulses except the frequency doubling pulse are blocked by a small-hole diaphragm, so that only the frequency doubling pulse is ensured to be incident to a probe of the spectrometer. The spectrometer transmits the collected frequency multiplication pulse signals to a computer, and the frequency multiplication pulse signals are processed by the computer to obtain a scanning spectrum line trace graph.
In this embodiment, the software system selects a BP (back-ward propagation) neural network structure, and compared with a multi-layer perceptron, the BP neural network adopts a back propagation optimization parameter weight, so that the accuracy is higher and more reliable, and compared with a convolutional neural network structure, the BP neural network structure is simpler.
According to the image matrix with the scanning spectrum line trace to be detected being 128 x 3, the number of neurons of an input layer of a model is set to 49152, an input layer bias unit is added, the number of neurons of a hidden layer is set to 1, the number of neurons of the hidden layer is set to 49152, the bias unit of the hidden layer is added, the number of neurons of the output layer is set to 1, according to 128 points of both output pulse amplitude data and phase data, the number of neurons of the output layer is set to 256, a relu activation function is used for introducing nonlinear factors, and the neural network is completed, as shown in fig. 3.
In this embodiment, the training set of the BP neural network model adopts a constructed simulated scanning spectrum trace set, and the construction method and process of the training set are as follows:
(A) And (3) placing the femtosecond laser to be detected with the wavelength of 800nm in front of a built double-beam autocorrelation optical path system, and adjusting the collimation of an optical path to enable two pulses to be detected to be focused on a BBO crystal to generate second harmonic.
(B) Step-length displacement of a stepping motor is controlled by using a program written by labview, delay is realized, scanning spectrums of each displacement are collected by a spectrometer, and a scanning spectrum line trace diagram is generated at a computer end.
(C) Collecting pulse amplitude data from documents and experiments, and sampling 128 points of each group of collected pulse amplitude data;
(D) 1-30 data are selected from 128 sampling points of the amplitude data to carry out random digitizing processing with different degrees, specifically:
as shown in fig. 4a, the collected pulse amplitude data is a gaussian pulse, which is a continuous gaussian curve, 128 parts are divided on the abscissa, each part takes the value of the middle point to represent the part, a scatter diagram shown in fig. 4b is obtained, the scatter diagram is composed of 128 points, random number processing is carried out on the 128 points to different degrees, random number processing is carried out on 6 data points in fig. 4, and as shown in fig. 4c, the data of the 27 th, 45 th, 61 th, 70 th, 81 th and 101 th points are given with a value between 0 and 1 different from the original value, so that the scatter diagram shown in fig. 4c is obtained, and the process is the random number processing.
(E) And generating pulse phase data by the processed 6 pulse amplitude data and the function according to the formula:
in the middle ofThe scanning spectrum is the second harmonic, E (t) is the pulse to be detected, E (t-tau) is the time-shifted copy of the pulse to be detected, namely the gate function, omega is the pulse frequency to be detected, and tau is the delay.
And calculating to obtain simulated scanning spectrum line trace data, and generating 6 simulated scanning spectrum line trace diagrams with different signal to noise ratios. Fig. 4d is a fitted view of the scatter plot shown in fig. 4 c.
Through the randomization processing, the generated simulated scanning spectrum line trace graph presents different signal to noise ratios, and can simulate the real spectrum line trace graph.
In the training set of this embodiment, if 30 data are selected for randomization, the same group of noiseless pulse amplitudes and phases corresponds to 30 simulated scan spectrum linemaps with different signal-to-noise ratios.
In this embodiment, a spectrometer is used to collect a delayed scanning spectrum trace of a frequency multiplication pulse generated through an autocorrelation optical path, and a PCGPA algorithm is used to perform inversion calculation to obtain amplitude pulse width phase information of the pulse to be tested, which is used as a test data set.
Inputting the constructed training data set into a BP neural network model, training and optimizing the model, wherein a loss function adopts cross entropy:
when loss is reduced to 10 -3 And is considered to be expected.
And inputting the test data set and a part of training data set into a trained model, predicting pulse information to be tested, and comparing the pulse information with the original data to test the performance of the BP neural network.
Through prediction and comparison, under the condition that the scanning spectrum line trace shows a high signal-to-noise ratio (without randomization), the reconstruction accuracy of the BP neural network is about 95.1 percent (the traditional PCGPA is 95.3 percent), which is basically consistent with the traditional method, and under the condition that the scanning spectrum line trace shows a low signal-to-noise ratio (the number of randomization is 10), the reconstruction accuracy of the BP neural network is 61.7 percent (the traditional PCGPA is 40 percent), which is greatly improved compared with the traditional method. In terms of reconstruction speed, the time for reconstructing data from a set of scanning spectrum linegrams by the BP neural network is 0.5 seconds (the traditional PCGPA is about 30 minutes), and compared with the traditional method, the BP neural network has great improvement.
Embodiment 2 this embodiment provides an ultrashort pulse reconstruction method based on a BP neural network, and the flow is shown in fig. 5, and includes the following steps:
(1) Acquiring a scanning spectrum line trace diagram of a short wave pulse to be detected by using the built double-beam autocorrelation optical path system taking the BBO crystal as a nonlinear crystal;
(2) And inputting the acquired scanning spectrum line trace graph into a BP neural network model built in a computer, and outputting corresponding amplitude and phase after model prediction.
Claims (6)
1. An ultrashort pulse reconstruction system based on a BP neural network is characterized in that: the system comprises a double-beam autocorrelation optical path unit and a main control unit; the double-beam autocorrelation optical path unit is used for acquiring a scanning spectrum line trace of the ultrashort pulse; the main control unit is used for controlling the system and processing data and comprises a control module and a data processing module; the data processing module comprises a BP neural network model; the BP neural network model predicts an input scanning spectrum line trace graph and outputs the amplitude and the phase of ultrashort pulses to be detected; the training set of the BP neural network model consists of simulated scanning spectrum line trace diagrams, and the same group of noiseless pulse amplitude and phase correspond to n simulated scanning spectrum line trace diagrams with different signal to noise ratios; n is more than or equal to 1 and less than or equal to 30;
the simulated scanning spectrum line trace map is obtained by the following steps:
(1) Collecting pulse amplitude data, and sampling N points of each group of collected pulse amplitude data, wherein N is more than or equal to 100 and less than or equal to 500;
(2) N data are selected from N sampling points of the pulse amplitude data, and random numeralization processing is carried out to different degrees;
(3) And (3) generating pulse amplitude data processed in the step (2) and pulse phase data generated by a function according to a formula:
in the middle ofThe method is characterized in that the method is a scanning spectrum of a second harmonic, E (t) is a pulse to be detected, E (t-tau) is a time-shifted copy of the pulse to be detected, namely a gate function, omega is the pulse frequency to be detected, and tau is a delay amount;
and calculating to obtain simulated scanning spectrum line trace data, and generating n simulated scanning spectrum line trace diagrams with different signal to noise ratios.
2. The BP neural network-based ultrashort pulse reconstruction system according to claim 1, wherein: the random digitizing processes of different degrees are as follows: respectively taking different random numbers between 0 and 1 for the n selected pulse amplitude data; and the selected random number is different from the corresponding original value.
3. The BP neural network-based ultrashort pulse reconstruction system according to claim 1, wherein: the control module is used for controlling the delay system of the double-beam autocorrelation optical path unit to generate delay.
4. The BP neural network-based ultrashort pulse reconstruction system according to claim 3, wherein: the control module is also used for controlling the spectrometer to continuously scan the frequency multiplication pulse spectrum while the delay system generates delay so as to generate a delayed scanning spectrum linetrace diagram.
5. The BP neural network-based ultrashort pulse reconstruction system according to claim 1, wherein: the nonlinear crystal in the double-beam autocorrelation optical path unit is barium metaborate crystal; the cutting angle of the barium metaborate crystal selects a main section collinear phase matching frequency multiplication angle, and the calculation method is as follows:
let two pulses to be measured be k 1 And k 2 The frequency multiplication pulse is k 3 The phase matching condition under the main section condition is that wave vector superposition is satisfied, namely
Decomposing the vector to obtain a formula:
because the three pulses satisfy the vector superposition
Square adding the formulas (2) and (3), substitutingThe following steps are obtained:
from the frequency multiplication, ω 3 ＝2ω 1 ＝2ω 2 Therefore:
wherein n is p (θ 3 ) For the included angle theta between the frequency doubling light and the main axis 3 Refractive index of time-doubled light; n is n p (θ 1 )、n p (θ 2 ) For the included angle between the pulse to be measured and the main shaft to be theta 1 、θ 2 The refractive index of the pulse to be measured;
the refractive index of the frequency doubling light e is calculated according to the formula:
wherein n is p (θ 3 ) Namely, isCombining the formulas (5) and (6), and obtaining theta by matlab plotting 3 Along with theta 1 Is a change map of the image vertex corresponding to θ 3 The value is the desired cutting angle.
6. An ultrashort pulse reconstruction method based on a BP neural network, which is characterized by using the system as defined in claim 1, comprising the following steps:
(1) Acquiring a scanning spectrum line trace diagram of an ultrashort pulse to be detected;
(2) And inputting the scanning spectrum line trace graph into a pre-trained BP neural network model, and outputting the amplitude and the phase of the ultrashort pulse to be detected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111383297.0A CN114216575B (en) | 2021-11-22 | 2021-11-22 | Ultrashort pulse reconstruction system and method based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111383297.0A CN114216575B (en) | 2021-11-22 | 2021-11-22 | Ultrashort pulse reconstruction system and method based on BP neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114216575A CN114216575A (en) | 2022-03-22 |
| CN114216575B true CN114216575B (en) | 2024-04-05 |
Family
ID=80697707
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111383297.0A Active CN114216575B (en) | 2021-11-22 | 2021-11-22 | Ultrashort pulse reconstruction system and method based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114216575B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5530544A (en) * | 1992-10-26 | 1996-06-25 | Sandia Corporation | Method and apparatus for measuring the intensity and phase of one or more ultrashort light pulses and for measuring optical properties of materials |
| CN101526465A (en) * | 2009-04-22 | 2009-09-09 | 天津大学 | Quick multi-wavelength tissue optical parameter measuring device and trans-construction method |
| WO2021123481A1 (en) * | 2019-12-19 | 2021-06-24 | Universidad De Salamanca | Method and system for the temporal and spectral characterization of the amplitude and phase of ultrashort laser pulses |
| CN113063507A (en) * | 2021-03-26 | 2021-07-02 | 中国科学院物理研究所 | Ultra-short pulse width prediction method based on convolutional neural network |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6504612B2 (en) * | 2000-11-14 | 2003-01-07 | Georgia Tech Research Corporation | Electromagnetic wave analyzer |
| US9423307B2 (en) * | 2013-02-20 | 2016-08-23 | Mesa Photonics, LLC | Method and apparatus for determining wave characteristics using interaction with a known wave |
| US10908026B2 (en) * | 2016-08-10 | 2021-02-02 | Sphere Ultrafast Photonics, S.A. | System and method for calculating the spectral phase of laser pulses |
-
2021
- 2021-11-22 CN CN202111383297.0A patent/CN114216575B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5530544A (en) * | 1992-10-26 | 1996-06-25 | Sandia Corporation | Method and apparatus for measuring the intensity and phase of one or more ultrashort light pulses and for measuring optical properties of materials |
| CN101526465A (en) * | 2009-04-22 | 2009-09-09 | 天津大学 | Quick multi-wavelength tissue optical parameter measuring device and trans-construction method |
| WO2021123481A1 (en) * | 2019-12-19 | 2021-06-24 | Universidad De Salamanca | Method and system for the temporal and spectral characterization of the amplitude and phase of ultrashort laser pulses |
| CN113063507A (en) * | 2021-03-26 | 2021-07-02 | 中国科学院物理研究所 | Ultra-short pulse width prediction method based on convolutional neural network |
Non-Patent Citations (1)
| Title |
|---|
| Deep learning reconstruction of ultrashort pulses;TOM ZAHAVY.et.al;Optica;20180518;第5卷(第5期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114216575A (en) | 2022-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Stratoudaki et al. | Laser induced ultrasonic phased array using full matrix capture data acquisition and total focusing method | |
| US6504612B2 (en) | Electromagnetic wave analyzer | |
| JP4790560B2 (en) | Single terahertz wave time waveform measurement device | |
| Hoelen et al. | Image reconstruction for photoacoustic scanning of tissue structures | |
| US6219142B1 (en) | Method and apparatus for determining wave characteristics from wave phenomena | |
| WO2006020341A2 (en) | Characterization of materials with optically shaped acoustic waveforms | |
| Chen et al. | Terahertz-wave imaging system based on backward wave oscillator | |
| Bolzonello et al. | Versatile setup for high-quality rephasing, non-rephasing, and double quantum 2D electronic spectroscopy | |
| JP2010048721A (en) | Terahertz measuring device | |
| CN116519601A (en) | Photoacoustic microscopic imaging system and method based on Airy light beam combined sparse sampling | |
| CN114216575B (en) | Ultrashort pulse reconstruction system and method based on BP neural network | |
| CN114324177B (en) | Laser ultrasonic nondestructive testing device and method | |
| CN109186769B (en) | A method of the measurement ellipse inclined rate of chirped pulse | |
| CN108344711A (en) | A kind of method and system improving terahertz pulse imaging resolution | |
| Cao et al. | Correction algorithm of the frequency-modulated continuous-wave LIDAR ranging system | |
| JPWO2004113885A1 (en) | Optical waveform measuring device and measuring method thereof, complex refractive index measuring device and measuring method thereof, and computer program recording medium recording the program | |
| Trofimov et al. | Highly effective method for temporal terahertz spectroscopy under the condition of random probe signals | |
| US4178079A (en) | Sub-picosecond optical gating by degenerate four-wave mixing | |
| CN214893682U (en) | Quick ultrahigh-resolution transient absorption spectrum measuring device | |
| US11768113B2 (en) | Light pulse signal processing system comprising a cylindrical lens to provide a signal light pulse having a spatial angle chirp incident on a pair of long mirrors at different angles | |
| RU2359265C1 (en) | Ultrasonic introscopy device | |
| Chang et al. | Frequency-chirped readout of spatial-spectral absorption features | |
| JP2000088657A (en) | Method for measuring wavelength of extremely short light pulse | |
| Li et al. | Research on singular value decomposition denoising algorithm based on a genetic algorithm and FFT and its application | |
| Brinks | Time-Resolved X-Ray Diffraction from Coherent Acoustic Phonons Analyzed by Combining Microscopic Modeling with Deep Learning Based Strain Retrieval |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |