Optical fiber laser intelligent mode locking control method and system based on EPC

Abstract

The invention discloses an optical fiber laser intelligent mode locking control method and system based on an electric polarization controller EPC. The system drives a closed-loop control system through an evolutionary algorithm to achieve optical fiber laser intelligent mode locking. The method specifically comprises the following steps: generating a driving signal, acting on a control module to drive an EPC, and adjusting the polarization state in a laser cavity; acquiring acquisition data, and synchronously acquiring time-space intensity evolution of pulses and real-time detection of spectrums through an acquisition module; a calculation program is executed, a calculation module calculates a fitness function value based on the collected data, an evolutionary algorithm is executed, and a new generation of voltage combination is generated; and the new voltage combination acts on the EPC through the feedback module for iteration until the control signal meets the mode locking requirement. According to the invention, the technical problems of low repeatability and poor stability of manual mold locking are solved, the time cost is reduced, and the production efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology, specifically to an intelligent mode-locking control method and system for fiber lasers based on EPC. Background Technology

[0002] Ultrafast fiber lasers generate ultrashort pulse sequences that form femtosecond pulse chains in the time domain and exhibit equally spaced comb-like spectral lines in the frequency domain. This characteristic makes them a core light source in cutting-edge fields such as optical frequency combs, astronomical spectral calibration, and attosecond physics. Mode-locking is the core method for generating ultrashort pulses in laser technology. Essentially, it forces all longitudinal modes within the laser cavity to maintain a fixed phase relationship, causing them to coherently superimpose to form periodic ultrashort pulses. Mode-locking requires the introduction of a periodic modulation mechanism within the cavity, ensuring low loss in the high-intensity pulse portion and high loss in the low-intensity continuous light portion.

[0003] Passive mode-locking technology for ultrafast fiber lasers is the core means to achieve highly stable femtosecond pulse output, and its key lies in the precise control of the nonlinear polarization evolution effect within the laser cavity. Traditional mode-locking relies on manually adjusting the mechanical knob of the polarization controller (PC) to change the birefringence state within the cavity by squeezing the fiber. This manual operation mode has significant drawbacks: due to the complex nonlinear dynamics within the laser, achieving a specific mode-locking state usually requires adjusting multiple control parameters. Operators must rely on experience to blindly search through a vast multidimensional parameter space, which is not only time-consuming (up to several hours) but also results in extremely poor reproducibility of mode-locking results due to differences in technique; the mode-locking characteristics of different adjustments within the same system can fluctuate by more than 30%. Furthermore, the polarization state after mechanical adjustment is susceptible to changes in ambient temperature, leading to mode-locking degradation or even loss of lock. This not only hinders the study of the dynamic characteristics of fiber lasers but also restricts the reliability of applications in industrial settings.

[0004] To improve automation, researchers have attempted to use motor-driven PCs to replace manual operation. While this approach alleviates the burden on human resources, its lifespan is often less than 100,000 operations due to limitations in the response speed of mechanical transmission components (typically requiring more than 200ms for adjustment steps) and inherent wear issues, making it difficult to meet the continuous operation requirements of industrial equipment.

[0005] In recent years, EPC (Electronic Control Process) has become a research hotspot due to its millisecond-level electronic control response capability. However, most existing EPC control schemes employ preset voltage sequence scanning or simple feedback control, which are essentially still local parameter optimizations. These methods encounter two major dilemmas in complex multidimensional control spaces: first, they cannot effectively distinguish between mode-locked states with similar physical characteristics; second, they lack the ability to identify pseudo-mode-locked states such as noise impulses and relaxation oscillations, leading to frequent local optima in the optimization process. Summary of the Invention

[0006] The purpose of this invention is to provide an intelligent mode-locking control method and system for fiber lasers based on EPC, addressing the aforementioned problems.

[0007] The technical solution of the present invention is as follows: A method for intelligent mode-locking control of fiber lasers based on EPC includes the following steps: The control module applies randomly generated combinations of voltage signals to the EPC to regulate the state of the fiber laser; The acquisition module acquires data and simultaneously obtains the spatiotemporal intensity evolution of the pulse and the real-time detection of its spectrum. The calculation module calculates the fitness function value based on the collected data and generates new voltage combinations through an evolutionary algorithm; The new voltage combination is applied to the EPC iteratively until the fiber laser reaches the mode-locked state.

[0008] Furthermore, the fitness function is calculated based on the collected spatiotemporal intensity evolution information and spectral information, wherein the spatiotemporal intensity evolution information includes one or more of the following: pulse autocorrelation trace, radio frequency spectrum, or time-domain waveform characteristics; and the spectral information includes one or more of the following: spectral bandwidth, spectral shape, or spectral stability characteristics.

[0009] Furthermore, the fitness function is: , in, , , Used to eliminate relaxation oscillations and noise-like pulses. Used to distinguish between stable mode-locking and breather mode-locking, when When the value is close to zero, it indicates that the fiber laser is in a stable mode-locked state. When the value is far from zero, it indicates that the laser may be subject to breathing mode lock.

[0010] The above method addresses the shortcomings of traditional manual mode-locking, which relies on operators visually observing waveforms on spectrometers or oscilloscopes, leading to strong subjectivity and an inability to quantitatively assess mode-locking quality. By constructing a fitness function using multi-dimensional quantitative parameters, the mode-locking state is transformed into a calculable mathematical index, eliminating subjectivity. The fitness function serves as the optimization objective of the evolutionary algorithm, guiding it to automatically evolve towards a "better mode-locking state," achieving targeted and efficient search. Furthermore, the fitness function includes stability parameters, enabling the algorithm to actively suppress environmental disturbances, maintain long-term mode-locking, and enhance the robustness and stability of the system.

[0011] Furthermore, the evolutionary algorithm includes a genetic algorithm or a differential evolution algorithm.

[0012] Furthermore, the genetic algorithm specifically includes: Fifty superior individuals were selected through a roulette wheel selection process. Random values ​​were then generated by computer to determine whether gene crossover and mutation occurred within the group, thus generating a new generation of voltage signal combinations. Crossover occurs when the random value generated by the computer is less than the crossover probability; mutation occurs when the random value generated by the computer is less than the crossover probability. When two individuals cross over, one of the four voltage values ​​is selected for exchange. When mutation occurs, the voltage value 0-5V is converted into 14-bit binary, and one bit is randomly selected for transformation between 0 and 1. The new generation voltage signal is applied to the EPC through the FPGA and DAC. After continuous iteration, the maximum fitness function value and the average fitness function value converge to the optimal value.

[0013] Using the above method, EPC needs to adjust four voltage channels to change the polarization state inside the laser cavity. The evolutionary algorithm explores the voltage combination space in parallel without human intervention, completing an automated global search and saving mode-locking time. The evolutionary algorithm escapes the local optimum region through crossover and mutation mechanisms to achieve global optimal mode-locking. The evolutionary algorithm responds to environmental disturbances in real time and automatically reconverges to the mode-locking point when the external environment changes.

[0014] This application also includes an EPC-based intelligent mode-locking control system for fiber lasers. The system is a closed-loop control system and uses an EPC-based intelligent mode-locking control method for fiber lasers, including: Control module: Applies voltage signals to the EPC to regulate the state of the fiber laser; Acquisition module: Connects to the output of the fiber laser to acquire the laser's time and frequency information; Calculation module: Connects to the acquisition module, processes and analyzes the acquired data, and determines the next steps for adjusting the control module; Feedback module: Connects the calculation module and the EPC interface, and feeds back the information processed by the calculation module to the EPC.

[0015] Furthermore, the control module includes an electropolarization controller (EPC), which comprises four piezoelectric ceramics positioned at 45-degree angles to each other, each piezoelectric ceramic being controlled by a voltage signal V1 / V2 / V3 / V4.

[0016] With the above modules, compared with PC, EPC's response time is on the order of milliseconds, which can quickly and dynamically change the polarization state of the pulse inside the laser, reducing the time required for the algorithm to find the required laser state.

[0017] Furthermore, the acquisition module includes an analog-to-digital converter (ADC) that synchronously detects and acquires the spatiotemporal intensity evolution information and spectral information of the pulses output by the fiber laser after EPC regulation, converts the acquired analog signals into digital signals, and processes and compensates the signals through software algorithms.

[0018] The modules described above provide quantized input to the evolutionary algorithm, and this process does not introduce additional noise or signal distortion.

[0019] Furthermore, the computing module utilizes devices with high computing power and speed, including CPUs, MCUs, CPLDs, computing chips with real-time feedback, or computing devices composed of these. It characterizes the mode-locked state, calculates fitness function values ​​based on the collected spatiotemporal intensity evolution information and spectral information, executes an evolutionary algorithm, and generates a new generation of EPC driving voltage combinations based on the fitness function values.

[0020] Furthermore, the feedback module includes an FPGA and a digital-to-analog converter (DAC) to convert the next-generation voltage combination into an analog signal and feed it back to the EPC in real time for iteration until the laser reaches the mode-locked state.

[0021] The above modules improve voltage control accuracy and reduce feedback time.

[0022] Compared with existing technologies, the advantages of this invention are: 1. Traditional mode-locking requires manual adjustment of the PC to achieve mode-locking of fiber lasers. However, due to the complex nonlinear dynamics within the laser, even slight parameter differences can lead to different mode-locking characteristics, making it difficult to achieve the optimal mode-locking state and consuming a significant amount of time. This invention achieves intelligent mode-locking of fiber lasers by using voltage-driven EPC, a data acquisition module to collect data, a calculation module to perform fitness function calculations and evolutionary algorithms, and a feedback module to apply new voltage combinations to the EPC and iterate. This eliminates the tedious process of relying on operator experience and repeated manual adjustments inherent in traditional methods. 2. When the external environment changes, such as due to vibration, temperature drift, or other factors, the laser may transition from a mode-locked state to a unlocked state. If the laser is encapsulated as part of a system or instrument, re-mode-locking becomes more cumbersome and time-consuming. This invention utilizes a closed-loop control system that continuously monitors and fine-tunes the voltage as needed once the target state is reached. This effectively resists environmental disturbances and maintains long-term stable operation in the mode-locked state. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the system structure of this application.

[0024] Figure 2 This is a flowchart illustrating the method described in this application.

[0025] Figure 3 This is a schematic diagram of the internal structure of the EPC.

[0026] Figure 4 This is a schematic diagram of genetic operations.

[0027] Figure 5 This is the time-domain plot when mode-locked.

[0028] Figure 6 This diagram illustrates the principle of evolutionary algorithms.

[0029] Figure 7 This is a schematic diagram of the evolution of the fitness function.

[0030] Figure 8 This is the time-domain diagram when the mode-locked state is reached. Detailed Implementation

[0031] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0032] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0033] Please see Figure 1-8 A smart mode-locking control method for fiber lasers based on EPC, such as Figure 2 As shown, it includes the following steps: N sets of drive voltage signal combinations are randomly generated and applied to the EPC. Each set contains four independent voltage values ​​(V1, V2, V3, V4). The generated voltage signals are combined and input into the fiber laser system to drive the EPC to change the polarization state inside the laser cavity; The ADC acquires the spatiotemporal intensity evolution and spectral information of the fiber laser, including time-domain data, spectral bandwidth, and spectral shape. Based on the collected spectral and temporal data, the fitness function value corresponding to each group of voltage signals is calculated. This function quantifies and evaluates the mode-locking state and quality. Based on the fitness function value, select X groups of the current optimal voltage signal combinations, and perform genetic operations on the selection results based on the evolutionary algorithm to generate a new generation of voltage signal combinations; The new generation of voltage signals is applied to the EPC via FPGA and DAC; The above method is executed iteratively until the fitness function value reaches a preset threshold or the iteration termination condition is met, thus achieving a stable mode-locked state.

[0034] Genetic operations include: wheel selection, which ensures that individuals with higher fitness have a higher probability of being selected; simple crossover, which can accelerate the convergence of the fitness function value, reduce the number of iterations, and save modulus locking time; and Gaussian mutation, which prevents the algorithm from getting trapped in local optima and failing to find the global optimum.

[0035] The iteration terminates under one of the following conditions: (1) The difference between the maximum and average fitness function values ​​over K consecutive generations < ; (2) Reach the preset maximum number of iterations .

[0036] This application also includes an intelligent mode-locking control system for fiber lasers based on EPC, such as... Figure 1 As shown, it includes a control module, a data acquisition module, a calculation module, and a feedback module, and can execute according to the intelligent mode-locking control method described above.

[0037] Control Module: The EPC consists of four piezoelectric ceramics positioned at 45° angles to each other. Each piezoelectric ceramic is controlled by a voltage signal (V1, V2, V3, V4), such as... Figure 3 As shown, its response time is 0.4ms, which can quickly change the polarization state of the pulses inside the fiber laser, which is beneficial for the algorithm to find the required laser state in a short time. The acquisition module includes an ADC, which is connected to the laser output and calculation module to synchronously detect and acquire the spatiotemporal intensity evolution information and spectral information of the pulses output by the fiber laser after EPC adjustment. The calculation module is a computer that deploys the evolutionary algorithm, which is connected to the acquisition module and the feedback module to calculate the fitness function value, execute the evolutionary algorithm, and generate a new generation of voltage control commands. The feedback module includes an FPGA and a DAC. The FPGA controls the DAC chip through the SPI interface to convert the new generation of voltage combinations into analog signals and feed them back to the EPC in real time.

[0038] The calculation module executes the following process: N sets of random voltage combinations are generated during initialization; X sets of superior individuals are selected using a roulette wheel, with individuals with higher fitness having a greater probability of being selected from the population. (Refer to...) Figure 4As shown, the selected superior individuals are grouped in pairs, and random values ​​are generated by computer to determine whether gene crossover and mutation occur within the group; finally, the optimal retention strategy is used to update the population.

[0039] The feedback module includes an FPGA and a DAC. The FPGA receives digital voltage commands generated by the evolutionary algorithm and sends them to the DAC. The DAC converts the digital commands into high-precision analog voltages to drive the EPC. The next-generation voltage combination is converted into an analog signal and fed back to the EPC in real time for iteration until the laser reaches the mode-locked state from the unmode-locked state.

[0040] In another specific embodiment, examples of intelligent stabilizing mode-locking and breathing sub-mode-locking are given: 100 sets of randomly generated voltage signal combinations drive the EPC; at this time, the fiber laser has not achieved stable mode-locking, and the output time-domain waveform exhibits relaxation oscillation, with irregular peaks and no fixed repetition period. Figure 5 As shown; subsequently, the ADC acquires the spatiotemporal intensity evolution and spectral information of the fiber laser; based on the acquired spectral and temporal data, the computer calculates the fitness function value corresponding to each group of voltage signals, and the fitness function is designed as F:

[0041] in, , , Used to eliminate relaxation oscillations and noise-like pulses. Used to distinguish between stable mode-locking and breather mode-locking, when When the value is close to zero, it indicates that the fiber laser is in a stable mode-locked state. When the value is far from zero, it indicates that the laser may be subject to breathing mode-locking. An evolutionary algorithm is executed, using a roulette wheel to select 50 superior individuals. A computer-generated random value determines whether gene crossover and mutation occur within each group, generating a new generation of voltage signal combinations. Crossover occurs when the computer-generated random value is less than the crossover probability; mutation occurs when the computer-generated random value is less than the crossover probability. When two individuals crossover, one of four voltage values ​​is selected for exchange. During mutation, the voltage value (0-5V) is converted into 14-bit binary, and one bit is randomly selected and toggled between 0 and 1. (Refer to...) Figure 6 As shown; the new generation voltage signal is applied to the EPC through the FPGA and DAC. After 11 generations, the maximum fitness function value and the average fitness function value converge to the optimal value. (Refer to...) Figure 7 As shown; a stable mode-locked state is finally achieved, referencing... Figure 8 As shown, the output time-domain signal at this time is a periodic pulse sequence with a fixed repetition period.

[0042] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A method for intelligent mode-locking control of fiber lasers based on EPC, characterized in that, Includes the following steps: The control module applies randomly generated combinations of voltage signals to the EPC to regulate the state of the fiber laser; The acquisition module acquires data and simultaneously obtains the spatiotemporal intensity evolution of the pulse and the real-time detection of its spectrum. The calculation module calculates the fitness function value based on the collected data and generates new voltage combinations through an evolutionary algorithm; The new voltage combination is applied to the EPC iteratively until the fiber laser reaches the mode-locked state.

2. The intelligent mode-locking control method for fiber lasers based on EPC according to claim 1, characterized in that, The fitness function is calculated based on the collected spatiotemporal intensity evolution information and spectral information, wherein the spatiotemporal intensity evolution information includes one or more of the following: pulse autocorrelation trace, radio frequency spectrum, or time-domain waveform characteristics; and the spectral information includes one or more of the following: spectral bandwidth, spectral shape, or spectral stability characteristics.

3. The intelligent mode-locking control method for fiber lasers based on EPC according to claim 1 or 2, characterized in that, The fitness function is: ， in, , , Used to eliminate relaxation oscillations and noise-like pulses. Used to distinguish between stable mode-locking and breather mode-locking, when When the value is close to zero, it indicates that the fiber laser is in a stable mode-locked state. When the value is far from zero, it indicates that the laser may be subject to breathing mode lock.

4. The intelligent mode-locking control method for fiber lasers based on EPC according to claim 1, characterized in that, The evolutionary algorithm includes genetic algorithm or differential evolution algorithm.

5. The intelligent mode-locking control method for fiber lasers based on EPC according to claim 4, characterized in that, The genetic algorithm specifically includes: Fifty superior individuals were selected through a roulette wheel selection process. Random values ​​were then generated by computer to determine whether gene crossover and mutation occurred within the group, thus generating a new generation of voltage signal combinations. Crossover occurs when the random value generated by the computer is less than the crossover probability; mutation occurs when the random value generated by the computer is less than the crossover probability. When two individuals cross over, one of the four voltage values ​​is selected for exchange. When mutation occurs, the voltage value 0-5V is converted into 14-bit binary, and one bit is randomly selected for transformation between 0 and 1. The new generation voltage signal is applied to the EPC through the FPGA and DAC. After continuous iteration, the maximum fitness function value and the average fitness function value converge to the optimal value.

6. A smart mode-locking control system for fiber lasers based on EPC, characterized in that, The system is a closed-loop control system, using an EPC-based intelligent mode-locking control method for fiber lasers as described in any one of claims 1-5, comprising: Control module: Applies voltage signals to the EPC to regulate the state of the fiber laser; Acquisition module: Connects to the output of the fiber laser to acquire the laser's time and frequency information; Calculation module: Connects to the acquisition module, processes and analyzes the acquired data, and determines the next steps for adjusting the control module; Feedback module: Connects the calculation module and the EPC interface, and feeds back the information processed by the calculation module to the EPC.

7. The intelligent mode-locking control system for fiber lasers based on EPC according to claim 6, characterized in that, The control module includes an electropolarization controller (EPC), which comprises four components spaced 45 degrees apart. 。 The piezoelectric ceramics at the corners are each controlled by a voltage signal V1 / V2 / V3 / V4.

8. The intelligent mode-locking control system for fiber lasers based on EPC according to claim 6, characterized in that, The acquisition module includes an analog-to-digital converter (ADC), which synchronously detects and acquires the spatiotemporal intensity evolution and spectral information of the pulses output by the fiber laser after EPC regulation, converts the acquired analog signals into digital signals, and processes and compensates the signals through software algorithms.

9. The intelligent mode-locking control system for fiber lasers based on EPC according to claim 6, characterized in that, The computing module uses a device with strong computing power and fast computing speed, including a CPU, MCU, CPLD, computing chips with real-time feedback, or computing devices composed of these.

10. The intelligent mode-locking control system for fiber lasers based on EPC according to claim 6, characterized in that, The feedback module includes an FPGA and a digital-to-analog converter (DAC), which converts the next-generation voltage combination into an analog signal and feeds it back to the EPC in real time for iteration until the laser reaches the mode-locked state.

Images