A method for improving fiber coupling efficiency based on bat algorithm

Abstract

The application discloses a kind of based on bat algorithm's optical fiber coupling efficiency promotion method, including the following steps: collecting optical power data and preprocessing;Coupling parameter search space model is constructed;Improved latent space reinforcement-bat algorithm is initialized, and initial coupling parameter vector is output;Introduce reversible flow coding to generate latent space vector;The optimal action vector is calculated by differentiable strategy flow mapping mechanism and output;Current iteration optimal bat individual is obtained;The coupling parameter vector of global optimal bat individual is output;Improved latent space reinforcement-bat algorithm is recalled again to carry out parameter optimization calculation, effectively improves the optimization speed and global optimal stability of optical fiber coupling, significantly reduces coupling loss, improves system automation adjustment precision and optical power stability.

Description

Technical Field

[0001] This invention relates to the field of optical communication control technology, and in particular to a method for improving optical fiber coupling efficiency based on the bat algorithm. Background Technology

[0002] Fiber optic coupling systems are crucial components in optical communication, optical sensing, and optoelectronic testing equipment, and their coupling efficiency directly impacts the system's optical power transmission performance and signal transmission stability. Achieving high-precision, robust, automated fiber optic coupling has been a significant research direction in optical manufacturing and control, particularly in applications such as fiber optic devices, fiber optic arrays, and photonic chip packaging. Traditional fiber optic coupling methods often combine mechanical adjustments with empirical parameter tuning, relying on manual optimization through displacement platforms and optical power detection devices. While simple to implement, these methods are time-consuming, have poor repeatability, and are extremely sensitive to environmental disturbances, making it difficult to meet high-precision and real-time requirements.

[0003] With the development of intelligent control and optimization algorithms, researchers have gradually introduced intelligent optimization algorithms into the fiber coupling process to achieve global optimization search for multiple parameters. Currently, commonly used intelligent optimization algorithms include particle swarm optimization, genetic algorithms, and ant colony optimization. These algorithms can improve the automation level and search efficiency of fiber coupling to some extent. However, existing algorithms generally have the following shortcomings: First, the traditional bat algorithm is prone to getting trapped in local optima in high-dimensional coupling parameter spaces, especially in multi-parameter nonlinear coupling systems, where the algorithm has slow convergence speed and insufficient global optimization capability; second, optical power signals in fiber coupling systems are often accompanied by noise fluctuations and dynamic drift, resulting in large fluctuations in the fitness function, making traditional optimization algorithms prone to oscillations and unstable searches during iterative updates; third, existing algorithms do not fully consider the linear constraints and physical feasibility between coupling parameters, and the generated parameter solutions may exceed the actual operating range of the system, affecting the practicality of the algorithm. Summary of the Invention

[0004] One objective of this invention is to propose a method for improving fiber coupling efficiency based on the bat algorithm. This invention fully utilizes optical power signal preprocessing technology, coupling parameter search space modeling technology, latent space reversible flow coding mechanism and reinforcement strategy network structure to introduce an improved latent space reinforcement-bat algorithm in the fiber coupling process to dynamically optimize fiber coupling parameters and achieve intelligent coupling control with high efficiency, low error and strong robustness.

[0005] A method for improving fiber coupling efficiency based on the bat algorithm according to an embodiment of the present invention includes the following steps:

[0006] Acquire and preprocess optical power data from the fiber optic coupling system;

[0007] Based on the preprocessed optical power data, the coupling parameter vector and value range of the optical fiber coupling system are set, and a coupling parameter search space model is constructed.

[0008] Based on the coupling parameter search space model, an improved latent space reinforcement-bat algorithm is initialized and the initial coupling parameter vector is output.

[0009] The initial coupling parameter vector is input into the latent space coding module of the improved latent space enhancement-bat algorithm, and reversible flow coding is introduced to generate latent space vectors.

[0010] The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module, and the optimal action vector is calculated and output through a differentiable policy flow mapping mechanism.

[0011] The optimal action vector is input into the bat individual evolution update module of the improved latent space reinforcement-bat algorithm to obtain the optimal bat individual in the current iteration.

[0012] The current best bat individual is input into the adaptive adjustment and termination control module of the improved latent space enhancement-bat algorithm. The loudness and pulse emission rate of the current best bat individual are adaptively corrected, and the latent space enhancement strategy parameters are adjusted according to the feedback signal of optical power data. The coupling parameter vector of the global best bat individual is output.

[0013] Based on the coupling parameter vector of the globally optimal bat individual, the control actuator adjusts the fiber optic coupling system to the corresponding coupling parameter state and locks it. During the operation of the fiber optic coupling system, the optical power signal is continuously monitored. When the optical power signal is detected to be less than the preset threshold, the improved latent space enhancement-bat algorithm is called again to perform parameter optimization calculation.

[0014] Optionally, the preprocessing includes noise filtering, smoothing, and normalization preprocessing.

[0015] Optionally, the construction of the coupling parameter search space model specifically includes:

[0016] Based on the preprocessed optical power signal, optical power response features are extracted, including the rate of change of optical power amplitude, the trend of coupling loss, and the steady-state power distribution of the system.

[0017] Define a coupling parameter vector, and set lower and upper limits for each parameter component based on the variation range of optical power response characteristics to form a lower limit vector and an upper limit vector. The coupling parameter vector includes fiber end position parameters, fiber end angle parameters, and optical component spacing parameters.

[0018] The feasible region of the coupling parameters is established based on the lower limit vector and the upper limit vector. The linear dependency between the components of each coupling parameter and the feasible state of the system are defined by the linear constraint matrix and the linear constraint vector, and the feasible region boundary model is obtained.

[0019] Under the constraints of the feasible domain boundary model, the parameter resolution step size is set according to the dynamic variation range of the preprocessed optical power data and the sensitivity of the coupling parameters. The step size quantization and normalization of each component of the coupling parameter vector are performed using the boundary information of the lower limit vector and the upper limit vector to obtain the quantized coupling parameter vector and the normalized coupling parameter vector.

[0020] The lower bound vector, upper bound vector, linear constraint matrix, linear constraint vector, parameter resolution step size, quantized coupling parameter vector, and normalized coupling parameter vector are combined to form a coupling parameter search space model.

[0021] Optionally, the output of the initial coupling parameter vector specifically includes:

[0022] The improved latent space enhancement-bat algorithm is loaded and initialized based on the coupled parameter search space model. The initialization parameters include population size, maximum number of iterations, search frequency range, initial loudness and initial pulse emission rate.

[0023] Projecting the initialization parameters onto the linear constraint matrix and linear constraint vector of the coupling parameter search space model generates candidate coupling parameter vectors. The generation of the candidate coupling parameter vectors is based on the coupling parameter samples generated by the initialization parameters. The coupling parameter samples are substituted into the constraint conditions defined by the linear constraint matrix and linear constraint vectors. The coupling parameter samples that do not meet the constraint conditions are corrected, and the coupling parameter samples that meet all constraint conditions are obtained as candidate coupling parameter vectors.

[0024] For each candidate coupling parameter vector, an initial coupling parameter vector is generated for each bat individual within the range of the lower bound vector and the upper bound vector.

[0025] Optionally, the generation of the latent space vector specifically includes:

[0026] The initial coupling parameter vector is normalized by mapping each parameter value of the initial coupling parameter vector to a normalized interval of zero to one proportionally, thereby generating a normalized coupling parameter vector.

[0027] A coupling parameter correlation matrix is ​​constructed based on the normalized coupling parameter vector, and the coupling parameter sensitivity coefficient is calculated. The construction of the coupling parameter correlation matrix involves performing covariance analysis on the normalized values ​​of each coupling parameter under multiple sets of optical power data, calculating the linear correlation coefficient and nonlinear coupling degree between each parameter component, and combining them by rows and columns to form the coupling parameter correlation matrix. The calculation process of the coupling parameter sensitivity coefficient involves applying small perturbations to each component of the normalized coupling parameter vector in sequence, recording the changes in optical power output before and after the perturbation, and obtaining the local response degree of the component to the optical power output by comparing the ratio of the change in optical power output to the corresponding parameter perturbation amount. The ratio is defined as the coupling parameter sensitivity value of the component.

[0028] The normalized coupling parameter vector and the coupling parameter sensitivity coefficient are input into the latent space coding module of the improved latent space enhancement-Bat algorithm. The normalized coupling parameter vector is then weighted by a feature weighting layer to obtain a weighted normalized coupling parameter vector.

[0029] The weighted normalized coupled parameter vector is used for nonlinear feature mapping using reversible flow coding to output a latent space vector, which refers to the state of an individual bat in the latent space feature domain.

[0030] Optionally, the output of the optimal action vector specifically includes:

[0031] The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module. Linear mapping and nonlinear activation calculation are performed through a differentiable policy flow mapping mechanism to obtain latent space features. The latent space reinforcement policy network module includes an input layer, a hidden layer and an action generation layer.

[0032] The latent space features are input into the hidden layer of the latent space reinforcement strategy network module, and feature mapping is performed through the weight matrix and a non-linear activation function is applied to output the original action vector.

[0033] The original motion vector is subjected to amplitude constraint processing to obtain a standardized motion output;

[0034] The standardized motion output is categorized into frequency adjustment, velocity correction, position disturbance, loudness adjustment, and pulse emission rate adjustment according to parameter type. Based on preset scaling factors and mapping rules, the amplitude of each component is corrected and constrained to obtain the optimal motion vector.

[0035] Optionally, obtaining the optimal bat individual in the current iteration specifically includes:

[0036] Based on the frequency adjustment amount in the optimal action vector, the search frequency of the individual bat is updated to obtain the updated frequency adjustment amount, which is the sum of the search frequency and the frequency adjustment amount of the previous iteration.

[0037] Based on the velocity correction in the optimal action vector and the coupling parameter vector of the optimal bat individual, the velocity vector of the bat individual is updated to obtain the updated velocity vector. The updated velocity vector is the sum of the original velocity vector and the velocity correction, and is formed by weighting and correcting the difference in coupling parameters between the bat individual and the global optimal individual in combination with the search frequency.

[0038] Based on the updated velocity vector and the position perturbation in the optimal motion vector, the coupling parameter vector of the individual bat is updated to obtain the updated position vector, which is obtained by combining the original position vector, the updated velocity vector, and the position perturbation.

[0039] The updated frequency adjustment, the updated velocity vector, and the updated position vector are concatenated to form the updated bat individual coupling parameter vector.

[0040] The updated bat individual coupling parameter vector is input into the optical fiber coupling system. The corresponding optical power output signal is collected by the control actuator, and the fitness value of the bat individual is calculated. The bat individual with the highest fitness value is selected as the optimal bat individual for the current iteration. The calculation process of the fitness value of the bat individual is based on the intensity of the optical power output signal. The collected optical power signal is compared with the target maximum optical power, and the minimum coupling loss is used as the evaluation criterion to calculate the fitness value of each bat individual.

[0041] Optionally, the output of the coupling parameter vector of the globally optimal bat individual specifically includes:

[0042] Adaptively correct the loudness adjustment and pulse emission rate adjustment of the current best bat individual to obtain the correction result, and synchronously update the parameter state of the improved latent space enhancement-bat algorithm with the correction result.

[0043] By updating the parameter state of the improved latent space reinforcement-bat algorithm, the policy parameters of the latent space reinforcement policy network module are adaptively updated to form the updated policy parameters. The update process adjusts the policy parameters along the direction of increasing reward value.

[0044] The termination condition is determined based on the updated strategy parameters. The maximum number of iterations is set to a preset upper limit. When the number of iterations reaches the preset upper limit, the coupling parameter vector of the globally optimal bat individual is output.

[0045] Optionally, the re-invocation of the improved latent space reinforcement-bat algorithm specifically includes:

[0046] Based on the coupling parameter vector of the globally optimal individual bat, the coupling parameter state of the fiber optic coupling system is adjusted to the state corresponding to the globally optimal coupling parameter vector by controlling the actuator, and a coupling parameter state locking operation is performed to form the globally optimal coupling state.

[0047] Based on the global optimal coupling state, online monitoring parameters for optical power signals are set, and optical power signals are collected according to the set sampling period. The moving average value of optical power is calculated within each monitoring window. The online monitoring parameters include sampling period, monitoring window length, power threshold, and power drift threshold.

[0048] When the sliding average optical power is less than the preset power threshold, a re-optimization trigger signal is generated. The generation of the re-optimization trigger signal is based on the real-time optical power monitoring module. When the sliding average optical power is lower than the preset power threshold within a continuous sampling period, the monitoring module determines that the system coupling state deviates from the optimal range and automatically outputs a digital trigger signal to the control unit.

[0049] When the re-optimization trigger signal is detected to be active, the current coupling parameter status of the fiber optic coupling system is read and input to the control unit along with the current optical power signal.

[0050] An optimized startup command is generated in the control unit, the optimization process of the improved latent space enhancement-bat algorithm is invoked again, and parameter re-optimization calculation is performed.

[0051] The beneficial effects of this invention are:

[0052] This invention achieves significant technical results by introducing an improved latent space reinforcement-bat algorithm into the fiber optic coupling process, realizing intelligent global optimization and dynamic adaptive adjustment of coupling parameters. The method uses optical power signals as optimization feedback. First, it establishes a search space model of the coupling parameters of the fiber optic coupling system. Then, it utilizes a reversible flow coding mechanism to map high-dimensional parameters into the latent space for compressed expression, effectively reducing coupling interference and search complexity. The algorithm executes reinforcement learning-style policy flow mapping within the latent space. Through the optimal action output of the policy network, it achieves efficient evolutionary updates of bat individuals in parameters such as frequency, speed, and position, making the search process more global and directional.

[0053] By dynamically feeding back optical power signals and adaptively correcting strategy parameters, this invention can adjust action outputs in real time during iteration and automatically correct coupling errors, thereby achieving rapid and stable convergence in multi-parameter nonlinear systems involving fiber end position, angle, and optical component spacing. This method ensures the physical feasibility of parameter values ​​during initialization, implements reinforcement learning-driven optimal actions during evolution, and achieves systematic control of globally optimal parameters during output, effectively avoiding the problems of traditional bat algorithms such as getting trapped in local optima, search oscillations, and insufficient convergence accuracy.

[0054] Compared to existing fiber coupling optimization methods based on genetic algorithms, particle swarm optimization, and traditional bat algorithms, this invention introduces latent space feature representation and differentiable strategy mapping mechanisms into the algorithm structure, enabling the search process to have self-learning and self-evolution capabilities, significantly improving the response speed and optimization accuracy of optical power signals. By maintaining the consistency of the mapping between latent and original spaces through reversible flow coding, the continuity and reversibility of parameter updates are strengthened, ensuring the stability and physical realizability of the optimization results.

[0055] In summary, this invention achieves intelligent optimization of the entire process of fiber coupling, from parameter modeling, feature mapping, strategy output to adaptive adjustment, significantly improving fiber coupling efficiency and system robustness. It has advantages such as fast convergence speed, strong global optimization capability, excellent noise resistance and high adaptability, providing an efficient, stable and learning-capable solution for automatic coupling control of complex fiber systems. Attached Figure Description

[0056] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0057] Figure 1 This is an overall flowchart of a method for improving fiber coupling efficiency based on the bat algorithm proposed in this invention.

[0058] Figure 2 This is a schematic diagram of the module structure of the improved latent space reinforcement-bat algorithm, which is a fiber coupling efficiency improvement method based on the bat algorithm proposed in this invention. Detailed Implementation

[0059] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0060] refer to Figure 1-2 A method for improving fiber coupling efficiency based on the bat algorithm includes the following steps:

[0061] Acquire and preprocess optical power data from the fiber optic coupling system;

[0062] Based on the preprocessed optical power data, the coupling parameter vector and value range of the optical fiber coupling system are set, and a coupling parameter search space model is constructed.

[0063] Based on the coupling parameter search space model, an improved latent space reinforcement-bat algorithm is initialized and the initial coupling parameter vector is output.

[0064] The initial coupling parameter vector is input into the latent space coding module of the improved latent space enhancement-bat algorithm, and reversible flow coding is introduced to generate latent space vectors.

[0065] The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module, and the optimal action vector is calculated and output through a differentiable policy flow mapping mechanism.

[0066] The optimal action vector is input into the bat individual evolution update module of the improved latent space reinforcement-bat algorithm to obtain the optimal bat individual in the current iteration.

[0067] The current best bat individual is input into the adaptive adjustment and termination control module of the improved latent space enhancement-bat algorithm. The loudness and pulse emission rate of the current best bat individual are adaptively corrected, and the latent space enhancement strategy parameters are adjusted according to the feedback signal of optical power data. The coupling parameter vector of the global best bat individual is output.

[0068] Based on the coupling parameter vector of the globally optimal bat individual, the control actuator adjusts the fiber optic coupling system to the corresponding coupling parameter state and locks it. During the operation of the fiber optic coupling system, the optical power signal is continuously monitored. When the optical power signal is detected to be less than the preset threshold, the improved latent space enhancement-bat algorithm is called again to perform parameter optimization calculation.

[0069] In this embodiment, the preprocessing includes noise filtering, smoothing, and normalization preprocessing.

[0070] In this embodiment, the construction of the coupling parameter search space model specifically includes:

[0071] Based on the preprocessed optical power signal, optical power response features are extracted, including the rate of change of optical power amplitude, the trend of coupling loss, and the steady-state power distribution of the system.

[0072] Define a coupling parameter vector, and set lower and upper limits for each parameter component based on the variation range of optical power response characteristics to form a lower limit vector and an upper limit vector. The coupling parameter vector includes fiber end position parameters, fiber end angle parameters, and optical component spacing parameters.

[0073] The feasible region of the coupling parameters is established based on the lower limit vector and the upper limit vector. The linear dependency between the components of each coupling parameter and the feasible state of the system are defined by the linear constraint matrix and the linear constraint vector, and the feasible region boundary model is obtained.

[0074] Under the constraints of the feasible domain boundary model, the parameter resolution step size is set according to the dynamic variation range of the preprocessed optical power data and the sensitivity of the coupling parameters. The step size quantization and normalization of each component of the coupling parameter vector are performed using the boundary information of the lower limit vector and the upper limit vector to obtain the quantized coupling parameter vector and the normalized coupling parameter vector.

[0075] The lower bound vector, upper bound vector, linear constraint matrix, linear constraint vector, parameter resolution step size, quantized coupling parameter vector, and normalized coupling parameter vector are combined to form a coupling parameter search space model.

[0076] In this embodiment, the output of the initial coupling parameter vector specifically includes:

[0077] The improved latent space enhancement-bat algorithm is loaded and initialized based on the coupled parameter search space model. The initialization parameters include population size, maximum number of iterations, search frequency range, initial loudness and initial pulse emission rate.

[0078] Projecting the initialization parameters onto the linear constraint matrix and linear constraint vector of the coupling parameter search space model generates candidate coupling parameter vectors. The generation of the candidate coupling parameter vectors is based on the coupling parameter samples generated by the initialization parameters. The coupling parameter samples are substituted into the constraint conditions defined by the linear constraint matrix and linear constraint vectors. The coupling parameter samples that do not meet the constraint conditions are corrected, and the coupling parameter samples that meet all constraint conditions are obtained as candidate coupling parameter vectors.

[0079] For each candidate coupling parameter vector, an initial coupling parameter vector is generated for each bat individual within the range of the lower bound vector and the upper bound vector.

[0080] In this embodiment, the generation of the latent space vector specifically includes:

[0081] The initial coupling parameter vector is normalized by mapping each parameter value of the initial coupling parameter vector to a normalized interval of zero to one proportionally, thereby generating a normalized coupling parameter vector.

[0082] A coupling parameter correlation matrix is ​​constructed based on the normalized coupling parameter vector, and the coupling parameter sensitivity coefficient is calculated. The construction of the coupling parameter correlation matrix involves performing covariance analysis on the normalized values ​​of each coupling parameter under multiple sets of optical power data, calculating the linear correlation coefficient and nonlinear coupling degree between each parameter component, and combining them by rows and columns to form the coupling parameter correlation matrix. The calculation process of the coupling parameter sensitivity coefficient involves applying small perturbations to each component of the normalized coupling parameter vector in sequence, recording the changes in optical power output before and after the perturbation, and obtaining the local response degree of the component to the optical power output by comparing the ratio of the change in optical power output to the corresponding parameter perturbation amount. The ratio is defined as the coupling parameter sensitivity value of the component.

[0083] The normalized coupling parameter vector and the coupling parameter sensitivity coefficient are input into the latent space coding module of the improved latent space enhancement-Bat algorithm. The normalized coupling parameter vector is then weighted by a feature weighting layer to obtain a weighted normalized coupling parameter vector.

[0084] The weighted normalized coupled parameter vector is used for nonlinear feature mapping using reversible flow coding to output a latent space vector, which refers to the state of an individual bat in the latent space feature domain.

[0085] In this embodiment, the output of the optimal action vector specifically includes:

[0086] The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module. Linear mapping and nonlinear activation calculation are performed through a differentiable policy flow mapping mechanism to obtain latent space features. The latent space reinforcement policy network module includes an input layer, a hidden layer and an action generation layer.

[0087] The latent space features are input into the hidden layer of the latent space reinforcement strategy network module, and feature mapping is performed through the weight matrix and a non-linear activation function is applied to output the original action vector.

[0088] The original motion vector is subjected to amplitude constraint processing to obtain a standardized motion output;

[0089] The standardized motion output is categorized into frequency adjustment, velocity correction, position disturbance, loudness adjustment, and pulse emission rate adjustment according to parameter type. Based on preset scaling factors and mapping rules, the amplitude of each component is corrected and constrained to obtain the optimal motion vector.

[0090] In this embodiment, obtaining the optimal bat individual in the current iteration specifically includes:

[0091] Based on the frequency adjustment amount in the optimal action vector, the search frequency of the individual bat is updated to obtain the updated frequency adjustment amount, which is the sum of the search frequency and the frequency adjustment amount of the previous iteration.

[0092] Based on the velocity correction in the optimal action vector and the coupling parameter vector of the optimal bat individual, the velocity vector of the bat individual is updated to obtain the updated velocity vector. The updated velocity vector is the sum of the original velocity vector and the velocity correction, and is formed by weighting and correcting the difference in coupling parameters between the bat individual and the global optimal individual in combination with the search frequency.

[0093] Based on the updated velocity vector and the position perturbation in the optimal motion vector, the coupling parameter vector of the individual bat is updated to obtain the updated position vector, which is obtained by combining the original position vector, the updated velocity vector, and the position perturbation.

[0094] The updated frequency adjustment, the updated velocity vector, and the updated position vector are concatenated to form the updated bat individual coupling parameter vector.

[0095] The updated bat individual coupling parameter vector is input into the optical fiber coupling system. The corresponding optical power output signal is collected by the control actuator, and the fitness value of the bat individual is calculated. The bat individual with the highest fitness value is selected as the optimal bat individual for the current iteration. The calculation process of the fitness value of the bat individual is based on the intensity of the optical power output signal. The collected optical power signal is compared with the target maximum optical power, and the minimum coupling loss is used as the evaluation criterion to calculate the fitness value of each bat individual.

[0096] In this embodiment, the output of the coupling parameter vector of the globally optimal bat individual specifically includes:

[0097] Adaptively correct the loudness adjustment and pulse emission rate adjustment of the current best bat individual to obtain the correction result, and synchronously update the parameter state of the improved latent space enhancement-bat algorithm with the correction result.

[0098] By updating the parameter state of the improved latent space reinforcement-bat algorithm, the policy parameters of the latent space reinforcement policy network module are adaptively updated to form the updated policy parameters. The update process adjusts the policy parameters along the direction of increasing reward value.

[0099] The termination condition is determined based on the updated strategy parameters. The maximum number of iterations is set to a preset upper limit. When the number of iterations reaches the preset upper limit, the coupling parameter vector of the globally optimal bat individual is output.

[0100] In this embodiment, the re-invocation of the improved latent space enhancement-bat algorithm specifically includes:

[0101] Based on the coupling parameter vector of the globally optimal individual bat, the coupling parameter state of the fiber optic coupling system is adjusted to the state corresponding to the globally optimal coupling parameter vector by controlling the actuator, and a coupling parameter state locking operation is performed to form the globally optimal coupling state.

[0102] Based on the global optimal coupling state, online monitoring parameters for optical power signals are set, and optical power signals are collected according to the set sampling period. The moving average value of optical power is calculated within each monitoring window. The online monitoring parameters include sampling period, monitoring window length, power threshold, and power drift threshold.

[0103] When the sliding average optical power is less than the preset power threshold, a re-optimization trigger signal is generated. The generation of the re-optimization trigger signal is based on the real-time optical power monitoring module. When the sliding average optical power is lower than the preset power threshold within a continuous sampling period, the monitoring module determines that the system coupling state deviates from the optimal range and automatically outputs a digital trigger signal to the control unit.

[0104] When the re-optimization trigger signal is detected to be active, the current coupling parameter status of the fiber optic coupling system is read and input to the control unit along with the current optical power signal.

[0105] An optimized startup command is generated in the control unit, the optimization process of the improved latent space enhancement-bat algorithm is invoked again, and parameter re-optimization calculation is performed.

[0106] Example 1:

[0107] An automatic fiber optic coupling experimental system was used as the subject of this study. The system consists of a light source module, a collimating lens assembly, a six-dimensional precision displacement platform, a fiber clamping device, an optical power detection unit, and a control computer. The light source is a 1550nm semiconductor laser with a stable output power of 10mW; the optical fiber is a single-mode SMF-28e fiber with FC / APC end-face polishing; and an Agilent 8163B power meter with a measurement accuracy of ±0.01dB is used for detection.

[0108] During the experiment, the system's optical power data was first acquired and preprocessed using mean filtering and normalization to remove high-frequency noise. Then, a coupling parameter search space model was constructed based on the characteristics of optical power variation. The coupling parameter vector includes six components: X, Y, and Z displacement parameters at the fiber end, as well as pitch angle, yaw angle, and spacing parameters. The system limited the value range of each parameter to ±50μm and ±1.5°, respectively, and quantized them with steps of 0.2μm and 0.05°.

[0109] In the initialization phase, an improved latent space reinforcement-bat algorithm is used to generate 40 individual bats. The coupling parameter vector of each individual is mapped into the latent space through reversible flow encoding to form a latent feature vector. The initial loudness of the algorithm is set to 0.9, the pulse emission rate is set to 0.1, and the search frequency range is [0.5, 2.5]. In the evolution phase, the action output is calculated through the latent space reinforcement policy network module to achieve joint optimization of frequency adjustment, velocity correction, and position perturbation. After each iteration, the fitness value is calculated based on the optical power feedback signal, and the policy parameters are adaptively adjusted using a reinforcement learning reward mechanism.

[0110] To compare and verify the algorithm performance, three sets of experiments were conducted: the standard bat algorithm, the improved particle swarm optimization algorithm, and the improved latent space reinforcement-bat algorithm proposed in this invention. All experiments were performed under the same conditions, with each set running independently 20 times. The maximum optical power, the number of convergence iterations, the optimization time, and the optical power stability were used as the main performance indicators.

[0111] Table 1. Experimental results comparing fiber coupling performance based on different algorithms

[0112]

[0113] As shown in Table 1, under the same experimental conditions, the improved latent space enhancement-bat algorithm proposed in this invention significantly outperforms the traditional algorithm in fiber coupling optimization performance. Regarding optical power output, the maximum output power of the improved latent space enhancement-bat algorithm reaches -1.42 dBm, an improvement of approximately 0.4 dB compared to the standard bat algorithm and approximately 0.25 dB compared to the improved particle swarm optimization algorithm, indicating more accurate optical field matching and lower coupling loss.

[0114] In terms of convergence performance, the algorithm has an average of 96 iterations, which is about 42% less than the standard bat algorithm and about 32% less than the improved particle swarm optimization algorithm, demonstrating faster convergence speed and better optimization efficiency. This is mainly due to the collaborative optimization mechanism of latent space mapping and reinforcement policy network.

[0115] In terms of optimization efficiency, the improved latent space reinforcement-bat algorithm has an average optimization time of only 15.7 seconds, which is about 41% shorter than the standard bat algorithm, achieving fast and high-precision online parameter optimization. Regarding system stability, the optical power stability decreased from 0.048 to 0.021, output fluctuations were significantly reduced, and the algorithm's ability to suppress noise and drift was significantly enhanced.

[0116] Furthermore, the improved latent space reinforcement-bat algorithm achieved a global optimal hit rate of 93.2% in multiple experiments, significantly higher than the comparative algorithms, indicating its strong global search capability and result consistency. In summary, the method of this invention achieves significant improvements in fiber coupling efficiency, convergence speed, optimization accuracy, and system robustness, verifying its superiority and practical value in complex multi-parameter fiber coupling systems.

Claims

1. A method for improving fiber coupling efficiency based on the bat algorithm, characterized in that, Includes the following steps: Acquire and preprocess optical power data from the fiber optic coupling system; Based on the preprocessed optical power data, the coupling parameter vector and value range of the optical fiber coupling system are set, and a coupling parameter search space model is constructed. Based on the coupling parameter search space model, an improved latent space reinforcement-bat algorithm is initialized and the initial coupling parameter vector is output. The initial coupling parameter vector is input into the latent space coding module of the improved latent space enhancement-bat algorithm, and reversible flow coding is introduced to generate latent space vectors. The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module, and the optimal action vector is calculated and output through a differentiable policy flow mapping mechanism. The optimal action vector is input into the bat individual evolution update module of the improved latent space reinforcement-bat algorithm to obtain the optimal bat individual in the current iteration. The current best bat individual is input into the adaptive adjustment and termination control module of the improved latent space enhancement-bat algorithm. The loudness and pulse emission rate of the current best bat individual are adaptively corrected, and the latent space enhancement strategy parameters are adjusted according to the feedback signal of optical power data. The coupling parameter vector of the global best bat individual is output. Based on the coupling parameter vector of the globally optimal bat individual, the control actuator adjusts the fiber optic coupling system to the corresponding coupling parameter state and locks it. During the operation of the fiber optic coupling system, the optical power signal is continuously monitored. When the optical power signal is detected to be less than the preset threshold, the improved latent space enhancement-bat algorithm is called again to perform parameter optimization calculation. The construction of the coupling parameter search space model specifically includes: Based on the preprocessed optical power signal, optical power response features are extracted, including the rate of change of optical power amplitude, the trend of coupling loss, and the steady-state power distribution of the system. Define a coupling parameter vector, and set lower and upper limits for each parameter component based on the variation range of optical power response characteristics to form a lower limit vector and an upper limit vector. The coupling parameter vector includes fiber end position parameters, fiber end angle parameters, and optical component spacing parameters. The feasible region of the coupling parameters is established based on the lower limit vector and the upper limit vector. The linear dependency between the components of each coupling parameter and the feasible state of the system are defined by the linear constraint matrix and the linear constraint vector, and the feasible region boundary model is obtained. Under the constraints of the feasible domain boundary model, the parameter resolution step size is set according to the dynamic variation range of the preprocessed optical power data and the sensitivity of the coupling parameters. The step size quantization and normalization of each component of the coupling parameter vector are performed using the boundary information of the lower limit vector and the upper limit vector to obtain the quantized coupling parameter vector and the normalized coupling parameter vector. The lower bound vector, upper bound vector, linear constraint matrix, linear constraint vector, parameter resolution step size, quantized coupling parameter vector, and normalized coupling parameter vector are combined to form a coupling parameter search space model. The generation of the latent space vector specifically includes: The initial coupling parameter vector is normalized by mapping each parameter value of the initial coupling parameter vector to a normalized interval of zero to one proportionally, thereby generating a normalized coupling parameter vector. A coupling parameter correlation matrix is ​​constructed based on the normalized coupling parameter vector, and the coupling parameter sensitivity coefficient is calculated. The construction of the coupling parameter correlation matrix involves performing covariance analysis on the normalized values ​​of each coupling parameter under multiple sets of optical power data, calculating the linear correlation coefficient and nonlinear coupling degree between each parameter component, and combining them by rows and columns to form the coupling parameter correlation matrix. The calculation process of the coupling parameter sensitivity coefficient involves applying small perturbations to each component of the normalized coupling parameter vector in sequence, recording the changes in optical power output before and after the perturbation, and obtaining the local response degree of the component to the optical power output by comparing the ratio of the change in optical power output to the corresponding parameter perturbation amount. The ratio is defined as the coupling parameter sensitivity value of the component. The normalized coupling parameter vector and the coupling parameter sensitivity coefficient are input into the latent space coding module of the improved latent space enhancement-Bat algorithm. The normalized coupling parameter vector is then weighted by a feature weighting layer to obtain a weighted normalized coupling parameter vector. The weighted normalized coupled parameter vector is used for nonlinear feature mapping using reversible flow coding to output a latent space vector, which refers to the state of an individual bat in the latent space feature domain.

2. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The preprocessing includes noise filtering, smoothing, and normalization preprocessing.

3. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The output of the initial coupling parameter vector specifically includes: The improved latent space enhancement-bat algorithm is loaded and initialized based on the coupled parameter search space model. The initialization parameters include population size, maximum number of iterations, search frequency range, initial loudness and initial pulse emission rate. Projecting the initialization parameters onto the linear constraint matrix and linear constraint vector of the coupling parameter search space model generates candidate coupling parameter vectors. The generation of the candidate coupling parameter vectors is based on the coupling parameter samples generated by the initialization parameters. The coupling parameter samples are substituted into the constraint conditions defined by the linear constraint matrix and linear constraint vectors. The coupling parameter samples that do not meet the constraint conditions are corrected, and the coupling parameter samples that meet all constraint conditions are obtained as candidate coupling parameter vectors. For each candidate coupling parameter vector, an initial coupling parameter vector is generated for each bat individual within the range of the lower bound vector and the upper bound vector.

4. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The output of the optimal action vector specifically includes: The latent space vector is input into the latent space reinforcement-bat algorithm latent space reinforcement policy network module. Linear mapping and nonlinear activation calculation are performed through a differentiable policy flow mapping mechanism to obtain latent space features. The latent space reinforcement policy network module includes an input layer, a hidden layer and an action generation layer. The latent space features are input into the hidden layer of the latent space reinforcement strategy network module, and feature mapping is performed through the weight matrix and a non-linear activation function is applied to output the original action vector. The original motion vector is subjected to amplitude constraint processing to obtain a standardized motion output; The standardized motion output is categorized into frequency adjustment, velocity correction, position disturbance, loudness adjustment, and pulse emission rate adjustment according to parameter type. Based on preset scaling factors and mapping rules, the amplitude of each component is corrected and constrained to obtain the optimal motion vector.

5. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The specific steps for obtaining the optimal bat individual in the current iteration include: Based on the frequency adjustment amount in the optimal action vector, the search frequency of the individual bat is updated to obtain the updated frequency adjustment amount, which is the sum of the search frequency and the frequency adjustment amount of the previous iteration. Based on the velocity correction in the optimal action vector and the coupling parameter vector of the optimal bat individual, the velocity vector of the bat individual is updated to obtain the updated velocity vector. The updated velocity vector is the sum of the original velocity vector and the velocity correction, and is formed by weighting and correcting the difference in coupling parameters between the bat individual and the global optimal individual in combination with the search frequency. Based on the updated velocity vector and the position perturbation in the optimal motion vector, the coupling parameter vector of the individual bat is updated to obtain the updated position vector, which is obtained by combining the original position vector, the updated velocity vector, and the position perturbation. The updated frequency adjustment, the updated velocity vector, and the updated position vector are concatenated to form the updated bat individual coupling parameter vector. The updated bat individual coupling parameter vector is input into the optical fiber coupling system. The corresponding optical power output signal is collected by the control actuator, and the fitness value of the bat individual is calculated. The bat individual with the highest fitness value is selected as the optimal bat individual for the current iteration. The calculation process of the fitness value of the bat individual is based on the intensity of the optical power output signal. The collected optical power signal is compared with the target maximum optical power, and the minimum coupling loss is used as the evaluation criterion to calculate the fitness value of each bat individual.

6. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The output of the coupling parameter vector of the globally optimal bat individual specifically includes: Adaptively correct the loudness adjustment and pulse emission rate adjustment of the current best bat individual to obtain the correction result, and synchronously update the parameter state of the improved latent space enhancement-bat algorithm with the correction result. By updating the parameter state of the improved latent space reinforcement-bat algorithm, the policy parameters of the latent space reinforcement policy network module are adaptively updated to form the updated policy parameters. The update process adjusts the policy parameters in the direction of increasing reward value. The termination condition is determined based on the updated strategy parameters. The maximum number of iterations is set to a preset upper limit. When the number of iterations reaches the preset upper limit, the coupling parameter vector of the globally optimal bat individual is output.

7. The method for improving fiber coupling efficiency based on the bat algorithm according to claim 1, characterized in that, The re-invocation of the improved latent space reinforcement-bat algorithm specifically includes: Based on the coupling parameter vector of the globally optimal individual bat, the coupling parameter state of the fiber optic coupling system is adjusted to the state corresponding to the globally optimal coupling parameter vector by controlling the actuator, and a coupling parameter state locking operation is performed to form the globally optimal coupling state. Based on the global optimal coupling state, online monitoring parameters for optical power signals are set, and optical power signals are collected according to the set sampling period. The moving average value of optical power is calculated within each monitoring window. The online monitoring parameters include sampling period, monitoring window length, power threshold, and power drift threshold. When the sliding average optical power is less than the preset power threshold, a re-optimization trigger signal is generated. The generation of the re-optimization trigger signal is based on the real-time optical power monitoring module. When the sliding average optical power is lower than the preset power threshold within a continuous sampling period, the monitoring module determines that the system coupling state deviates from the optimal range and automatically outputs a digital trigger signal to the control unit. When the re-optimization trigger signal is detected to be active, the current coupling parameter status of the fiber optic coupling system is read and input to the control unit along with the current optical power signal. An optimized startup command is generated in the control unit, which re-invokes the optimization process of the improved latent space enhancement-bat algorithm and performs parameter re-optimization calculation.

Images