Orthogonal disturbance adaptive gain SPGD algorithm for laser coherent combination system

Abstract

The invention belongs to the field of laser coherent combination, discloses an orthogonal disturbance adaptive gain SPGD algorithm for a laser coherent combination system, and aims to solve the problem of low convergence speed when a traditional SPGD algorithm is applied to the coherent combination system. According to the method, in the iteration process, random disturbance voltage applied in a traditional SPGD algorithm is replaced with a pre-generated orthogonal matrix column vector, and self-adaptive adjustment of a gain coefficient is achieved in combination with a gain coefficient adjustment formula. According to the method, the phase of the laser beam can be efficiently controlled, the convergence speed of the algorithm is remarkably improved, and the orthogonal disturbance voltage matrix can be generated in advance and directly called in closed-loop iteration, so that hardware implementation is simpler and more convenient. In addition, the algorithm still shows excellent performance when being applied to a large-scale laser coherent combination system.

Description

Technical Field

[0001] The invention relates to the technical field of laser coherent synthesis, and in particular to an orthogonal perturbation adaptive gain SPGD algorithm for a laser coherent synthesis system. Background Art

[0002] High-energy lasers are increasingly used in defense, industry and other fields. Due to the influence of factors such as the nonlinear effect of the fiber gain medium and the thermal effect of the device, the output power of a single fiber laser is limited. The beam synthesis technology is used to coherently synthesize multiple laser beams so that the phases of each unit beam reaching the target surface are basically consistent, and finally the coherent superposition of each unit beam is achieved, which can ensure good beam quality while obtaining higher power output.

[0003] Fast and accurate control of the beam phase is one of the keys to achieve fiber laser coherent combining. According to the physical mechanism of phase control, phase control technology can be divided into passive phase control and active phase control. Since the passive phase control scheme is not suitable for high-power and large-element fiber laser coherent combining systems, most systems use active phase control technology. At present, active phase control methods mainly include direct detection and indirect detection. Direct detection measures the phase error between each unit laser by certain means and corrects it. Commonly used methods include heterodyne method and homodyne interferometry. Indirect detection continuously updates the control voltage through an algorithm to optimize the coherent combining effect. It mainly includes stochastic parallel gradient descent (SPGD) algorithm, machine learning algorithm and jitter algorithm. The indirect detection method usually only requires one photodetector to obtain the energy at the center of the far-field spot, which has the advantage of a simple control system structure.

[0004] Although the SPGD algorithm is widely used in laser coherent synthesis systems due to its advantages of parallel control and simple structure, the traditional SPGD algorithm has certain limitations. Specifically, the traditional SPGD algorithm uses a fixed gain coefficient and perturbation amplitude, which leads to a contradiction between the convergence speed and convergence effect of the algorithm. That is, when the gain coefficient or perturbation amplitude is small, the convergence speed of the algorithm is slow; and when these parameters are large, the algorithm is prone to fall into a local optimal solution. In addition, as the number of synthesized beam units increases, the number of iterations required for the convergence of the SPGD algorithm increases exponentially, making the implementation of the coherent synthesis system more difficult. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide an orthogonal perturbation adaptive gain SPGD algorithm for a laser coherent synthesis system. The orthogonal perturbation adaptive gain SPGD algorithm can efficiently control the phase of the laser beam and improve the algorithm convergence speed. When the algorithm is applied to a large-scale laser coherent synthesis system, it still has good performance.

[0006] The technical solution of the present invention to solve the above technical problems is:

[0007] The orthogonal perturbation adaptive gain SPGD algorithm used in the laser coherent synthesis system of the present invention corrects the phase of each laser. The laser coherent synthesis system mainly includes the following components: a seed laser, a fiber beam splitter, a phase modulator, a fiber collimation array, a focusing lens, a photodetector with a small hole, and an algorithm controller based on the orthogonal perturbation adaptive gain SPGD algorithm of the present invention. The phase correction process of the laser coherent synthesis system is as follows: the seed laser is divided into M beams through the fiber beam splitter, and the M beams are output after passing through the optical fiber, the phase modulator and the fiber collimation array in turn. The output light beam is focused by a focusing lens and finally projected onto the photodetector with a small hole, thereby measuring the optical power value of the M beams. The main control chip (FPGA) reads the optical power value as an algorithm controller and uses it as a performance evaluation function. The algorithm controller uses the orthogonal perturbation adaptive gain SPGD algorithm to calculate the M-dimensional control voltage matrix, and applies the control voltage to the phase modulator in parallel, thereby realizing the correction of the phase of each laser.

[0008] An orthogonal perturbation adaptive gain SPGD algorithm for a laser coherent synthesis system replaces the random perturbation voltage applied in a traditional SPGD algorithm with a pre-generated orthogonal matrix column vector during the iteration process, and combines the gain coefficient adjustment formula to achieve adaptive adjustment of the gain coefficient, thereby significantly improving the convergence speed of the algorithm; the algorithm comprises the following steps:

[0009] (S1), using the optical power values ​​of the M laser beams detected by the photodetector with a pinhole as a performance evaluation function, setting a gain coefficient, a phase control voltage, and calculating an initial performance evaluation function;

[0010] (S2) Generate an M×M dimensional random number matrix with a size between -0.5 and 0.5 in advance, orthogonalize the matrix and record the orthogonal matrix as G. The algorithm recycles the column vector of the orthogonal matrix G as the disturbance voltage δu during the n-th iteration process. (n) ;

[0011] (S3) Apply positive and negative disturbances to the phase modulator, use a photodetector to detect the optical power values ​​under the positive and negative disturbance voltages respectively, and calculate the corresponding change in the performance evaluation function

[0012] (S4), based on the adaptive gain coefficient update formula: Dynamically adjust the gain coefficient for each iteration;

[0013] (S5), according to the updated gain coefficient, adjust the phase control voltage matrix of this iteration and apply it to the phase modulator: u (n) =u (n-1) +γ (n) δu (n) δJ (n) ,u (n) is the M-dimensional phase control voltage matrix, u (n-1) is the M-dimensional phase control voltage matrix after the last iteration;

[0014] (S6), the updated system performance evaluation function J is detected by the photoelectric detector (n) And determine whether it meets the convergence conditions. If it does, the algorithm ends; if not, repeat the above steps.

[0015] Preferably, in step (S2), the orthogonal matrix G is obtained by orthogonalizing an M×M dimensional initial random matrix whose size is between -0.5 and 0.5, and the column vector of the matrix is ​​recycled in each iteration to replace the random perturbation voltage in the traditional SPGD algorithm.

[0016] Preferably, in step (S4), the gain coefficient updating formula is: Among them, γ (n) is the gain coefficient in the nth iteration process, which is used to replace the fixed gain coefficient to ensure that the algorithm can optimize the phase control effect in different iteration stages.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] 1. The orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system proposed in the present invention improves the perturbation mechanism in the traditional SPGD algorithm. Specifically, in the iterative process of the algorithm, the random perturbation voltage in the traditional SPGD algorithm is replaced with a pre-generated orthogonal matrix column vector. The column vectors of the orthogonal matrix are orthogonal to each other and have strong irrelevance, so that the phase space can be explored more efficiently. This improvement significantly reduces the redundancy and repeatability of the random perturbation, avoids the problem of unbalanced phase updates that may occur in the traditional random perturbation method, and thus accelerates the convergence process of the algorithm. In addition, when the algorithm is deployed on an FPGA, the pre-generated orthogonal matrix can be directly called to generate the perturbation signal without the need to design an additional pseudo-random signal generator, which simplifies the hardware implementation and improves the implementation efficiency of the system.

[0019] 2. The orthogonal perturbation adaptive gain SPGD algorithm used in the laser coherent synthesis system of the present invention adopts an adaptive gain coefficient adjustment mechanism. This mechanism effectively reduces the sensitivity of the traditional SPGD algorithm to the gain coefficient setting, enhances the stability and robustness of the algorithm, and enables the laser coherent synthesis system to reach the optimal coherence state in a shorter time, thereby improving the overall performance of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 A system block diagram of an orthogonal perturbation adaptive gain SPGD algorithm for a laser coherent synthesis system of the present invention;

[0021] Figure 2 It is a flow chart of the orthogonal perturbation adaptive gain SPGD algorithm for the laser coherent synthesis system of the present invention;

[0022] Figure 3 The evaluation function iteration curve of the orthogonal perturbation adaptive gain SPGD algorithm simulated 300 times in the embodiment of the present invention;

[0023] Figure 4 It is a comparison diagram of evaluation function iteration curves of an embodiment of the present invention; DETAILED DESCRIPTION

[0024] The present invention is further described in detail below in conjunction with embodiments and drawings, but the embodiments of the present invention are not limited thereto.

[0025] See also Figure 1 , the laser coherent synthesis system mainly includes the following components: a seed laser, a fiber beam splitter, a phase modulator, a fiber collimation array, a focusing lens, a photodetector with a pinhole, and an algorithm controller based on the orthogonal perturbation adaptive gain SPGD algorithm of the present invention. The phase correction process of the laser coherent synthesis system is as follows: the seed laser is divided into 7 beams by the fiber beam splitter, and the 7 beams are output after passing through the optical fiber, the phase modulator and the fiber collimation array in turn. The output light beam is focused by a focusing lens and finally projected onto the photodetector with a pinhole, so as to measure the optical power values ​​of the 7 beams. The main control chip (FPGA) reads the optical power value as an algorithm controller and uses it as a performance evaluation function. The algorithm controller uses the orthogonal perturbation adaptive gain SPGD algorithm to calculate the 7-dimensional control voltage matrix, and applies the control voltage to the phase modulator in parallel, so as to realize the correction of the phase of each laser.

[0026] The orthogonal perturbation adaptive gain SPGD algorithm was numerically simulated using MATLAB software. Each beam was simplified into a fundamental mode Gaussian beam with the same polarization. The number of coherent synthesis paths was set to 7. Each beam was arranged in a regular hexagonal array to obtain a high duty cycle, and the initial phase of the beam obeyed a random distribution of 0 to 2π. The power in the bucket was used as the system evaluation function, with a value range of 0 to 1.

[0027] See also Figure 2 , the orthogonal perturbation adaptive gain SPGD algorithm includes the following steps:

[0028] (S1), the optical power value of the 7 laser beams detected by the photodetector with a small hole after coherent synthesis is used as the performance evaluation function J, the gain coefficient γ is set to 4, and the phase control voltage u (0) ={0,…,0}, and calculate the initial performance evaluation function J (0) ;

[0029] (S2) Use the rand function in Matlab software to generate a 7×7 dimensional random number matrix with a size between -0.5 and 0.5, and call the orth function to orthogonalize it to obtain the orthogonal matrix G. The algorithm cyclically uses the column vector of the orthogonal matrix G as the disturbance voltage δu during the n-th iteration process. (n) ;

[0030] (S3) Apply positive and negative disturbances to the phase modulator, use a photodetector to detect the optical power value under the positive and negative disturbance voltages respectively, and calculate the corresponding change δJ of the performance evaluation function (n) =J + (n) -J - (n) ;

[0031] (S4), based on the adaptive gain coefficient update formula: Dynamically adjust the gain coefficient for each iteration;

[0032] (S5), according to the updated gain coefficient, adjust the phase control voltage matrix of this iteration and apply it to the phase modulator: u (n) =u (n-1) +γ (n) δu (n) δJ (n) ,u (n) is the 7-dimensional phase control voltage matrix, u (n-1) is the 7-dimensional phase control voltage matrix after the last iteration;

[0033] (S6), the updated system performance evaluation function J is detected by the photoelectric detector (n) In this embodiment, when the evaluation function value reaches 0.9, the algorithm is considered to have converged.(n) Whether the convergence condition is met, if it is met, the algorithm ends, if not, repeat the above steps.

[0034] Figure 3 The evaluation function iteration curve of the orthogonal perturbation adaptive gain SPGD algorithm in 300 simulations is shown.

[0035] To verify the performance of the orthogonal perturbation adaptive gain SPGD algorithm of the present invention, the algorithm of the present invention and the traditional SPGD algorithm (where the gain coefficient is set to 2) were simulated 300 times respectively, and the iterative curves of the evaluation function were averaged for comparison. Figure 4 The comparison diagram of the evaluation function iteration curve of the embodiment of the present invention and the traditional SPGD algorithm is shown. The results show that the orthogonal perturbation adaptive gain SPGD algorithm of the present invention has a significantly faster convergence speed than the traditional algorithm, so the method of the present invention has better correction ability.

[0036] The above is a preferred embodiment of the present invention, but the embodiment of the present invention is not limited to the above content. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. An orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system, characterized in that: In the iterative process, the random disturbance voltage applied in the traditional SPGD algorithm is replaced with a pre-generated orthogonal matrix column vector, and the adaptive adjustment of the gain coefficient is realized by combining the gain coefficient adjustment formula, thereby improving the convergence speed and phase control accuracy of the algorithm; the algorithm is applied to the laser coherent synthesis system to improve the coherent synthesis efficiency of the beam by controlling the phase of the laser beam. The system mainly includes a seed laser, a fiber beam splitter, a phase modulator, a fiber collimation array, a focusing lens, a photodetector with a pinhole, and a controller based on the above algorithm.

2. The orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system according to claim 1, characterized in that: The orthogonal perturbation adaptive gain SPGD algorithm includes the following steps: (S1), taking the optical power value of M laser beams detected by the photodetector with a pinhole as the performance evaluation function J, setting the gain coefficient γ, the phase control voltage, and calculating the initial performance evaluation function J (0) ; (S2) Generate an M×M dimensional random number matrix with a size between -0.5 and 0.5 in advance, orthogonalize the matrix and record the orthogonal matrix as G. The algorithm recycles the column vector of the orthogonal matrix G as the disturbance voltage δu during the n-th iteration process. (n) ; (S3) Apply positive and negative disturbances to the phase modulator, use a photodetector to detect the optical power value under the positive and negative disturbance voltages respectively, and calculate the corresponding change δJ of the performance evaluation function (n) =J + (n) -J - (n) ; (S4), based on the adaptive gain coefficient update formula: Dynamically adjust the gain coefficient for each iteration; (S5), according to the updated gain coefficient, adjust the phase control voltage matrix of this iteration and apply it to the phase modulator: u (n) =u (n-1) +γ (n) δu (n) δJ (n) ,u (n) is the M-dimensional phase control voltage matrix, u (n-1) is the M-dimensional phase control voltage matrix after the last iteration; (S6), the updated system performance evaluation function J is detected by the photoelectric detector (n) And determine whether it meets the convergence conditions. If it does, the algorithm ends; if not, repeat the above steps.

3. The orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system according to claim 1, characterized in that: The main control chip of the algorithm controller is a field programmable gate array (FPGA), and the hardware acceleration of the algorithm is realized by FPGA. The calculated M-dimensional phase control voltage matrix is ​​applied to the phase modulator in parallel to realize dynamic adjustment of the light beam phase; the performance evaluation function includes but is not limited to the bucket power detected by the photodetector, the maximum output power, the main lobe power, the synthetic beam quality factor or a combination thereof.

4. The orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system according to claim 2, characterized in that: In step (S2), the orthogonal matrix G is obtained by orthogonalizing an M×M dimensional initial random matrix with a size between -0.5 and 0.5, and the column vector of the matrix is ​​cyclically used in each iteration to replace the random perturbation voltage in the traditional SPGD algorithm.

5. The orthogonal perturbation adaptive gain SPGD algorithm for laser coherent synthesis system according to claim 2, characterized in that: In step (S4), the gain coefficient update formula is: Among them, γ (n) is the gain coefficient in the nth iteration process, which is used to replace the fixed gain coefficient to ensure that the algorithm can optimize the phase control effect in different iteration stages.

Images