Optimization method for improving gain of inter-band cascade laser

Abstract

The invention relates to an optimization method for improving the gain of an inter-band cascade laser, and the method comprises the steps: obtaining the structure parameters of a target inter-band cascade laser model, and carrying out the preprocessing; selecting an energy band model of the target inter-band cascade laser model; obtaining a parameter matrix of the model, and obtaining an energy eigenvalue and a wave function of a valence band according to the parameter matrix; determining a momentum matrix element and a Fermi level, and determining an optical gain value of the model according to the momentum matrix element and the Fermi level; and optimizing the target inter-band cascade laser model by using different optimization algorithms to obtain structure parameters at the optimal gain. The method has the beneficial effects that a large number of nonlinear and multi-dimensional parameter relationships can be processed, an accurate mapping relationship between the optical gain and various parameters is helped to be constructed, and the dependence on an accurate physical model is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of mid-infrared filters, and more specifically, to an optimization method for improving the gain of an inter-band cascade laser. Background Art

[0002] Since the 1950s, the mid-infrared region has attracted increasing attention for its military applications. Interband cascade lasers are semiconductor lasers based on quantum well or quantum cascade structures, which are widely used in gas detection, infrared spectroscopy and industrial monitoring in the mid-infrared (3–5 μm) band.

[0003] The wavelength range of the mid-infrared region is 3 to 30 microns, which is regarded as the molecular fingerprint band, where the spectrum provides key information about the material structure for physical, chemical and biological sciences. Therefore, interband cascade lasers in the mid-infrared field show great potential not only in scientific research, but also in daily life applications, military applications and even in material exploration in extraterrestrial space. The key to interband cascade lasers lies in their unique layered structure, which enables carriers to be re-injected and cascaded in multiple layers, thereby enhancing light generation. The interband cascade laser structure consists of a series of gain regions and injection regions, each of which has the function of interband transition to achieve light amplification.

[0004] With the advancement of machine learning technology, more and more machine learning methods are being applied to the optimization of laser structures. Traditional machine learning, such as Newton's method, uses the second-order derivative information (Hessian matrix) of the objective function to update parameters, but it has disadvantages such as high computational complexity, unsuitability for high-dimensional problems, and large memory requirements. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the prior art and to propose an optimization method for improving the gain of an inter-band cascade laser.

[0006] In a first aspect, an optimization method for improving the gain of an inter-band cascade laser is provided, comprising:

[0007] Step 1, obtaining the structural parameters of the target interband cascade laser model and performing preprocessing;

[0008] Step 2, selecting the band model of the target interband cascade laser model;

[0009] Step 3: obtaining a parameter matrix of the model according to the structural parameters and the energy band model, and obtaining the energy eigenvalue and wave function of the valence band according to the parameter matrix;

[0010] Step 4, determining a momentum matrix element and a Fermi level according to the energy eigenvalue and the wave function, and determining an optical gain value of the model according to the momentum matrix element and the Fermi level;

[0011] Step 5: Optimize the target interband cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

[0012] Preferably, in step 1, for the target inter-band cascade laser model, the gain layer is set to be multiple layers as a basic gain structure; the materials and thicknesses of different gain layers are different; and the preprocessing includes: defining the parameters of each gain layer in a dictionary form and unifying the units.

[0013] Preferably, in step 2, the energy band model is an eight-band model.

[0014] Preferably, in step 5, the optimization algorithm is RMSProp algorithm or AdamW algorithm.

[0015] In a second aspect, an optimization system for improving the gain of an inter-band cascade laser is provided, which is used to perform any method described in the first aspect, including:

[0016] A first acquisition module is used to acquire structural parameters of a target inter-band cascade laser model and perform preprocessing;

[0017] A selection module for selecting an energy band model of a target interband cascade laser model;

[0018] A second acquisition module is used to acquire a parameter matrix of the model according to the structural parameters and the energy band model, and to acquire an energy eigenvalue and a wave function of the valence band according to the parameter matrix;

[0019] A determination module, used to determine a momentum matrix element and a Fermi level according to the energy eigenvalue and the wave function, and to determine an optical gain value of a model according to the momentum matrix element and the Fermi level;

[0020] The optimization module is used to optimize the target inter-band cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

[0021] According to a third aspect, a computer storage medium is provided, wherein a computer program is stored in the computer storage medium; when the computer program is executed on a computer, the computer executes any method described in the first aspect.

[0022] In a fourth aspect, an electronic device is provided, including:

[0023] Memory, used to store computer programs;

[0024] A processor is used to execute the computer program to implement any method as described in the first aspect.

[0025] The beneficial effects of the present invention are as follows: the optical gain of the interband cascade laser depends on a variety of physical parameters (material properties, carrier concentration, temperature, etc.), and the relationship between them is complex. The present invention utilizes machine learning methods to process a large number of nonlinear, multi-dimensional parameter relationships, helps to construct an accurate mapping relationship between optical gain and a variety of parameters, and reduces dependence on precise physical models. In traditional methods, structural parameters need to be optimized through cumbersome trial and error methods or time-consuming numerical calculations. The application of machine learning can quickly find the optimal combination of laser design parameters and optical gain, providing a faster optimization design process. The thickness of each layer in the interband cascade laser is changed by the above-mentioned gradient descent method to find the optimal optical gain. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 A flow chart of an optimization method for improving the gain of an inter-band cascade laser provided by the present invention;

[0027] Figure 2 A structural diagram of the inter-band cascade laser quantum well provided by the present invention;

[0028] Figure 3a and Figure 3b A schematic diagram of the heterogeneous energy dispersion of the valence band and conduction band provided by the present invention;

[0029] Figure 4 A schematic diagram of the search path of the AdamW optimizer provided by the present invention;

[0030] Figure 5 The momentum matrix element diagram under the eight-band model provided by the present invention;

[0031] Figure 6 This is the optical gain diagram under the eight-band model provided by the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described below in conjunction with embodiments. The description of the following embodiments is only used to help understand the present invention. It should be noted that for ordinary persons in the art, without departing from the principle of the present invention, the present invention can also be modified in some ways, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

[0033] Embodiment 1:

[0034] In order to solve the problems of the prior art, Embodiment 1 of the present application provides an optimization method for improving the gain of an inter-band cascade laser, comprising:

[0035] Step 1: Obtain the structural parameters of the target interband cascade laser model and perform preprocessing.

[0036] Specifically, in step 1, obtain the structure of the interband cascade laser that you want to optimize (such as AlSb / InAs / InGaSb / InAs / AlSb, etc.), and find the parameters of the required materials in reference books and papers, such as lattice, constant band gap, spin-orbit energy, etc., and put all the parameters into the data set to prepare for subsequent calculations. It should be noted that the material parameters are used with known parameters and structures.

[0037] In addition, for the target inter-band cascade laser model, the gain layer is set to be multi-layer as a basic gain structure; the materials and thicknesses of different gain layers are different; the preprocessing includes: specifying the parameters of each gain layer in the form of a dictionary in python and unifying the units.

[0038] Step 2: Select the band model of the target interband cascade laser model.

[0039] Specifically, the choice of band model depends on the required calculation accuracy and specific application. Different approximate models are selected to describe the band structure. The model is usually used to calculate the electronic properties, band gaps and behavior of semiconductor materials in low-dimensional structures such as quantum wells or quantum dots, which provide basic data for calculating optical gain.

[0040] Step 3: Obtain a parameter matrix of the model according to the structural parameters and the energy band model, and obtain the energy eigenvalue and wave function of the valence band according to the parameter matrix.

[0041] Specifically, the material parameters are selected according to step 1 to fill in the matrix. Different models have different matrices. In order to obtain the energy eigenvalues ​​and wave functions of the valence band in semiconductor materials, the ARPACK package used to quickly find the eigenvalues / eigenvectors of large sparse matrices is selected to solve the model.

[0042] Step 4: Determine the momentum matrix element and the Fermi level according to the energy eigenvalue and the wave function, and determine the optical gain value of the model according to the momentum matrix element and the Fermi level.

[0043] In step 3, the energy eigenvalue and wave function have been obtained. The momentum matrix element and Fermi level are determined based on the values, and then substituted into the formula of optical gain to determine the size of the optical gain of the measured model.

[0044] Step 5: Optimize the target interband cascade laser model using different optimization algorithms, compare it with the structure before optimization, and obtain the structural parameters at the optimal gain.

[0045] Specifically, an optimization program is designed to adjust the material thickness to obtain the optimal optical gain problem. Generally speaking, a numerical optimization program is an iterative algorithm. For the existing iterative algorithms, including RMSProp, AdamW algorithm, etc., different optimization algorithms are selected for comparison to find the best convergence point and obtain the best optical gain. The machine learning method used in this application mainly includes RMSProp, AdamW algorithm and momentum method in step 5.

[0046] Embodiment 2:

[0047] Based on Example 1, Example 2 of the present application provides a more specific optimization method for improving the gain of an inter-band cascade laser, including:

[0048] Step 1: Obtain the structural parameters of the target interband cascade laser model and perform preprocessing.

[0049] For example, optical parameters of type II semiconductor materials for the target interband cascade laser model are shown in Table 1.

[0050] Table 1

[0051]

[0052] In addition, a five-layer structure is applied as the basic gain structure, the initial thickness is set, and the parameters of each layer of material are specified in the form of a dictionary in python, and the units are unified. The initial units of the thickness of each layer are given in Table 2.

[0053] Table 2

[0054]

[0055] Step 2: Select the band model of the target interband cascade laser model.

[0056] For example, in step 2, the energy band model is an eight-band model. The eight-band model is mainly in the form of an 8*8 matrix, which represents the energy interaction of these quantum states by constructing a Hamiltonian containing eight ground states, and takes into account the coupling effect between the valence band and the conduction band. In this way, the energy distribution and transition behavior between the conduction band, heavy holes, light holes and spin-split bands can be accurately described. Each element in the matrix can be found in Table 1. The electronic properties, band gaps and properties of semiconductor materials in low-dimensional structures such as quantum wells or quantum dots are calculated, including the band gaps of the conduction band, valence band (heavy holes, light holes and spin-orbit split bands). The results show Figure 3a and Figure 3b The electron energy bands (E1, E2 and E3) shown in the figure are all semi-parabolic, while the valence subbands represented by hh (heavy hole state), lh (light hole state) and so (spin-orbit splitting band) do not have parabolic properties.

[0057] Step 3: Obtain a parameter matrix of the model according to the structural parameters and the energy band model, and obtain the energy eigenvalue and wave function of the valence band according to the parameter matrix.

[0058] For example, the Hamiltonian matrix of the eight-band model is expressed as:

[0059]

[0060] Step 4: Determine the momentum matrix element and the Fermi level according to the energy eigenvalue and the wave function, and determine the optical gain value of the model according to the momentum matrix element and the Fermi level.

[0061] For example, Figure 5 The momentum matrix element diagram under this model is given, which shows the size of the momentum matrix element at different wave vector positions. At the same time, the dongl matrix elements of the three bands hh (heavy hole state), lh (light hole state) and so (spin-orbit splitting band) are given. Substituting them into the formula of optical gain, the size of the optical gain of the measured model is determined. The size of the optical gain model under this model is as follows Figure 6 As shown, the optical gain has been plotted as a function of the photon energy. It can be seen that when the zero gain condition f is satisfied c -f v >0, the nanoheterostructure will produce optical gain, otherwise the gain will not exist.

[0062] Step 5: Optimize the target interband cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

[0063] Repeat steps 3 and 4 to optimize the models with different structures and materials using different optimization algorithms. In the example, the AdamW optimization algorithm is selected, and its search path is Figure 4 As shown in , compared with the structure before optimization, the structural parameters for achieving the best gain are finally obtained. The results are shown in Table 2. It can be seen that after optimization, the optical gain is greatly increased, which is valuable in practical applications and helps to find the theoretical optimal thickness under various field strengths, which is conducive to the design and development of quantum well lasers such as interband cascade lasers.

[0064] It should be noted that the parts in this embodiment that are the same or similar to those in Embodiment 1 can be referenced to each other and will not be described in detail in this application.

[0065] Embodiment 3:

[0066] On the basis of embodiments 1 and 2, embodiment 3 of the present application provides an optimization system for improving the gain of an inter-band cascade laser, including:

[0067] A first acquisition module is used to acquire structural parameters of a target inter-band cascade laser model and perform preprocessing;

[0068] A selection module for selecting an energy band model of a target interband cascade laser model;

[0069] A second acquisition module is used to acquire a parameter matrix of the model according to the structural parameters and the energy band model, and to acquire an energy eigenvalue and a wave function of the valence band according to the parameter matrix;

[0070] A determination module, used to determine a momentum matrix element and a Fermi level according to the energy eigenvalue and the wave function, and to determine an optical gain value of a model according to the momentum matrix element and the Fermi level;

[0071] The optimization module is used to optimize the target inter-band cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

[0072] Specifically, the system provided in this embodiment is a system corresponding to the method provided in Embodiments 1 and 2. Therefore, the parts in this embodiment that are the same or similar to Embodiments 1 and 2 can be referenced to each other and will not be repeated in this application.

Claims

1. An optimization method for improving the gain of an inter-band cascade laser, characterized in that: include: Step 1, obtaining the structural parameters of the target interband cascade laser model and performing preprocessing; Step 2, selecting the band model of the target interband cascade laser model; Step 3: obtaining a parameter matrix of the model according to the structural parameters and the energy band model, and obtaining the energy eigenvalue and wave function of the valence band according to the parameter matrix; Step 4, determining a momentum matrix element and a Fermi level according to the energy eigenvalue and the wave function, and determining an optical gain value of the model according to the momentum matrix element and the Fermi level; Step 5: Optimize the target interband cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

2. The optimization method for improving the gain of an interband cascade laser according to claim 1, characterized in that: In step 1, for the target interband cascade laser model, the gain layer is set to multiple layers as a basic gain structure; The materials and thicknesses of different gain layers are different. The preprocessing includes: defining the parameters of each gain layer in a dictionary form and unifying the units.

3. The optimization method for improving the gain of an interband cascade laser according to claim 1, characterized in that: In step 2, the energy band model is an eight-band model.

4. The optimization method for improving the gain of an interband cascade laser according to claim 1, characterized in that: In step 5, the optimization algorithm is the RMSProp algorithm or the AdamW algorithm.

5. An optimization system for increasing the gain of an inter-band cascade laser, characterized in that: Used to perform the method according to any one of claims 1 to 4, comprising: A first acquisition module is used to acquire structural parameters of a target inter-band cascade laser model and perform preprocessing; A selection module for selecting an energy band model of a target interband cascade laser model; A second acquisition module is used to acquire a parameter matrix of the model according to the structural parameters and the energy band model, and to acquire an energy eigenvalue and a wave function of the valence band according to the parameter matrix; A determination module, used to determine a momentum matrix element and a Fermi level according to the energy eigenvalue and the wave function, and to determine an optical gain value of a model according to the momentum matrix element and the Fermi level; The optimization module is used to optimize the target inter-band cascade laser model using different optimization algorithms to obtain the structural parameters at the optimal gain.

6. A computer storage medium, characterized in that: The computer storage medium stores a computer program; when the computer program is executed on a computer, the computer executes any one of the methods described in claims 1 to 6.

7. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the method according to any one of claims 1 to 6.

Images