A Design Method, Device and Storage Medium for Traveling-Wave Electrodes of Electro-Optic Modulators

Abstract

The present invention discloses a design method, device and storage medium for a traveling-wave electrode of an electro-optic modulator. The method process is as follows: determine the process parameters to be optimized, performance indicators, key process parameters, initially selected performance indicators of the traveling-wave electrode, and set constraint conditions; obtain the initial values of multiple groups of process parameters to be optimized and the evaluation results of the initially selected performance indicators, and thus select the initial values of several groups of process parameters to be optimized as initial conditions respectively; then establish an objective function, and use the external penalty function method to transform the optimization problem of the objective function of the traveling-wave electrode design with constraints into an unconstrained optimization problem; finally, use the AdaGrad algorithm to perform iterative solution in combination with the initial conditions to obtain the process parameter values of the traveling-wave electrode that meet the design requirements; the device and storage medium are used to implement the above method. The method of the present invention has the advantages of high efficiency, robustness and ease of use. Non-professional personnel can quickly generate a feasible solution based on the present invention through preset performance indicators.

Description

Technical Field

[0001] The present invention relates to the field of traveling wave electrode design methods, and specifically to a traveling wave electrode design method, device, and storage medium for an electro-optic modulator. Background Art

[0002] With the rapid development of artificial intelligence technology, optical computing has become an important direction to break through the bottleneck of traditional electronic computing due to its characteristics of high parallelism and low energy consumption. As a core device in optical communication and optical computing systems, the performance of an electro-optic modulator determines the efficiency and accuracy of optical signal modulation. With the continuous increase in the demand for electro-optic modulators, higher requirements are put forward for the design of electro-optic modulators: on the one hand, it is necessary to meet high-performance indicators under complex multi-physical field coupling, and on the other hand, it is necessary to adapt to a rapidly iterative development cycle to cope with market changes.

[0003] The traditional optimization method for the traveling wave electrode of an electro-optic modulator heavily relies on the empirical adjustment of designers, resulting in a long development cycle and high trial-and-error costs. On the one hand, it is difficult to meet the requirements of the optical computing system for rapid device iteration, and on the other hand, it severely limits the possibility of non-professional personnel entering this field. Therefore, there is an urgent need for an electro-optic modulator design method with high efficiency, robustness, and ease of use, which can not only reduce the learning and operation thresholds of non-professional personnel who have not designed in this field through an automated process, but also provide a high-precision optimization tool for professional personnel in this field, thereby promoting the large-scale application of optical computing technology. Summary of the Invention

[0004] The present invention provides a traveling wave electrode design method, device, and storage medium for an electro-optic modulator to solve the problems of relying on manual experience and difficulty for non-professional personnel to participate in the design existing in the traveling wave electrode of the existing electro-optic modulator.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0006] A traveling wave electrode design method for an electro-optic modulator, the process is as follows:

[0007] Determine the process parameters to be optimized and the key process parameters among them, determine the performance indicators and select the initial performance indicators from them, and set the constraint conditions for the key process parameters and performance indicators;

[0008] Based on the evaluation results of the initial performance indicators corresponding to each initial value of multiple groups of process parameters to be optimized, select several groups of initial values of the process parameters to be optimized as initial conditions;

[0009] Establish an objective function for traveling wave electrode design with the performance indicators as variables, and use the constraint conditions of the key process parameters and performance indicators as the constraints of the objective function;

[0010] Transform the objective function optimization problem with constraints into an unconstrained optimization problem;

[0011] Use the AdaGrad algorithm with the process parameters to be optimized as variables to iterate on the unconstrained optimization problem; at each iteration, use the initial value of a set of process parameters to be optimized in the initial conditions as the variable initial point, solve to obtain the variable optimal solution and the corresponding iterative result of the unconstrained optimization problem; at multiple iterations, use different sets of initial values as the variable initial point, obtain the variable optimal solutions and the corresponding iterative results for each iteration; select the variable optimal solution corresponding to the iterative result that meets the requirements as the process parameter value of the traveling wave electrode that meets the design requirements.

[0012] Further, the structural types of the traveling wave electrodes include ordinary traveling wave electrodes and capacitive load type traveling wave electrodes.

[0013] Further, the process parameters to be optimized include signal electrode width, electrode spacing, ground electrode width, signal electrode thickness, and ground electrode thickness, and the key process parameters are signal electrode width and electrode spacing.

[0014] Further, the performance indicators include microwave refractive index, characteristic impedance, insertion loss, and half-wave voltage, and select the microwave refractive index and characteristic impedance as the initially selected performance indicators.

[0015] Further, construct an initially selected performance indicator evaluation result calculation function with the initially selected performance indicators as variables, and the initially selected performance indicator evaluation result calculation function includes an initially selected performance indicator objective function and an initially selected performance indicator penalty term;

[0016] Calculate the initially selected performance indicator evaluation results and the initially selected performance indicator penalty terms of each set of initial values of the process parameters to be optimized based on the initially selected performance indicator evaluation result calculation function;

[0017] When selecting, take the initial values of several sets of process parameters to be optimized with the initially selected performance indicator evaluation results less than the preset threshold as the initial conditions, or take the initial values of several sets of process parameters to be optimized with the penalty term being 0 as the initial conditions.

[0018] Further, sort the iterative results of the unconstrained optimization problem corresponding to the variable optimal solutions obtained at each iteration in ascending order, and take the variable optimal solution corresponding to the smallest iterative result as the process parameter value of the traveling wave electrode that meets the design requirements.

[0019] Further, after each gradient update in the Adagrad algorithm, project the parameters to the nearest discrete point.

[0020] An electronic device includes a processor and a memory. When the processor reads and runs program instructions in the memory, it executes the above-mentioned design method for the traveling-wave electrode of the electro-optic modulator.

[0021] A storage medium stores program instructions. When the program instructions are read and run, they execute the above-mentioned design method for the traveling-wave electrode of the electro-optic modulator.

[0022] The present invention is a design method for the traveling-wave electrode of an electro-optic modulator based on the collaborative optimization of the external penalty function method and the AdaGrad algorithm, and is particularly applicable to the structural parameter optimization design of the traveling-wave electrode in a lithium niobate (LiNbO 3 ) electro-optic modulator.

[0023] The method of the present invention solves the problems of low optimization efficiency and slow convergence speed caused by multi-parameter coupling, non-linear constraints, and discrete process limitations in traditional designs by integrating the constraint processing ability of the external penalty function method and the adaptive gradient optimization characteristics of the AdaGrad algorithm, and is applicable to the development of high-performance electro-optic modulators in the fields of high-speed optical communication, microwave photonics, and integrated photonic chips. This method has the advantages of high efficiency, robustness, and ease of use, and can realize the automatic optimization of the traveling-wave electrode of the electro-optic modulator.

[0024] In the present invention, by using the external penalty function method to handle the constraint conditions and combining the characteristics of the adaptive step size of the Adagrad algorithm, the optimization speed can be significantly improved while ensuring the accuracy. Through simulation experiments, it is found that even in the case of a large optimization range, excellent performance can still be guaranteed. The following technical advantages are specifically realized:

[0025] (1) High-efficiency optimization and fast convergence: Based on the adaptive gradient adjustment mechanism of the AdaGrad algorithm, it can dynamically allocate the parameter update step size, significantly improving the optimization efficiency under the problem of multi-parameter coupling. Compared with the traditional gradient descent method, the convergence speed is increased by more than 40%, and it effectively avoids falling into local optimal solutions, especially suitable for the search in large-scale parameter spaces.

[0026] (2) Low threshold and high applicability: The automatic optimization process greatly reduces the dependence on the professional experience of designers. Non-professionals who have not been involved in this field can quickly generate feasible solutions through preset performance indicators; at the same time, professionals in this field can flexibly adjust the penalty factor (σ), tolerance parameters (Δn, ΔZ), constraint ranges (n2 - n1, Z2 - Z1), and weight coefficients (α, β, a, b, c, d) to adapt to different process platforms (such as lithium niobate thin films, lithium niobate bulk materials, silicon-based materials), electrode types (such as ordinary traveling-wave electrodes, capacitively loaded traveling-wave electrodes), and different process requirements (such as hollowing out the bottom, sio2 covering).

[0027] (3) Strong engineering practicability: By using the parameter projection technology, the continuous optimization results are mapped to discrete process nodes (such as the electrode width stepping by 0.1 μm), and the manufacturable layout parameters are directly output, avoiding the iterative process of "design - simulation - process adaptation" in the traditional method and reducing the trial - production cost. Brief Description of the Drawings

[0028] Figure 1 is the flowchart of the method of the embodiment of the present invention.

[0029] Figure 2 is the schematic diagram of the traveling - wave electrode structure of the electro - optic modulator involved in the embodiment of the present invention.

[0030] Figure 3 is the cross - sectional view of the traveling - wave electrode structure of the electro - optic modulator involved in the embodiment of the present invention.

[0031] Figure 4 is the schematic diagram of the iterative convergence of the Adagrad algorithm in the embodiment of the present invention.

[0032] Figure 5 is the page effect diagram of the embodiment of the present invention. Detailed Embodiment

[0033] In order to enable those skilled in the art to better understand the solution of the present invention, the following will describe in detail the embodiments of the present invention in combination with the drawings and embodiments, so as to fully understand how the present invention uses technical means to solve technical problems and the implementation process of achieving corresponding technical effects. The embodiments of the present invention and each feature in the embodiments can be combined with each other without conflict, and the formed technical solutions are all within the protection scope of the present invention.

[0034] Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work shall fall within the protection scope of the present invention.

[0035] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification and claims of the present invention and the above - mentioned drawings are intended to cover non - exclusive inclusion.

[0036] Such as Figure 1 shown, for the problem of low optimization efficiency caused by multi - parameter coupling, non - linear constraints and discrete process limitations in traditional designs, this embodiment discloses a design method of a traveling - wave electrode of an electro - optic modulator combining the external penalty function method and the Adagrad algorithm, and the process is as follows:

[0037] Step 1. According to the actual process design requirements, determine the structural type of the traveling-wave electrode of the electro-optic modulator. According to the structural type of the traveling-wave electrode of the electro-optic modulator, determine the process parameters to be optimized for the traveling-wave electrode, and clarify the key process parameters among the process parameters to be optimized.

[0038] Moreover, based on the electro-optic effect and optical waveguide characteristics of the traveling-wave electrode material of the electro-optic modulator, establish a physical model of the traveling-wave electrode of the electro-optic modulator, thereby clarifying the performance indicators of the traveling-wave electrode and the key performance indicators among the performance indicators.

[0039] Meanwhile, set the constraint conditions for the key process parameters and the constraint conditions for all performance indicators.

[0040] In this embodiment, two typical structural types of the traveling-wave electrode of the electro-optic modulator are taken as examples for illustration. The two typical structural types of the traveling-wave electrode are respectively a common traveling-wave electrode and a capacitive load type traveling-wave electrode. The structures of these two typical traveling-wave electrodes are the same, as Figure 2 、 Figure 3 shown. A signal electrode 2, a ground electrode 3, and a component 4 are arranged on a substrate 1. There are two ground electrodes 3, and the two ground electrodes 3 are respectively located outside the two symmetric sides of the signal electrode 2. There is an optical waveguide 4 between each ground electrode 3 on each side and the signal electrode 2.

[0041] Among the above two typical structural types of the traveling-wave electrode, the process parameters to be optimized include the width w of the signal electrode 2, the electrode spacing g (i.e., the spacing between the signal electrode 2 and each ground electrode 3 is g), the width wg of the ground electrode 3, the thickness h1 of the signal electrode 2, and the thickness h2 of the ground electrode 3. Among them, the width w of the signal electrode 2 and the electrode spacing g are the key process parameters.

[0042] When each process parameter to be optimized is determined, the traveling-wave electrode of the electro-optic modulator has certain physical properties. In this embodiment, physical models are respectively established for the above two typical structural types of the traveling-wave electrode, and the determined performance indicators include the microwave refractive index n, the characteristic impedance Z, the insertion loss S21, and the half-wave voltage V. And the microwave refractive index n and the characteristic impedance Z are selected as the initially selected performance indicators.

[0043] Moreover, in this embodiment, constraint conditions for key process parameters and performance indicators that are of great significance for evaluating the performance of the electro-optic modulator are also established. Among the key process parameters, the constraint condition for the width w of the signal electrode is: w1 < w < w2, where w1 is the minimum constraint for the width of the signal electrode and w2 is the maximum constraint for the width of the signal electrode. Among the key process parameters, the constraint condition for the electrode spacing g is: g1 < g < g2, where g1 is the minimum constraint for the electrode spacing and g2 is the maximum constraint for the electrode spacing. Among the performance indicators, the constraint condition for the microwave refractive index n is: n1 < n < n2, where n1 is the minimum constraint for the microwave refractive index and n2 is the maximum constraint for the microwave refractive index. Among the performance indicators, the constraint condition for the characteristic impedance Z is: Z1 < Z < Z2, where Z1 is the minimum constraint for the characteristic impedance and Z2 is the maximum constraint for the characteristic impedance. Among the performance indicators, the constraint condition for the half-wave voltage V is V < V0, where V0 is the maximum constraint for the half-wave voltage. Among the performance indicators, since the insertion loss S21 has little significance for evaluating the performance of the above two typical traveling-wave electrode electro-optic modulators, the constraint for the insertion loss S21 is not established.

[0044] Step 2: Obtain the initial values of multiple groups of process parameters to be optimized for the traveling-wave electrode.

[0045] Construct a calculation function for the evaluation result of the initial selected performance indicators of the traveling-wave electrode with the initially selected performance indicators as variables. The calculation function for the evaluation result of the initial selected performance indicators includes an initial selected performance indicator objective function and an initial selected performance indicator penalty term.

[0046] Based on the calculation function for the evaluation result of the initial selected performance indicators, calculate the evaluation result of the initial selected performance indicators of the traveling-wave electrode corresponding to the initial values of each group of process parameters to be optimized, as well as the initial selected performance indicator penalty term.

[0047] Then, based on the evaluation result of the initial selected performance indicators of the traveling-wave electrode corresponding to the initial values of each group of process parameters to be optimized, take the initial values of several groups of process parameters to be optimized with the evaluation result of the initial selected performance indicators less than the preset threshold as the initial conditions, where the preset threshold is determined based on the design experience of experts for different types of electro-optic modulators; or take the initial values of several groups of process parameters to be optimized with the penalty term being 0 as the initial conditions.

[0048] In this embodiment, the initial values of the process parameters to be optimized are obtained by performing an equally-spaced preliminary linear scan on the process parameters to be optimized at a relatively large interval to determine the approximate positions of the feasible solutions of the process parameters to be optimized. For the above two typical types of traveling-wave electrode structures, in this embodiment, the initial values of each set of process parameters to be optimized refer to the initial values of the signal electrode width w, the electrode pitch g, the ground electrode width wg, the signal electrode thickness h1, and the ground electrode thickness h2. Multiple sets of initial values of the signal electrode width w, the electrode pitch g, the ground electrode width wg, the signal electrode thickness h1, and the ground electrode thickness h2 are obtained through the preliminary linear scan.

[0049] For the above two typical types of traveling-wave electrode structures, the initially selected performance index evaluation result calculation function P(n, Z) constructed in this embodiment is as follows:

[0050] P(n, Z) = f(n, Z) + σp(n, Z)

[0051] In the formula: P(n, Z) is defined as the initially selected performance index evaluation result calculation function, which is constructed with the initially selected performance index microwave refractive index n and characteristic impedance Z as variables. The smaller the value of P(n, Z), the better the matching result.

[0052] f(n, Z) is the objective function, which is constructed with the initially selected performance index microwave refractive index n and characteristic impedance Z as variables.

[0053] p(n, Z) is the penalty term, which is constructed with the initially selected performance index microwave refractive index n and characteristic impedance Z as variables.

[0054] σ is a sufficiently large number, called the penalty factor.

[0055] In this embodiment, for the above two typical types of traveling-wave electrode structures, the constructed objective function f(n, Z) is as shown in the following formula:

[0056] min f(n, Z) = α * |n - no| + β * |Z - Zo|

[0057] The constructed penalty term p(n, Z) is as shown in the following formula:

[0058]

[0059] Where: α and β are constants. no is the group refractive index of the optical waveguide. Zo is the port characteristic impedance. η ≥ 1, and usually η = 2 can be taken. Δn is the microwave refractive index tolerance caused by the scan interval in the preliminary linear scan, and the range is 0.1 < Δn < 0.3. ΔZ is the characteristic impedance tolerance caused by the scan interval in the preliminary linear scan, and the range is 10Ω < ΔZ < 15Ω.

[0060] In this embodiment, based on the evaluation results of the initially selected performance indicators, the function P(n, Z) is calculated, and the evaluation results of the initially selected performance indicators (n, Z) of the traveling-wave electrodes corresponding to the initial values of each group of process parameters to be optimized are calculated, as well as the penalty term p(n, Z) for the initially selected performance indicators.

[0061] Finally, take the initial values of several groups of process parameters to be optimized whose evaluation results of the initially selected performance indicators are less than the preset threshold as the initial conditions, or take the initial values of several groups of process parameters to be optimized with a penalty term of 0 as the initial conditions.

[0062] Step 3: Taking all the performance indicators of the traveling-wave electrode as variables, establish the objective function for the design of the traveling-wave electrode, and use the constraint conditions of the key process parameters and the constraint conditions of the performance indicators that are of great significance to the performance evaluation of the electro-optic modulator as the constraints of the objective function for the design of the traveling-wave electrode.

[0063] In this embodiment, for the above two typical types of traveling-wave electrode structures, the established objective function g(n, Z, V, S21) for the design of the traveling-wave electrode takes the four performance indicators of the microwave refractive index n, characteristic impedance Z, insertion loss S21, and half-wave voltage V as variables. The constraints of the objective function for the design of the traveling-wave electrode include the constraint conditions of the performance indicators n1 < n < n2, Z1 < Z < Z2, V < V0, and the constraint conditions of the key process parameters w1 < w < w2, g1 < g < g2.

[0064] Step 4: Adopt the exterior penalty function method to transform the optimization problem of the objective function for the design of the traveling-wave electrode with constraints established in Step 3 into an unconstrained optimization problem.

[0065] In this embodiment, for the above two typical types of traveling-wave electrode structures, the transformed unconstrained optimization problem is shown as the following formula:

[0066] Q(n, Z, V, S21, w, g) = g(n, Z, V, S21) + σq(n, Z, V, w, g)

[0067] Where: Q(n, Z, V, S21, w, g) represents the function of the transformed unconstrained optimization problem. g(n, Z, V, S21) represents the objective function for the design of the traveling-wave electrode constructed in Step 3. q(n, Z, V, w, g) represents the penalty function for the constraints of the objective function for the design of the traveling-wave electrode. σ is a sufficiently large number, called the penalty factor.

[0068] Specifically, the objective function g(n, Z, V, S21) for the design of the traveling-wave electrode is shown as the following formula:

[0069] g(n, Z, V, S21) = a * |n - no| + b * |Z - Zo| + c * |V| + d * |S21|

[0070] Where: a, b, c, d are constants; n is the microwave refractive index; no is the group refractive index of the optical waveguide; Z is the characteristic impedance; Zo is the port characteristic impedance; V is the half-wave voltage; S21 is the insertion loss.

[0071] The penalty function q(n, Z, V, w, g) for the objective function constraint of the traveling-wave electrode design is shown as follows:

[0072] q(n, Z, V, w, g) = s(n, Z, V) + t(w, g)

[0073] Where: s(n, Z, V) represents the constraint conditions of the performance indicators that are important for evaluating the performance of the electro-optic modulator; t(w, g) represents the constraint conditions of the key process parameters.

[0074] The constraint conditions s(n, Z, V) of the performance indicators and the constraint conditions t(w, g) of the key process parameters are shown as follows:

[0075] s(n, Z, V) = |min{n - n1, 0}| η + |min{n2 - n, 0}| η + |min{Z - Z1, 0}| η + |min{Z2 - Z, 0}| η + |min{V0 - V, 0}| η

[0076] t(w, g) = |min{w - w1, 0}| η + |min{w2 - w, 0}| η + |min{g - g1, 0}| η + |min{g2 - g, 0}| η .

[0077] Step 5: Use the AdaGrad algorithm, with the process parameters to be optimized as variables, perform gradient iteration calculations on the unconstrained optimization problem; each time during iteration, use the initial value of one set of the process parameters to be optimized in the initial conditions as the variable initial point for calculation, thereby solving for the optimal solution of the variable and the iteration result of the unconstrained optimization problem corresponding to the optimal solution of the variable for that iteration; during multiple iterations, use the initial values of different sets of the process parameters to be optimized in the initial conditions as the variable initial points, thereby obtaining the iteration results of the unconstrained optimization problems corresponding to the optimal solutions of the variables for each iteration through multiple iterations; then based on the iteration results, select the optimal solution of the variable corresponding to the iteration result that meets the requirements as the process parameter value of the traveling-wave electrode that meets the design requirements.

[0078] In this embodiment, the AdaGrad algorithm is used to perform gradient iteration and solution on the function of the unconstrained optimization problem established in Step 4, as Figure 4As shown below, the process is as follows:

[0079] (S1) Determine the initial search step size according to the determined search interval range.

[0080] (S2) According to the determined search interval and search step size, with the process parameters to be optimized as variables, use the AdaGrad algorithm to iteratively solve the unconstrained optimization problem. Each time during iteration, take the initial value of one set of the process parameters to be optimized in the initial conditions obtained in step 2 as the variable initial point for calculation, and obtain the optimal solution of the variable and the corresponding iterative result of the unconstrained optimization problem. During multiple iterations, take the initial values of different sets of the process parameters to be optimized in the initial conditions obtained in step 2 as the variable initial points for calculation respectively, and thus obtain the optimal solutions of the variables corresponding to each iteration and the corresponding iterative results of the unconstrained optimization problem. Each iterative solution is as shown in the following formula:

[0081]

[0082] In the formula: represents the function value of the unconstrained optimization problem under the k-th iteration, that is, the k-th iteration result.

[0083] represents the variable at the k-th iteration. The initial point of the variable is the initial value of one set of the process parameters to be optimized in the initial conditions. Each time during the iteration process, solve with x as the variable to obtain the optimal solution of the variable x for the current iteration.

[0084] represents the penalty factor at the k-th iteration.

[0085] .

[0086] represents a function with x as the independent variable and the function value of the traveling-wave electrode design objective function g(n, Z, V, S21) as the dependent variable.

[0087] is the penalty term.

[0088] By using the Adagrad algorithm for iterative solution, obtain the optimal solution in the case where the penalty factor is . The optimal solution of the variable obtained by the k-th outer penalty iteration solution , where represents the optimal solution at the k-th iteration is related to the penalty factor, so it is expressed as a function with the penalty factor as the variable, that is, 。

[0089] (S3) When , the algorithm terminates, and the point is the approximate optimal solution of the constrained optimization problem, where represents the initial point of the k-th iteration, represents the initial point of the k-th iteration corresponding to the obtained penalty term; otherwise, update the penalty factor and repeat the iteration. Among them, is the magnification factor, which is taken as = 10 in this embodiment; is the allowable error, which is taken as = 0.05 in this embodiment.

[0090] Each iteration in this embodiment actually includes two layers of iteration, namely the optimization iteration of the Adagrad algorithm in step S2 and the iteration of the outer penalty (i.e., the penalty factor) in step S3.

[0091] After each gradient update of the Adagrad algorithm, the parameters are projected onto the nearest discrete point to meet the discreteness characteristics in the engineering optimization problem.

[0092] (S4) Sort the iteration results corresponding to the variable optimal solutions obtained in each iteration and take the variable optimal solution corresponding to the smallest one of the iteration results as the traveling-wave electrode process parameter values that meet the design requirements, that is, the process parameter values of the signal electrode width w, electrode spacing g, ground electrode width wg, signal electrode thickness h1, and ground electrode thickness h2 that meet the design requirements, thereby completing the design work of the traveling-wave electrode of the electro-optic modulator.

[0093] This embodiment also discloses an electronic device, including a processor and a memory. When the processor reads and runs the program instructions stored in the storage medium of the memory, it executes the above-mentioned traveling-wave electrode design method of the electro-optic modulator.

[0094] This embodiment also discloses a storage medium. When the program instructions in the storage medium are read and run by any device, the above-mentioned traveling-wave electrode design method of the electro-optic modulator is executed.

[0095] The method of this embodiment has been verified through actual tests, and an electro-optic modulator physical object that meets the performance indicators has been successfully designed and manufactured, which can realize the full-process automated design from model establishment to parameter optimization. For the specific page effect, please refer to Figure 5 as shown.

[0096] The preferred embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings. The embodiments described in the present invention are merely descriptions of the preferred embodiments of the present invention, and do not limit the concept and scope of the present invention. Among the various specific technical features described in the above specific embodiments, they can be combined in any suitable manner without contradiction. As long as such a combination does not violate the idea of the present invention, it should also be regarded as the content disclosed in this disclosure. To avoid unnecessary repetition, the present invention will not separately describe various possible combination methods.

[0097] The present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention and without departing from the design idea of the present invention, various variations and improvements made by those skilled in the art to the technical solution of the present invention should fall within the protection scope of the present invention. The technical content claimed by the present invention has been fully recorded in the claims.

Claims

1. A method for designing a traveling wave electrode of an electro-optic modulator, characterized in that: The process is as follows: Determine the process parameters to be optimized and the key process parameters therein, determine the performance indicators and select the initial performance indicators therefrom, and set the constraints of the key process parameters and performance indicators; Based on the preliminary performance index evaluation results corresponding to each of the multiple groups of initial values ​​of the process parameters to be optimized, several groups of initial values ​​of the process parameters to be optimized are selected as initial conditions; The objective function of the traveling wave electrode design is established with the performance index as a variable, and the constraints of the key process parameters and performance indexes are used as constraints of the objective function; Convert the constrained objective function optimization problem into an unconstrained optimization problem; The AdaGrad algorithm is used to iterate the unconstrained optimization problem with the process parameters to be optimized as variables; In each iteration, a set of initial values ​​of the process parameters to be optimized in the initial conditions are used as the initial points of the variables, and the optimal solutions of the variables and the corresponding iterative results of the unconstrained optimization problem are solved; in multiple iterations, different groups of initial values ​​are used as the initial points of the variables, and the optimal solutions of the variables and the corresponding iterative results of each iteration are obtained; the optimal solutions of the variables corresponding to the iterative results that meet the requirements are selected as the traveling wave electrode process parameter values ​​that meet the design requirements.

2. The method for designing a traveling wave electrode of an electro-optic modulator according to claim 1, characterized in that: The structural types of the traveling wave electrode include ordinary traveling wave electrode and capacitive load type traveling wave electrode.

3. The method for designing a traveling wave electrode of an electro-optic modulator according to claim 2, characterized in that: The process parameters to be optimized include signal electrode width, electrode spacing, ground electrode width, signal electrode thickness, and ground electrode thickness, among which the key process parameters are signal electrode width and electrode spacing.

4. The method for designing a traveling wave electrode of an electro-optic modulator according to claim 2, characterized in that: The performance indicators include microwave refractive index, characteristic impedance, insertion loss, and half-wave voltage, and the microwave refractive index and characteristic impedance are selected as the primary performance indicators.

5. The method for designing a traveling wave electrode of an electro-optic modulator according to claim 1, characterized in that: Using the preliminary selected performance indicator as a variable to construct a preliminary selected performance indicator evaluation result calculation function, the preliminary selected performance indicator evaluation result calculation function includes a preliminary selected performance indicator objective function and a preliminary selected performance indicator penalty item; Based on the preliminary performance index evaluation result calculation function, the preliminary performance index evaluation results and the preliminary performance index penalty items of each group of initial values ​​of the process parameters to be optimized are calculated; When selecting, the initial values ​​of several groups of process parameters to be optimized whose initial performance index evaluation results are less than the preset threshold are taken as initial conditions, or the initial values ​​of several groups of process parameters to be optimized whose penalty items are 0 are taken as initial conditions.

6. The method for designing a traveling wave electrode of an electro-optic modulator according to claim 1, characterized in that: The iterative results of the unconstrained optimization problem corresponding to the optimal solutions of the variables obtained in each iteration are sorted by size, and the optimal solution of the variables corresponding to the smallest iterative result is taken as the traveling wave electrode process parameter value that meets the design requirements.

7. A method for designing a traveling wave electrode of an electro-optic modulator according to any one of claims 1 to 6, characterized in that: The Adagrad algorithm projects the parameters to the nearest discrete point after each gradient update.

8. An electronic device comprising a processor and a memory, characterized in that: When the processor reads and runs the program instructions in the memory, it executes the method for designing a traveling wave electrode of an electro-optic modulator as claimed in any one of claims 1 to 7.

9. A storage medium storing program instructions, characterized in that: When the program instructions are read and executed, the method for designing traveling wave electrodes of an electro-optic modulator as claimed in any one of claims 1 to 7 is executed.

Images