Design optimization method and device of sisl device, server and storage medium

Abstract

The application discloses a kind of SISL device design optimization method, device, server and storage medium, belong to device optimization technical field.It includes: SISL complete structure is decomposed into simple structure and transition structure, and simple structure and transition structure are respectively regarded as simple structure fine model and transition structure fine model;First equivalent circuit with the same circuit structure of simple structure is constructed, and the first equivalent circuit is regarded as simple structure coarse model;Second equivalent circuit with the same circuit structure of transition structure is constructed, and the second equivalent circuit is regarded as transition structure coarse model;Based on the fine model and coarse model of simple structure and transition structure, the optimal solution of simple structure circuit size and the preliminary solution of transition structure circuit size are obtained;Based on the optimal solution, the preliminary solution is optimized to obtain the optimal solution of transition structure circuit size.Through dimensionality reduction decomposition and replacement model auxiliary, the efficient and accurate design of SISL device is realized.

Description

Technical Field

[0001] This invention relates to the field of device optimization technology, and in particular to a design optimization method, apparatus, server, and storage medium for SISL devices. Background Technology

[0002] Substrate Integrated Suspended Line (SISL) is a transmission structure based on alternating layers of dielectrics and metals. Its core advantages lie in its high integration and self-encapsulation, along with low dielectric loss and strong design flexibility. Leveraging these advantages, SISL has demonstrated broad application potential in several key areas, including microwave and millimeter-wave circuits, satellite communications, radar systems, and high-speed digital circuits, becoming one of the important research directions in the current high-frequency transmission field.

[0003] Currently, in the design process of SISL devices, the High Frequency Structure Simulator (HFSS) is a core technology widely relied upon in the industry. The core value of HFSS lies in its ability to provide high-fidelity simulation results, accurately simulating the electromagnetic characteristics of SISL transmission structures. This provides data support for device performance verification and preliminary determination of structural parameters, and is the technical foundation for ensuring the accuracy of SISL device design.

[0004] Despite the high fidelity of HFSS, it has significant technical drawbacks, including extremely high computational resource consumption, long simulation time, and the need for frequent large-scale numerical calculations, which restricts the speed of SISL device design iteration. Summary of the Invention

[0005] This invention provides a design optimization method, apparatus, server, and storage medium for SISL devices to solve the technical problems of high computational resource consumption and long simulation time caused by existing technologies that rely on HFSS full-wave electromagnetic simulation.

[0006] In a first aspect, embodiments of the present invention provide a design optimization method for SISL devices, including:

[0007] Based on the decomposition and coupling formula of SISL devices, the complete SISL structure is decomposed into a simplified structure and a transitional structure, and the simplified structure and the transitional structure are respectively used as the simplified structure fine model and the transitional structure fine model.

[0008] Construct a first equivalent circuit with the same circuit structure as the simplified structure, and use the first equivalent circuit as a coarse model of the simplified structure.

[0009] constructing a second equivalent circuit with the same circuit structure as the transition structure, and taking the second equivalent circuit as a transition structure coarse model;

[0010] obtaining the optimal solution of the simple structure circuit size based on the simple structure fine model and the simple structure coarse model;

[0011] obtaining the preliminary solution of the transition structure circuit size based on the transition structure fine model and the transition structure coarse model;

[0012] constructing a transition single-line structure based on the simple structure circuit, the optimal solution of the simple structure circuit size and the preliminary solution of the transition structure circuit size;

[0013] constructing a transition single-line structure substitute model based on the transition single-line structure and a space mapping principle, adjusting the circuit size of the transition single-line structure with the circuit size of the transition single-line structure as an initial value, and determining the circuit size of the transition single-line structure as the optimal solution of the transition structure circuit size when an error between an output response of the transition single-line structure substitute model and an output response of the simple single-line structure converges to a tolerance range.

[0014] In a second aspect, an embodiment of the present application further provides a design optimization device of an SISL device, which comprises:

[0015] a decomposition module configured to decompose an SISL complete structure into a simple structure and a transition structure based on an SISL decomposition coupling formula, and take the simple structure and the transition structure as a simple structure fine model and a transition structure fine model respectively;

[0016] a first equivalent circuit construction module configured to construct a first equivalent circuit with the same circuit structure as the simple structure, and take the first equivalent circuit as a simple structure coarse model;

[0017] a second equivalent circuit construction module configured to construct a second equivalent circuit with the same circuit structure as the transition structure, and take the second equivalent circuit as a transition structure coarse model;

[0018] an optimal solution acquisition module configured to obtain the optimal solution of the simple structure circuit size based on the simple structure fine model and the simple structure coarse model;

[0019] a preliminary solution acquisition module configured to obtain the preliminary solution of the transition structure circuit size based on the transition structure fine model and the transition structure coarse model;

[0020] a construction module configured to construct a transition single-line structure based on the simple structure circuit, the optimal solution of the simple structure circuit size and the preliminary solution of the transition structure circuit size;

[0021] An optimization module is configured to construct a transition single-line structure substitute model based on a transition single-line structure and a space mapping principle; adjust circuit sizes of the transition single-line structure with the circuit sizes of the transition single-line structure as initial values; and determine the circuit sizes of the transition single-line structure as an optimal solution of the transition structure circuit sizes when an error between an output response of the transition single-line structure substitute model and an output response of the simple single-line structure converges to a tolerance range.

[0022] In a third aspect, an embodiment of the present application further provides a server, comprising:

[0023] one or more processors;

[0024] a storage device configured to store one or more programs;

[0025] When the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the design optimization method of the SISL device provided by the above embodiment.

[0026] In a fourth aspect, an embodiment of the present application further provides a storage medium containing computer executable instructions, which, when executed by a computer processor, are configured to perform the design optimization method of the SISL device provided by the above embodiment.

[0027] The SISL device design optimization method, device, server and storage medium provided by the embodiment of the present application are based on a SISL device decoupling formula, the SISL complete structure is decomposed into a simple structure and a transition structure, and the simple structure and the transition structure are respectively taken as a simple structure fine model and a transition structure fine model; a first equivalent circuit with the same circuit structure as the simple structure is constructed, and the first equivalent circuit is taken as a simple structure coarse model; a second equivalent circuit with the same circuit structure as the transition structure is constructed, and the second equivalent circuit is taken as a transition structure coarse model; the optimal solution of the simple structure circuit size is obtained based on the simple structure fine model and the simple structure coarse model; the preliminary solution of the transition structure circuit size is obtained based on the transition structure fine model and the transition structure coarse model; the transition single-line structure is constructed according to the optimal solution of the simple structure circuit size and the preliminary solution of the transition structure circuit size based on the circuit of the simple structure; the transition single-line structure substitution model is constructed based on the transition single-line structure and the space mapping principle; the circuit size of the transition single-line structure is adjusted with the circuit size in the transition single-line structure as the initial value; when the error between the output response of the transition single-line structure substitution model and the output response of the simple single-line structure converges to the tolerance range, the circuit size of the transition single-line structure is determined as the optimal solution of the transition structure circuit size. Through the decomposition of the SISL complete structure, the complex optimization problem requiring full-wave simulation is converted into independent and collaborative optimization of the simple structure and the transition structure, the calculation resources are concentrated on the fine adjustment of the key part, and therefore the efficient and accurate design of the SISL device is realized, and the calculation bottleneck of the full-wave simulation is effectively avoided. BRIEF DESCRIPTION OF DRAWINGS

[0028] The accompanying drawings, which form a part of the present application, are used to provide further understanding of the present application, and serve as an aid in explaining the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application, and do not constitute improper limitations on the present application.

[0029] Figure 1 is a flowchart of the SISL device design optimization method provided by the first embodiment of the present application;

[0030] Figure 2 is a flowchart of obtaining the optimal solution of the simple structure circuit size of the SISL device design optimization method provided by the first embodiment of the present application;

[0031] Figure 3 is a flowchart of obtaining the preliminary solution of the transition structure circuit size of the SISL device design optimization method provided by the first embodiment of the present application;

[0032] Figure 4 is a flowchart of constructing the transition single-line structure of the SISL device design optimization method provided by the first embodiment of the present application;

[0033] Figure 5 is a structural diagram of a design optimization device of a SISL device provided by Embodiment Two of the present application;

[0034] Figure 6 is a structural diagram of a server provided by Embodiment Three of the present application. DETAILED DESCRIPTION

[0035] The present application will be further described below in conjunction with the accompanying drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the present application, but not to limit the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the present application are shown in the drawings, but not all the structures.

[0036] Embodiment One

[0037] Figure 1 is a flow chart of a design optimization method of a SISL device provided by Embodiment One of the present application. The present embodiment can be applicable to solve the scenario that the SISL device design iteration is slow due to the dependence on full-wave electromagnetic simulation, and specifically includes the following steps:

[0038] In step 110, based on the SISL device decomposition coupling formula, the SISL complete structure is decomposed into a simple structure and a transition structure, and the simple structure and the transition structure are respectively taken as a simple structure fine model and a transition structure fine model.

[0039] Substrate Integrated Suspended Line (SISL) is a new type of high-performance planar transmission line technology. Its core structure is to form an air cavity by hollowing out in a multi-layer dielectric plate, and to suspend the metal line of the main signal in the center of the air cavity. This unique design combines the low loss and high Q value advantages of traditional non-planar suspended strip line with the ease of integration and compact size of planar circuits. Since the signal line is mainly surrounded by low-loss air medium, the SISL structure can significantly reduce the conduction loss and dielectric loss, and at the same time, it naturally forms a shield with the upper and lower metal layers, and has excellent self-packaging characteristics and anti-interference ability. Therefore, SISL technology is particularly suitable for microwave and millimeter wave circuits, satellite communication front ends, radar modules, and high-speed digital signal transmission fields with stringent requirements on performance, integration, and cost, and has become one of the important technical paths to realize small-sized and high-performance electronic systems. However, to fully tap and optimize the performance potential of SISL devices, a modeling and design method that can balance accuracy and efficiency is needed. Therefore, optionally, the present embodiment introduces the SISL device decomposition coupling formula as the core theoretical tool.

[0040] For example, the SISL device decomposition coupling formula can be represented by the following formula: wherein, is the response matrix of the SISL complete structure, is the response matrix of the simple structure, is the response matrix of the transition structure, is the coupling matrix, is the boundary field coupling correction matrix, is the circuit size parameter of the simple structure, is the circuit size parameter of the transition structure, is the circuit size parameter of the SISL complete structure, and The principle of the SISL device decomposition coupling formula is that a complex SISL complete structure, whose overall electromagnetic response is not an indivisible black box, but can be composed of a simple structure and a transition structure, the responses of the simple structure and the transition structure are combined through a determined coupling relationship , and a correction term is added to compensate for the ignored boundary coupling effect.

[0041] Exemplarily, the decomposition of the SISL complete structure into the simple structure and the transition structure is based on the division strategy of the physical composition and functional characteristics. The simple structure corresponds to the intermediate dielectric substrate layer circuit in the SISL device that realizes the core circuit function, constitutes the main transmission part of the device, and this part is usually the main and regular part that realizes the core function of the circuit. The transition structure covers the upper and lower dielectric substrate layer circuits and the metal via connecting the layers, and mainly undertakes auxiliary functions such as vertical interconnection, electromagnetic shielding and field mode conversion. This division method decouples the complex complete three-dimensional structure into sub-units with clear functions, which are easy to analyze and optimize independently, and creates conditions for the subsequent collaborative design of accurate simulation and rapid optimization.

[0042] Exemplarily, in the design process of the SISL device, the model can refer to a calculation object used to predict what response the SISL device structure will produce under electromagnetic excitation. Among them, the fine model refers to a high-precision and high-fidelity model established by full-wave electromagnetic simulation software. It can extremely accurately simulate the real electromagnetic behavior of the structure, but the disadvantage is that the calculation cost is high and the time is extremely long. The simple structure fine model refers to the extraction of the core function part after the dimensionality reduction decomposition of the SISL complete structure. The input of the simple structure fine model is the geometric parameter , and the output is the electromagnetic response . The transition structure fine model refers to the remaining part of the complete structure except the simple structure, mainly including the upper and lower dielectric layers, metal vias and other three-dimensional structures for interconnection and packaging. The input of the transition structure fine model is the geometric parameter , and the output is the electromagnetic response .

[0043] Step 120: Construct a first equivalent circuit with the same circuit structure as the simplified structure, and use the first equivalent circuit as a rough model of the simplified structure.

[0044] Microstrip lines are a classic and widely used planar transmission line structure. Their basic structure consists of a conductor strip laid on the surface of a dielectric substrate and a metallic ground plane covering the entire bottom of the substrate. Signals propagate between the conductor strip and the ground plane, with their electromagnetic energy primarily confined to the dielectric and air near the strip.

[0045] For example, the simplified equivalent circuit structure refers to the first equivalent circuit constructed using microstrip lines in circuit simulation software (such as ADS). All key physical dimensions that determine the circuit performance of the first equivalent circuit (such as the linewidth, length, conductor gaps, spacing and length of coupling segments, etc.) must be completely consistent with the core circuit pattern dimensions in the established simplified structure fine model.

[0046] For example, in the design process of SISL devices, a coarse model refers to a fast but approximate equivalent circuit model built using circuit simulation software (such as ADS). It can qualitatively or semi-quantitatively simulate the circuit behavior of the structure. Its core advantage lies in its extremely fast computation speed, making it suitable for the initial optimization stage that requires a large number of iterations, but its accuracy is lower than that of full-wave electromagnetic simulation. A coarse model for a simplified structure refers to a fast but approximate first equivalent circuit model based on a circuit structure consistent with the simplified structure.

[0047] Step 130: Construct a second equivalent circuit with the same circuit structure as the transition structure, and use the second equivalent circuit as a coarse model of the transition structure.

[0048] For example, the transition structure identical circuit structure can refer to a second equivalent circuit constructed using microstrip lines in circuit simulation software (such as ADS). All critical physical dimensions of the second equivalent circuit that determine circuit performance must be identical to the core circuit pattern dimensions in the established fine-scale transition structure model. A coarse-scale transition structure model refers to a fast but approximate second equivalent circuit model based on the same transition structure circuit structure.

[0049] Step 140: Based on the simplified structure fine model and the simplified structure coarse model, obtain the optimal solution for the simplified structure circuit dimensions.

[0050] Figure 2 This is a flowchart illustrating the process of obtaining the optimal solution for the simplified structure circuit size using the design optimization method for SISL devices described in Embodiment 1 of the present invention. Figure 2 The process of obtaining the optimal solution for the dimensions of a simplified circuit structure may include the following steps:

[0051] Step 210: Set the sampling range.

[0052] Setting a sampling range defines clear parameter boundaries for the entire optimization process. Optionally, the sampling range is a pre-determined reasonable variation range of various key geometric parameters of a simplified structure, based on design experience and performance indicators. This range should not be too narrow, lest potential optimal solutions be missed, nor too wide, leading to low sampling efficiency or even unrealistic physical structures. Therefore, precise sampling range setting is the foundation for the effective implementation of all subsequent sampling, modeling, and optimization operations. It ensures that the optimization process takes place within a physically realizable and clearly defined efficient space, directly determining the quality of the final design results and the computational efficiency of the entire process.

[0053] Step 220: Within the sampling range, sample the simplified structural fine model to obtain sample data of the simplified structural fine model.

[0054] For example, sampling is performed within a pre-defined sampling range, and a series of representative combinations of geometric parameters are selected according to a specific experimental design method. For each sampling point, a full-wave electromagnetic simulation using HFSS is performed to obtain high-fidelity sample data for the simplified structural model, and its accurate electromagnetic response is calculated and recorded. Each parameter and response pair ( , This constitutes a set of sample data. This sample data realistically reflects the electromagnetic behavior of the real physical structure in key regions, providing reliable real data for subsequent calibration and construction of accurate alternative models using the principle of spatial mapping.

[0055] Optionally, the sampling operation can be orthogonal sampling, an efficient experimental design method. Its core idea is to use clever mathematical design (such as orthogonal arrays) in a multidimensional parameter space to ensure that each factor's different levels appear exactly once in all combinations of all other factor levels. This method ensures that sampling points are evenly distributed across all dimensions of the parameters to be optimized, thus exploring the entire design space unbiasedly and efficiently with minimal sampling. Compared to simple grid scanning or random sampling, it avoids data redundancy and significantly reduces the costly number of simulations required to build accurate alternative models.

[0056] Step 230: Within the sampling range, sample the simplified coarse model of the structure to obtain sample data of the simplified coarse model of the structure.

[0057] For example, the sampling of the coarse model of the simplified structure and the sampling of the fine model of the simplified structure in step 220 can use the same experimental design method and parameter sampling points. However, the core difference is that for each identical parameter combination, circuit simulation is performed using the coarse model (equivalent circuit) of the simplified structure established in ADS to quickly obtain its approximate electromagnetic response. Each parameter-response pair constitutes a set of sample data for the coarse model of the simplified structure. Although these sample data have lower accuracy, they improve efficiency and provide a data foundation for establishing the correlation between the fine and coarse models of the simplified structure in the subsequent spatial mapping process.

[0058] Step 240: Using the principle of spatial mapping, construct a simplified structure alternative model based on the sample data of the simplified structure fine model and the sample data of the simplified structure coarse model.

[0059] Spatial mapping is an efficient modeling and optimization technique. For example, by establishing and utilizing the mapping relationship between a high-precision but time-consuming simple structural fine model and a fast but low-precision simple structural coarse model, the fast computation capability of the coarse model and the high-precision advantage of the fine model can be combined.

[0060] For example, step 240 can be implemented as follows: First, a mapping network for learning the behavior of the coarse model is constructed. Based on the sampled data of the simplified coarse model, a multilayer perceptron (MLP) is developed to learn the input-output mapping relationship of the simplified coarse model. The MLP is a classic feedforward artificial neural network, consisting of a multilayer neuron network structure with an input layer, one or more hidden layers, and an output layer. Its core characteristic is that the neurons in the hidden layers use nonlinear activation functions. Through multi-layer nonlinear transformations, the MLP can approximate any complex nonlinear mapping relationship with arbitrary precision, thereby accurately learning the complex functional relationship from geometric parameters to circuit response. Based on this, the trained MLP is cascaded with the nonlinear input mapping network to form a complete spatial mapping model architecture. The nonlinear input mapping is responsible for correcting the input parameters and establishing a nonlinear transformation relationship between the parameter space of the simplified coarse model and the parameter space of the simplified fine model, thereby compensating for the differences between the two models. Finally, the model is trained and validated. High-fidelity sample data obtained from the simplified fine model is used to train the cascaded spatial mapping model end-to-end. By employing optimization algorithms such as backpropagation, the parameters of the input mapping network and the MLP network are adjusted synchronously, allowing the model's predicted values ​​to continuously approximate the actual response of the fine model. When the model's prediction error on the test set reaches the preset accuracy requirement (e.g., prediction error ≤ 2%), it indicates that the spatial mapping model has successfully captured the behavioral characteristics of the fine model and can serve as a simple structural replacement model that combines high accuracy and high efficiency, providing a reliable computational foundation for subsequent rapid optimization design.

[0061] Step 250: Adjust the circuit dimensions of the simplified structural alternative model.

[0062] Step 260: When the output response of the simplified structure replacement model meets the SISL device design specifications, the circuit size of the simplified structure replacement model is determined as the optimal solution for the simplified structure circuit size.

[0063] SISL device design metrics are quantitative standards set during the optimization process to determine whether the design results meet the final performance requirements. They ensure that the optimization algorithm always searches with the practical application goals of the device in mind and automatically terminates the iteration when the design performance meets the targets. Optionally, SISL device design metrics are specific performance criteria set for the design goals, which may include the following common requirements: first, frequency characteristic metrics, i.e., the target center frequency, operating bandwidth, or specific frequency band range that the device needs to achieve; second, loss and reflection metrics, i.e., the insertion loss and return loss of the device within the operating frequency band must reach preset thresholds; and third, selectivity or suppression metrics, i.e., the suppression level or isolation that the device needs to achieve in a specific out-of-band frequency band.

[0064] For example, optimization algorithms can be used to adjust the circuit dimensions of the simplified structure replacement model, try different size combinations on the simplified structure replacement model, and obtain its predicted performance to find the optimal solution. When a set of circuit dimensions is input into the simplified structure replacement model, and its output response (i.e., the predicted performance) meets the design specifications of the SISL device, the optimization loop will terminate. At this point, this set of circuit dimensions is determined as the optimal solution for the simplified structure circuit dimensions. The optimal solution for the simplified structure circuit dimensions can refer to the key geometric parameters of the core circuit determined through optimization. For example, this includes the precise width and length of the center conductor, and the optimal spacing between adjacent coupling lines.

[0065] Through the above steps, using the principle of spatial mapping, based on the simplified fine model and the simplified coarse model, sample data of the simplified fine model and the simplified coarse model are obtained, and finally the optimal solution of the simplified structure circuit size is obtained.

[0066] Step 150: Based on the fine model and coarse model of the transition structure, obtain a preliminary solution for the circuit dimensions of the transition structure.

[0067] Figure 3 This is a flowchart illustrating the preliminary solution for the transition structure circuit dimensions in the design optimization method for SISL devices described in Embodiment 1 of the present invention. Figure 3 The process of obtaining a preliminary solution for the dimensions of the transition structure circuit may include the following steps:

[0068] Step 310: Set the sampling range.

[0069] Step 320: Within the sampling range, sample the fine model of the transition structure to obtain sample data of the fine model of the transition structure.

[0070] Optionally, the sampling method described in step 220 can be used to select a series of representative combinations of geometric parameters. For each sampling point, a full-wave electromagnetic simulation using HFSS is performed to obtain high-fidelity sample data for the detailed model of the transition structure, and its precise electromagnetic response is calculated and recorded. Each parameter and response pair ( , This constitutes a set of sample data.

[0071] Step 330: Within the sampling range, sample the coarse model of the transition structure to obtain sample data of the coarse model of the transition structure.

[0072] For example, sampling the coarse model of the transition structure and sampling the fine model of the transition structure in step 320 can use the same experimental design method and parameter sampling points. However, the core difference is that for each identical parameter combination, circuit simulation is performed using the coarse model (equivalent circuit) of the transition structure established in ADS to quickly obtain its approximate electromagnetic response. Each parameter-response pair constitutes a set of sample data for the coarse model of the transition structure. Although these sample data have lower accuracy, their efficiency is improved compared to the fine model of the transition structure, providing a data foundation for establishing the correlation between the fine and coarse models of the transition structure in the subsequent spatial mapping process.

[0073] Step 340: Using the principle of spatial mapping, construct a transition structure alternative model based on the sample data of the fine transition structure model and the sample data of the coarse transition structure model.

[0074] For example, step 340 can be implemented in the same way as step 240. First, a mapping network is constructed to learn the behavior of the coarse transition structure model. Based on the sampled coarse transition structure model data, an MLP is developed to learn the input-output mapping relationship of the coarse transition structure model. Second, the trained MLP is cascaded with a nonlinear input mapping network to form a complete spatial mapping model architecture. Finally, the model is trained and validated using sample data obtained from the fine transition structure model. By optimizing the algorithm, the parameters of the input mapping network and the MLP network are adjusted synchronously so that the model's predicted values ​​continuously approach the actual response of the fine transition structure model. When the model's prediction error on the test set reaches the preset accuracy requirement, the spatial mapping model is used as the transition structure replacement model.

[0075] Step 350: Adjust the circuit dimensions of the transition structure replacement model.

[0076] Step 360: When the output response of the transition structure substitution model satisfies the preset convergence condition, the circuit size of the transition structure substitution model is determined as the preliminary solution of the transition structure circuit size.

[0077] Preset convergence criteria are quantitative standards set for the optimization process to determine whether the search can be terminated. They ensure that the optimization algorithm does not run indefinitely and automatically stops when a sufficiently good solution is found. Optionally, preset convergence criteria are quantitative termination criteria set for the optimization process, which can include the following judgment criteria: First, performance target tolerance, i.e., the error between the circuit performance predicted by the alternative model (such as S-parameters, center frequency, bandwidth, etc.) and the desired design target is less than a preset tolerance value. Second, parameter change, i.e., in consecutive iterations, the adjustment range of the circuit size has fallen below a set threshold, indicating that the structure tends to stabilize. Third, objective function improvement, i.e., the improvement range of the cost function or objective function used to evaluate circuit performance has significantly decreased during the iteration process, reflecting that the optimization process is gradually approaching its limit.

[0078] For example, optimization algorithms can be used to adjust the circuit dimensions of the transition structure replacement model, try different size combinations on the transition structure replacement model, and obtain its predictive performance to find the optimal solution. When a set of circuit dimensions is input into the transition structure replacement model, and its output response (i.e., the predicted performance) meets the preset convergence condition, the optimization loop will terminate. At this point, this set of circuit dimensions is determined as the preliminary solution for the transition structure circuit dimensions. The preliminary solution for the transition structure circuit dimensions can be the initial geometric parameters set to achieve electromagnetic coupling and impedance matching. For example, these include the initial diameter and array spacing of metal vias, the initial dimensions of grounding cavities in the upper and lower dielectric substrates, etc.

[0079] Through the above steps, using the principle of spatial mapping, based on the fine and coarse models of the transition structure, sample data of the fine and coarse models of the transition structure are obtained, and finally, a preliminary solution for the circuit dimensions of the transition structure is obtained.

[0080] Step 160: Based on the simplified circuit structure, construct a transitional single-line structure according to the optimal solution of the simplified circuit size and the preliminary solution of the transitional circuit size.

[0081] Although the previous steps have yielded independent solutions for the simplified and transitional structures, complex electromagnetic coupling effects occur at the interface when these structures are recombined into a complete structure. This coupling distorts the electromagnetic field distribution, preventing the overall performance from being the sum of the performances of the two independent components, often resulting in performance degradation (such as impedance mismatch, resonant frequency shift, and increased insertion loss). Simply piecing together the optimization results may not meet the final design specifications. Therefore, it is necessary to perform secondary fine-tuning of the parameters of one component while keeping the parameters of the other component constant, in order to compensate for and optimize the interaction between them.

[0082] Figure 4This is a flowchart illustrating the construction of a transitional single-line structure in the design optimization method for the SISL device described in Embodiment 1 of the present invention, as shown below. Figure 4 The construction of the transition single-line structure may include the following steps:

[0083] Step 410: Replace the circuit of the simplified structure with a simplified single-wire structure, which is generated using the first transmission line.

[0084] Transmission lines are fundamental structures used to guide electromagnetic wave energy from one point to another. In microwave and radio frequency circuits, when the physical dimensions of the circuit are comparable to the operating wavelength, the conductor is no longer merely a current path but must be considered a transmission line. Common types of transmission lines include microstrip lines, striplines, and coaxial lines. In circuit simulation, a transmission line can be modeled as an ideal element with a specific electrical length and characteristic impedance, allowing for the rapid and accurate simulation of signal propagation, delay, reflection, and impedance transformation behavior on a conductor.

[0085] For example, in the HFSS full-wave simulation environment, the original circuit pattern in the simplified structure can be replaced with a uniform transmission line structure formed by direct extension of two ports. This structure retains the original metallic material properties, and its line width is set to be the same as the port width of the original simplified structure. Consistent, the length is the optimal solution of the obtained simplified circuit structure dimensions. The corresponding core circuit length is determined. The purpose of replacing the simplified structure with a simplified single-wire structure is to simplify the complex circuit geometry into an equivalent uniform transmission line segment while ensuring the accuracy of electromagnetic simulation. This facilitates rapid analysis of its basic transmission characteristics in subsequent optimization and provides a stable and reliable interface benchmark for the co-optimization of the transition structure.

[0086] Step 420: Connect the transition structure using the first transmission line to construct a transition single-line structure. The length of the first transmission line is equal to the length of the simplified structure circuit determined by the optimal solution of the simplified structure circuit size. The circuit size of the transition structure corresponds to the transition structure circuit size determined by the preliminary solution of the transition structure circuit size.

[0087] For example, the purpose of this step is to construct a transition single-wire structure for response matching to verify and optimize the collaborative performance between the transition structure and the simplified structure. An optional implementation is to connect the transition structures on both sides using a first transmission line in the HFSS full-wave simulation environment, thereby forming a complete test cell. The transition single-wire structure retains the geometric complexity of the transition circuit while simulating the simplified structure through the first transmission line, providing a simulation model for subsequent response matching.

[0088] For example, the parameters in the transition single-line structure are derived from the optimization results of previous steps, wherein the width of the first transmission line and the width of the core circuit port are... Consistency is maintained to ensure impedance continuity. The length of the first transmission line is equal to the optimal solution obtained from the simplified structure. The length of the core circuit of the simplified structure is determined, thus accurately reproducing the characteristics of the simplified structure. Simultaneously, the initial geometric parameters of the transition section are derived from the preliminary solution of the transition structure circuit dimensions. This is determined as the starting point for secondary optimization, enabling effective coordination between the optimal solution of the simplified structure and the preliminary solution of the transition structure.

[0089] Step 170: Based on the transition single-line structure and the principle of spatial mapping, construct a replacement model for the transition single-line structure; using the circuit dimensions in the transition single-line structure as initial values, adjust the circuit dimensions of the transition single-line structure; when the error between the output response of the replacement model and the output response of the simplified single-line structure converges to the tolerance range, determine the circuit dimensions of the transition single-line structure as the optimal solution for the transition structure circuit dimensions.

[0090] For example, the replacement model for the transition single-line structure can be a high-speed and high-precision replacement model for the transition single-line structure. Optionally, the construction of the replacement model for the transition single-line structure is implemented in the following way: First, in ADS, a fast equivalent circuit is built as a coarse model of the transition single-line structure using a microstrip line of the same size as the physical structure in HFSS. Second, neural network technology is used to learn the nonlinear mapping relationship between the equivalent circuit and the transition single-line structure. When the error between the prediction result of the neural network and the result of the high-precision simulation of the transition single-line structure meets the preset accuracy requirements, the calibrated model is used as the replacement model for the transition single-line structure.

[0091] The quasi-Newton algorithm is an efficient multivariate nonlinear optimization method. Its core idea is to approximate the second derivative matrix (Hessian matrix) of the original problem using the gradient information of the objective function during iteration. In each iteration, the algorithm dynamically corrects its approximate Hessian matrix by utilizing the relationship between the current gradient and the parameter update, thereby achieving intelligent adjustment of the search direction and step size. Compared to the traditional Newton method, which requires direct calculation of the second derivative, the quasi-Newton algorithm significantly reduces computational complexity while maintaining a relatively fast convergence speed. It is suitable for engineering optimization problems with high computational costs and numerous design variables, such as electromagnetic parameter optimization.

[0092] The tolerance range is an acceptable error limit set for the optimization process. For example, the tolerance range defines the maximum allowable difference between the output responses of the transitional single-line structure alternative model and the simplified single-line structure.

[0093] For example, this step can be parameter optimization based on the initial solution of the transition structure. Using the circuit dimensions determined by the initial solution in the transition single-wire structure as the initial values ​​for iteration, a quasi-Newton algorithm is used to adjust this parameter set until the error between the output response of the alternative model of the transition single-wire structure and the output response of the simplified single-wire structure converges to a preset tolerance range. This convergence indicates that the transition structure and the simplified structure have achieved good electromagnetic compatibility and performance matching. At this point, the circuit dimensions of the corresponding transition single-wire structure are determined as the optimal solution for the transition structure circuit dimensions, signifying that the two structures have achieved synergistic optimization in system-level integration. The error is calculated using the following formula. ,in This represents the optimal solution for the size of the transition structure circuit. The error between the output response of the alternative model for the transitional single-line structure and the output response of the simple single-line structure. ,in, An alternative model for a transitional single-line structure in terms of parameters The output response below, The output response is for a simple single-wire structure. To ensure design parameters Realizable physical constraint functions. The core function of physical constraint functions is to ensure that adjustments to geometry are made. To avoid physical interference with other parts of the device, the formula is: ,in, Indicates the first The calculated distance between the transition single-line structure at each location and the nearest adjacent component (such as a metallized via, ground plane, or other transmission line). This represents the minimum safe distance required for this location. Physical constraint function. It plays a crucial role in ensuring design feasibility during the optimization process. The function defines the geometric dimensions mathematically. Manufacturing process constraints must be met, such as the minimum spacing of metal vias and the minimum width of dielectric layer structures—physically feasible conditions. The core objective is to guide the optimization results from ideal mathematical solutions to physically feasible solutions that meet actual production requirements, thereby avoiding invalid design schemes that are impossible to manufacture or structurally unstable.

[0094] For example, after obtaining the optimal solutions for the simplified structure and the transitional structure through the aforementioned steps, the optimal solutions for the simplified structure and the transitional structure are combined into an optimal solution parameter combination. The overall performance of the SISL device was verified by substituting the complete structure of the device into the model. The electromagnetic response was evaluated through full-wave electromagnetic simulation. If the results met the preset design specifications for the SISL device, it indicated that the design method based on the joint optimization of structural decomposition and equivalent models was effectively implemented, and the optimized design of the high-performance SISL device was successfully completed.

[0095] This embodiment decomposes the complete SISL structure into a simplified structure and a transition structure based on the SISL device decomposition coupling formula. The simplified structure and the transition structure are then used as fine models of the simplified structure and the transition structure, respectively. A first equivalent circuit with the same circuit structure as the simplified structure is constructed, and this first equivalent circuit is used as a coarse model of the simplified structure. A second equivalent circuit with the same circuit structure as the transition structure is also constructed, and this second equivalent circuit is used as a coarse model of the transition structure. Based on the fine and coarse models of the simplified structure, the optimal solution for the circuit dimensions of the simplified structure is obtained. Based on the fine and coarse models of the transition structure, the optimal solution for the circuit dimensions of the simplified structure is obtained. A coarse model is constructed to obtain a preliminary solution for the dimensions of the transition structure circuit. Based on the simplified structure circuit, and according to the optimal solution for the simplified structure circuit dimensions and the preliminary solution for the transition structure circuit dimensions, a transition single-wire structure is constructed. Based on the transition single-wire structure and the principle of space mapping, a replacement model for the transition single-wire structure is constructed. Using the circuit dimensions in the transition single-wire structure as initial values, the circuit dimensions of the transition single-wire structure are adjusted. When the error between the output response of the replacement model and the output response of the simplified single-wire structure converges to within the tolerance range, the circuit dimensions of the transition single-wire structure are determined as the optimal solution for the transition structure circuit dimensions. By decomposing the complete SISL structure, the complex optimization problem requiring full-wave simulation is transformed into independent and collaborative optimization of the simplified and transition structures. This allows computational resources to be concentrated on the fine-tuning of key parts, thereby achieving efficient and accurate design of SISL devices and effectively avoiding the computational bottleneck of full-wave simulation.

[0096] Example 2

[0097] Figure 5 This is a schematic diagram of the design optimization device for the SISL device provided in Embodiment 2 of the present invention, as shown below. Figure 5 As shown, the device includes:

[0098] The decomposition module 510 is used to decompose the complete SISL structure into a simplified structure and a transitional structure based on the SISL device decomposition coupling formula, and to use the simplified structure and the transitional structure as a detailed model of the simplified structure and a detailed model of the transitional structure, respectively.

[0099] The first equivalent circuit construction module 520 is used to construct a first equivalent circuit with the same circuit structure as the simplified structure, and to use the first equivalent circuit as a rough model of the simplified structure.

[0100] The second equivalent circuit construction module 530 is used to construct a second equivalent circuit with the same circuit structure as the transition structure, and to use the second equivalent circuit as a coarse model of the transition structure.

[0101] The optimal solution acquisition module 540 is used to obtain the optimal solution for the circuit size of the simplified structure based on the simplified structure fine model and the simplified structure coarse model;

[0102] The preliminary solution acquisition module 550 is used to obtain a preliminary solution of the circuit dimensions of the transition structure based on the fine model and the coarse model of the transition structure.

[0103] Module 560 is used to construct a transitional single-line structure based on the simplified circuit structure, according to the optimal solution of the simplified circuit size and the preliminary solution of the transitional circuit size.

[0104] The optimization module 570 is used to construct a transitional single-line structure replacement model based on the principle of transitional single-line structure and spatial mapping; adjust the circuit size of the transitional single-line structure with the circuit size in the transitional single-line structure as the initial value; and determine the circuit size of the transitional single-line structure as the optimal solution of the transitional structure circuit size when the error between the output response of the transitional single-line structure replacement model and the output response of the simplified single-line structure converges to the tolerance range.

[0105] The design optimization apparatus for SISL devices provided in this embodiment decomposes the complete SISL structure into a simplified structure and a transitional structure based on the SISL device decomposition coupling formula. The simplified structure and the transitional structure are respectively used as a fine model of the simplified structure and a fine model of the transitional structure. A first equivalent circuit with the same circuit structure as the simplified structure is constructed, and this first equivalent circuit is used as a coarse model of the simplified structure. A second equivalent circuit with the same circuit structure as the transitional structure is constructed, and this second equivalent circuit is used as a coarse model of the transitional structure. Based on the fine and coarse models of the simplified structure, the optimal solution for the circuit dimensions of the simplified structure is obtained. Based on the transitional structure... A preliminary solution for the circuit dimensions of the transition structure is obtained by constructing a refined model and a coarse model of the transition structure. Based on the circuit of the simplified structure, a transition single-wire structure is constructed according to the optimal solution of the simplified structure circuit dimensions and the preliminary solution of the transition structure circuit dimensions. Based on the transition single-wire structure and the principle of space mapping, a replacement model for the transition single-wire structure is constructed. Using the circuit dimensions in the transition single-wire structure as initial values, the circuit dimensions of the transition single-wire structure are adjusted. When the error between the output response of the replacement model and the output response of the simplified single-wire structure converges to the tolerance range, the circuit dimensions of the transition single-wire structure are determined as the optimal solution for the transition structure circuit dimensions. By decomposing the complete SISL structure, the complex optimization problem requiring full-wave simulation is transformed into independent and collaborative optimization of the simplified and transition structures. This allows computational resources to be concentrated on the fine-tuning of key parts, thereby achieving efficient and accurate design of SISL devices and effectively avoiding the computational bottleneck of full-wave simulation.

[0106] Based on the above embodiments, the optimal solution acquisition module includes:

[0107] The first setting unit is used to set the sampling range;

[0108] A sample data acquisition unit for a simplified structural fine model is used to sample the simplified structural fine model within the sampling range to obtain sample data of the simplified structural fine model.

[0109] A sample data acquisition unit for a simplified coarse model is used to sample the simplified coarse model within the sampling range to obtain sample data of the simplified coarse model.

[0110] The simplified structure replacement model construction unit is used to construct a simplified structure replacement model based on the sample data of the simplified structure fine model and the sample data of the simplified structure coarse model by utilizing the principle of spatial mapping.

[0111] The first adjustment unit is used to adjust the circuit dimensions of the simplified structural replacement model;

[0112] The simplified structure circuit size optimal solution acquisition unit is used to determine the circuit size of the simplified structure alternative model as the optimal solution for the simplified structure circuit size when the output response of the simplified structure alternative model meets the design specifications of the SISL device.

[0113] Based on the above embodiments, the preliminary solution acquisition module includes:

[0114] The second setting unit is used to set the sampling range;

[0115] A sample data acquisition unit for a fine model of a transition structure is used to sample the fine model of the transition structure within the sampling range to obtain sample data of the fine model of the transition structure.

[0116] A sample data acquisition unit for a coarse model of a transition structure is used to sample the coarse model of the transition structure within the sampling range to obtain sample data of the coarse model of the transition structure.

[0117] The transition structure replacement model construction unit is used to construct a transition structure replacement model based on the sample data of the fine transition structure model and the sample data of the coarse transition structure model by utilizing the principle of spatial mapping.

[0118] The second adjustment unit is used to adjust the circuit dimensions of the transition structure replacement model;

[0119] The preliminary solution acquisition unit for the transition structure circuit size is used to determine the circuit size of the transition structure substitution model as the preliminary solution for the transition structure circuit size when the output response of the transition structure substitution model satisfies the preset convergence condition.

[0120] Based on the above embodiments, the building module includes:

[0121] A simplified single-wire structure building unit is used to replace the circuit of the simplified structure with a simplified single-wire structure, wherein the simplified single-wire structure is generated using a first transmission line;

[0122] A transition single-line structure construction unit is used to connect the transition structure using the first transmission line to construct the transition single-line structure. The length of the first transmission line is equal to the length of the simplified structure circuit determined by the optimal solution of the simplified structure circuit size. The circuit size of the transition structure corresponds to the transition structure circuit size determined by the preliminary solution of the transition structure circuit size.

[0123] The design optimization apparatus for SISL devices provided in this embodiment of the invention can execute the design optimization method for SISL devices provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0124] Example 3

[0125] Figure 6 This is a schematic diagram of the structure of a server provided in Embodiment 3 of the present invention. Figure 6 A block diagram of an exemplary server 12 suitable for implementing embodiments of the present invention is shown. Figure 6 The server 12 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0126] like Figure 6 As shown, server 12 is presented in the form of a general-purpose computing server. The components of server 12 may include, but are not limited to: one or more processors or processing units 16, system memory 28, and bus 18 connecting different system components (including system memory 28 and processing unit 16).

[0127] Bus 18 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, the Industry Standard Architecture (ISA) bus, the Micro Channel Architecture (MAC) bus, the Enhanced ISA bus, the Video Electronics Standards Association (VESA) local bus, and the Peripheral Component Interconnect (PCI) bus.

[0128] Server 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by server 12, including volatile and non-volatile media, removable and non-removable media.

[0129] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache 32. Server 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be used to read and write non-removable, non-volatile magnetic media (… Figure 6 Not shown; usually referred to as a "hard drive"). Although Figure 6 Not shown, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk") and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments of the present invention.

[0130] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. Program modules 42 typically perform the functions and / or methods described in the embodiments of the present invention.

[0131] Server 12 can also communicate with one or more external devices 14 (e.g., keyboard, pointing server, display 24, etc.), and with one or more servers that enable users to interact with server 12, and / or with any server (e.g., network card, modem, etc.) that enables server 12 to communicate with one or more other computing servers. This communication can be performed via input / output (I / O) interface 22. Furthermore, server 12 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 20. As shown, network adapter 20 communicates with other modules of server 12 via bus 18. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with server 12, including but not limited to: microcode, server drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0132] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, such as implementing the design optimization method for SISL devices provided in the embodiments of the present invention.

[0133] Example 4

[0134] Embodiment 4 of the present invention also provides a storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a design optimization method for any of the SISL devices described in the above embodiments.

[0135] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0136] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0137] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including—but not limited to—wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0138] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0139] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. A design optimization method for SISL devices, characterized in that, include: Based on the SISL device decomposition and coupling formula, the complete SISL structure is decomposed into a simplified structure and a transitional structure. The simplified structure and the transitional structure are respectively used as a detailed model of the simplified structure and a detailed model of the transitional structure. The simplified structure is the intermediate dielectric substrate layer circuit in the complete SISL structure, and the transitional structure is the upper dielectric substrate layer circuit, the lower dielectric substrate layer circuit, and the metal via in the complete SISL structure. Construct a first equivalent circuit with the same circuit structure as the simplified structure, and use the first equivalent circuit as a coarse model of the simplified structure. Construct a second equivalent circuit with the same circuit structure as the transition structure, and use the second equivalent circuit as a coarse model of the transition structure; Based on the simplified fine model and the simplified coarse model, the optimal solution for the circuit dimensions of the simplified structure is obtained; Based on the fine and coarse models of the transition structure, a preliminary solution for the circuit dimensions of the transition structure is obtained; Based on the circuit with the simplified structure, and according to the optimal solution of the simplified structure circuit size and the preliminary solution of the transition structure circuit size, a transition single-line structure is constructed, including: replacing the circuit with the simplified structure with a simplified single-line structure, wherein the simplified single-line structure is generated using a first transmission line; The first transmission line is used to connect the transition structure to construct a transition single-line structure. The length of the first transmission line is equal to the length of the simplified structure circuit determined by the optimal solution of the simplified structure circuit size. The circuit size of the transition structure corresponds to the transition structure circuit size determined by the preliminary solution of the transition structure circuit size. Based on the principle of transitional single-line structure and spatial mapping, a replacement model for the transitional single-line structure is constructed. Taking the circuit size in the transitional single-line structure as the initial value, the circuit size of the transitional single-line structure is adjusted. When the error between the output response of the replacement model and the output response of the simplified single-line structure converges to the tolerance range, the circuit size of the transitional single-line structure is determined as the optimal solution for the circuit size of the transitional structure.

2. The method according to claim 1, characterized in that, The aforementioned SISL device decomposition coupling formula decomposes the complete SISL structure into a simplified structure and a transitional structure, including: Based on the SISL device decomposition and coupling formula, the complete SISL structure is decomposed into a simplified structure and a transitional structure. The SISL device decomposition and coupling formula is as follows: in, The response matrix of the complete SISL structure. For a simplified structural response matrix, The transition structure response matrix, Let be the coupling matrix. This is the boundary field coupling correction matrix. For circuit size parameters of a simplified structure, For the circuit size parameters of the transition structure, The circuit size parameters of the complete SISL structure and .

3. The method according to claim 1, characterized in that, The first equivalent circuit, which constructs the same circuit structure as the simplified structure, includes: Using microstrip lines, a first equivalent circuit with the same circuit structure as the simplified structure described above is constructed. Accordingly, the construction of the second equivalent circuit with the same circuit structure as the transition structure includes: Using microstrip lines, a second equivalent circuit with the same circuit structure as the transition structure is constructed.

4. The method according to claim 1, characterized in that, The step of adjusting the circuit dimensions of the transition single-wire structure, using the circuit dimensions in the transition single-wire structure as initial values, includes: Using the circuit dimensions in the transition single-wire structure as initial values, the circuit dimensions of the transition single-wire structure are adjusted using a quasi-Newton algorithm. When the error between the output response of the alternative model of the transitional single-wire structure and the output response of the simplified single-wire structure converges to within the tolerance range, the circuit dimensions of the transitional single-wire structure are determined as the optimal solution for the circuit dimensions of the transitional structure, and the error is calculated in the following manner: in This represents the optimal solution for the size of the transition structure circuit. The error between the output response of the alternative model for the transitional single-line structure and the output response of the simplified single-line structure; in, An alternative model for a transitional single-line structure in terms of parameters The output response below, The output response is for a simple single-wire structure. To ensure design parameters Realizable physical constraint functions.

5. The method according to claim 1, characterized in that, The process of obtaining the optimal solution for the simplified structure circuit dimensions based on the simplified fine model and the simplified coarse model includes: Set the sampling range; Within the sampling range, the simplified structural fine model is sampled to obtain sample data of the simplified structural fine model; Within the sampling range, the simplified coarse model of the structure is sampled to obtain sample data of the simplified coarse model of the structure; Using the principle of spatial mapping, a simplified structure replacement model is constructed based on the sample data of the simplified detailed model and the simplified coarse model. Adjust the circuit dimensions of the simplified structural alternative model; When the output response of the simplified structure replacement model meets the design specifications of the SISL device, the circuit size of the simplified structure replacement model is determined as the optimal solution for the simplified structure circuit size.

6. The method according to claim 1, characterized in that, The preliminary solution for the circuit dimensions of the transition structure, based on the fine and coarse models of the transition structure, includes: Set the sampling range; Within the sampling range, the fine model of the transition structure is sampled to obtain sample data of the fine model of the transition structure; Within the sampling range, the coarse model of the transition structure is sampled to obtain sample data of the coarse model of the transition structure; Using the principle of spatial mapping, a replacement model for the transition structure is constructed based on the sample data of the fine model of the transition structure and the sample data of the coarse model of the transition structure. Adjust the circuit dimensions of the transition structure replacement model; When the output response of the transition structure replacement model satisfies the preset convergence condition, the circuit dimensions of the transition structure replacement model are determined as the preliminary solution of the transition structure circuit dimensions.

7. A design optimization device for SISL devices, characterized in that, include: The decomposition module is used to decompose the complete SISL structure into a simplified structure and a transitional structure based on the SISL device decomposition coupling formula. The simplified structure and the transitional structure are respectively used as a detailed model of the simplified structure and a detailed model of the transitional structure. The simplified structure is the intermediate dielectric substrate layer circuit in the complete SISL structure, and the transitional structure is the upper dielectric substrate layer circuit, the lower dielectric substrate layer circuit and the metal via in the complete SISL structure. The first equivalent circuit construction module is used to construct a first equivalent circuit with the same circuit structure as the simplified structure, and the first equivalent circuit is used as a rough model of the simplified structure. The second equivalent circuit construction module is used to construct a second equivalent circuit with the same circuit structure as the transition structure, and the second equivalent circuit is used as a coarse model of the transition structure. The optimal solution acquisition module is used to obtain the optimal solution for the circuit dimensions of the simplified structure based on the simplified fine model and the simplified coarse model; The preliminary solution acquisition module is used to obtain a preliminary solution for the circuit dimensions of the transition structure based on the fine model and the coarse model of the transition structure. A construction module is used to construct a transitional single-line structure based on the simplified circuit structure, according to the optimal solution of the simplified circuit structure dimensions and the preliminary solution of the transitional circuit structure dimensions. The construction includes: replacing the simplified circuit structure with a simplified single-line structure, the simplified single-line structure being generated using a first transmission line; connecting the transitional structure using the first transmission line to construct the transitional single-line structure, wherein the length of the first transmission line is equal to the simplified circuit length determined by the optimal solution of the simplified circuit structure dimensions, and the circuit dimensions of the transitional structure correspond to the transitional circuit dimensions determined by the preliminary solution of the transitional circuit structure dimensions. An optimization module is used to construct a replacement model for the transitional single-line structure based on the principle of transitional single-line structure and spatial mapping; adjust the circuit size of the transitional single-line structure using the circuit size in the transitional single-line structure as the initial value; and determine the circuit size of the transitional single-line structure as the optimal solution for the circuit size of the transitional structure when the error between the output response of the replacement model and the output response of the simplified single-line structure converges to the tolerance range.

8. A server, characterized in that, The server includes: One or more processors; Storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the design optimization method for SISL devices as described in any one of claims 1-6.

9. A storage medium containing computer-executable instructions, which, when executed by a computer processor, are used to perform a design optimization method for a SISL device as described in any one of claims 1-6.

Images