PCB Layout Parameter Prediction Using Machine Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current PCB layout simulation software is time-consuming, requiring several minutes to simulate each layout parameter combination and potentially several days to complete a design for communication products with multiple circuit boards, leading to high labor and time costs.
Innovation Solution
A device with a PCB layout parameter setting system that uses machine learning models to learn from simulation software, reducing design time by establishing prediction models for interference parameters like impedance, insertion loss, and crosstalk, allowing for faster calculation of optimal layout parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If predetermined PCB layout simulation software is used to verify each layout parameter combination, then measurement precision of interference parameters is improved, but loss of time increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-collecting simulation data from the predetermined PCB layout simulation software to build a training dataset. This pre-processing step enables the subsequent machine learning model to make rapid predictions without requiring time-consuming real-time simulations, thus resolving the contradiction between measurement precision and time loss.
Solution Approach 2:
The patent creates a copy of the simulation software's functionality through a machine learning prediction model. Instead of repeatedly executing the original simulation software, the trained model replicates its interference parameter calculation capabilities, providing accurate predictions in fractions of a second compared to the minutes required by traditional simulation.
2Manufacturing precision
If comprehensive evaluation of multiple layout parameter combinations is performed, then manufacturing precision is improved, but productivity decreases due to extended design time
Solution Approach 1:
The patent changes the parameter representation by transforming physical layout parameters and interference parameters into standardized numerical features suitable for machine learning processing. This transformation enables efficient computation of multiple parameter combinations while maintaining the ability to evaluate design quality, thus improving productivity without sacrificing manufacturing precision.
Solution Approach 2:
The patent substitutes the mechanical simulation process with a machine learning-based prediction system. The traditional simulation software that requires minutes per calculation is replaced by a trained model that provides predictions in 0.02 seconds, dramatically increasing design throughput while maintaining evaluation accuracy.
3Reliability
If multiple layout parameter combinations are simulated to determine optimal design, then reliability of PCB layout is improved, but loss of time increases from several hours to several days
Solution Approach 1:
The patent enables continuous evaluation of multiple layout parameter combinations by replacing discrete, time-consuming simulations with a continuous machine learning prediction system. The trained model can rapidly assess numerous design options in sequence, allowing comprehensive reliability evaluation to be completed in minutes rather than days, thus resolving the contradiction between reliability and time loss.
Data Source
AI summary
A method for setting parameters in design of a printed circuit board (PCB) includes obtaining multiple combinations of layout parameters of a PCB and inputting the multiple combinations of layout parameters into a predetermined PCB layout simulation software to obtain multiple interference parameter combinations. The multiple combinations of layout parameters and the multiple interference parameter combinations are defined as training samples, and a predetermined network model is trained through the training samples to obtain a first prediction model. The first prediction model is trained and tested to obtain an impedance prediction model. When the multiple combinations of layout parameters are inputted to the impedance prediction model, only an average predetermined error is allowed between impedance values predicted by the impedance prediction model and impedance values calculated by the predetermined PCB layout simulation software, to enable acceptance of that combination.