Circuit Far Field Estimation via Emphasis Pattern Image
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating the far field of electromagnetic waves radiated from circuits using machine learning, such as deep learning, face challenges in accuracy due to variations in circuit positions and difficulties in distinguishing different circuits, leading to inconsistent estimation results.
Innovation Solution
The proposed solution involves generating an emphasis pattern image by emphasizing key elements of a circuit pattern image based on their type and influence on electromagnetic wave radiation, which is then used to estimate the far field using an existing learning model, thereby improving estimation accuracy and speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If deep learning is used to estimate far field of electromagnetic waves based on circuit images, then estimation speed is improved, but estimation accuracy deteriorates due to position variations and inability to distinguish different circuits
Solution Approach 1:
The patent segments the circuit image into multiple regions of interest based on proximity to the observation point. By dividing the circuit into near-field regions and far-field regions, the model can selectively process only the relevant portions that significantly influence electromagnetic wave radiation, thereby improving both accuracy and computational efficiency
Solution Approach 2:
The patent applies different processing weights to different regions of the circuit image based on their local characteristics and distance from the observation point. Regions closer to the observation point are assigned higher weights as they have greater influence on far-field radiation, while distant regions receive lower weights. This local differentiation enables the model to focus computational resources on critical areas
2Measurement precision
If the entire circuit image is processed to improve estimation accuracy, then measurement precision is improved, but device complexity and computational cost increase
Solution Approach 1:
The patent extracts only the necessary portions of the circuit image that significantly contribute to far-field electromagnetic wave radiation. By identifying and isolating regions of interest based on their spatial relationship to the observation point, the model eliminates redundant computational processing of irrelevant circuit portions, reducing overall computational complexity while maintaining accuracy
Solution Approach 2:
The patent applies partial processing to the circuit image by focusing computational effort only on regions that have substantial influence on the far-field characteristics. Instead of uniformly processing the entire circuit image, the model selectively intensifies processing for near-field regions while reducing or skipping processing for far-field regions, achieving optimal accuracy-efficiency balance
Data Source
AI summary
An estimation method is performed by a computer for estimating a far field of electromagnetic waves or heat. The method includes: generating an emphasis pattern image obtained by emphasizing each target element of a pattern image of a target circuit by an emphasizing method that corresponds to a type of each target element, with respect to the target element which is at least a part of elements included in the target circuit; and estimating the far field of electromagnetic waves or heat radiated from the target circuit by an existing estimation model using the emphasis pattern image.