Dual-Mode Power Amplifier Simulation with ML Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional electronic circuit simulators face challenges in accurately simulating nonlinear effects and dynamic performance characteristics, leading to inconsistencies and inefficiencies due to errors in modeling nonlinear components and complex signal interactions, which result in design errors, increased development time, and costs.
Innovation Solution
A machine learning-based post-processing system that uses a trained model to augment simulation results by learning from both simulation and measured data, reducing errors such as thermal modeling, surface mount component, and harmonic balance errors, thereby improving the accuracy of electronic circuit simulations without requiring extensive manual tuning or modification of the simulation model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional simulators use sophisticated nonlinear simulation models to accurately represent nonlinear components and complex signal interactions, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent introduces a machine learning model as an intermediary between the conventional simulator and the final results. The simulator generates initial simulation data, which is then processed by the ML model trained on measured data to correct systematic errors. This intermediary approach improves accuracy without requiring the simulator itself to become more complex.
Solution Approach 2:
The patent replaces the traditional mechanical approach of improving simulator accuracy through complex model modifications with a data-driven machine learning approach. Instead of adjusting simulation models to match reality, the ML model learns the discrepancies from measured data and applies corrections, substituting the mechanical model-tuning process with an intelligent correction system.
2Measurement precision
If designers manually tune simulation models through backfitting to match measured results, then measurement precision improves, but loss of time increases
Solution Approach 1:
The patent performs preliminary action by training the machine learning model once on measured data from actual circuit operation. This training phase captures the systematic errors and characteristics of the specific circuit implementation. Once trained, the model can be applied to simulate and predict performance under various conditions without requiring repeated manual tuning, saving significant design time.
Solution Approach 2:
The patent enables the simulation system to self-correct by using the machine learning model that automatically adjusts simulation results based on patterns learned from measured data. Instead of requiring designers to manually identify and correct errors, the system performs self-calibration through the ML model, which autonomously applies corrections to simulation outputs.
3Device complexity
If simulation models are created based on limited measurements, then device complexity decreases, but measurement precision worsens
Solution Approach 1:
The patent transforms the simulation approach by changing from using raw simulation outputs to using corrected outputs that incorporate measured data characteristics. The machine learning model learns the relationship between simulation parameters and actual measured behavior, enabling the system to maintain simple simulation models while achieving high precision through parameter correction based on training data.
Data Source
AI summary
A dual-mode power amplifier simulation system is provided. The dual-mode power amplifier system includes a trained machine learning model that is trained with first simulation results associated with a first power amplifier circuit and measured results associated with a physical implementation of the first power amplifier circuit. The trained machine learning model is configured to generate augmented simulation results. In addition, a simulator executes on one or more computer processors that simulates a dual-mode power amplifier circuit and generates dual-mode simulation results. A post processor including the trained machine learning model executes on one or more computing devices with computer-executable instructions that, when executed, causes the post processor to augment the dual-mode simulation results for the dual-mode power amplifier based on the trained machine learning model associated with the first power amplifier circuit.