Intelligent Electronic Device Zero-Crossing Estimation Under Signal Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for measuring line frequency in Intelligent Electronic Devices are prone to errors due to noise in electrical signals, which can lead to maloperation or failure of customer equipment.
Innovation Solution
An Intelligent Electronic Device equipped with sensors, analog-to-digital converters, and a processing module that uses a training dataset and hypothesis function to estimate zero-crossing positions and compute fundamental frequency and RMS voltage of power line signals, thereby improving measurement accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional power line frequency measurement methods (zero-crossing, digital Fourier Transform, phase-locked loop) are used, then the device can measure frequency, but measurement accuracy deteriorates due to noise on electrical signals
Solution Approach 1:
The patent applies preliminary action by collecting training data in advance that represents various noise conditions, then using this pre-collected data to train a machine learning model before actual frequency measurements are taken. This allows the system to have already learned how to handle different noise scenarios before encountering them during operation, improving measurement accuracy without adding real-time processing complexity.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between the raw electrical signal and the frequency measurement process. This intermediary has been trained to recognize patterns and filter noise, allowing it to mediate the measurement process by providing noise-resistant frequency estimates even when the input signals are corrupted by electromagnetic radiation or other noise sources.
2Measurement precision
If machine learning-based frequency measurement is implemented, then measurement accuracy improves, but device complexity increases due to training data collection and model training requirements
Solution Approach 1:
The patent applies self-service by enabling the Intelligent Electronic Device to collect its own training data from its operational environment and train its own machine learning model without requiring external intervention or specialized equipment. The device uses its existing sensors and processing capabilities to generate training datasets under various noise conditions and iteratively improve its frequency measurement algorithm, making the system self-improving and reducing deployment complexity.
Data Source
AI summary
Provided are a method and apparatus to measure line frequency. Specifically, an Intelligent Electronic Device employs a method in which a processor receives a training dataset including an input variable set and a corresponding zero-crossing position output variable, obtains a hypothesis function based on the training dataset, estimates zero-crossing positions using the hypothesis function and computes a fundamental frequency of the signal based on the estimated zero-crossing positions.