Power System State Prediction via Time Domain Signal Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional power system monitoring only allows for reaction to faults after they have occurred, leading to potential damage, and existing methods primarily focus on failure prediction in the frequency domain, limiting their application beyond fault detection.
Innovation Solution
A method and device that use the frequency spectrum of periodic waveform signals from power systems to predict future states by determining frequencies and complex amplitudes, allowing for the calculation of predicted values in the time domain, which can trigger control actions at optimal times to minimize negative effects and detect faults by comparing predicted and measured values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional monitoring methods are used to detect faults in power systems, then the system can react to faults after they have occurred, but potential damage cannot be prevented and response time is delayed
Solution Approach 1:
The patent applies preliminary action by performing Fast Fourier Transform on periodic waveform signals to predict future system states before faults actually occur. The method calculates predicted values at future time points and compares them with measured values to detect deviations early, enabling proactive fault detection and prevention rather than reactive response after damage has occurred.
2Reliability
If frequency domain methods are used for failure prediction, then fault detection is possible, but the application is limited solely to failure prediction and cannot provide broader system state prediction
Solution Approach 1:
The patent achieves universality by extending the Fast Fourier Transform method from solely failure prediction to comprehensive system state prediction. The method can predict various parameters including voltage, current, power, and other electric quantities by analyzing their periodic waveform signals, making it applicable to multiple aspects of power system monitoring and control beyond just failure detection.
3Loss of information
If the Fast Fourier Transform is performed on periodic waveform signals to determine frequency spectrum, then the state of the power system can be predicted, but the computational complexity increases
Solution Approach 1:
The patent applies the taking out principle by extracting only the essential frequency components and complex amplitudes from the complete frequency spectrum obtained through Fast Fourier Transform. Instead of processing all frequency data, the method identifies and uses only the significant peaks that contribute to predicting the system state, thereby reducing computational complexity while maintaining prediction accuracy.
Data Source
AI summary
A method to predict a state of a power system and a device to perform the method are presented, where the method includes the step of performing a Fast Fourier Transform on samples of a periodic waveform signal in order to determine a frequency spectrum of the signal, where the signal was derived from at least one measurable electric quantity of the power system and is representative for the state of the power system. Frequencies and related complex amplitudes at selected peaks in the frequency spectrum are determined and stored together with a time stamp. Afterwards, a predicted value of the periodic waveform signal in the time domain is determined at a prediction time by calculating a sum of sinusoidal signals at the prediction time, where each of the sinusoidal signals is characterized by one of the determined frequencies and its related complex amplitude as well as the time stamp.