Immediate detection of faults in systems plays an important role in improving their reliability, productivity, and safety. Hence I need suggestions on mathematical principles or transformations that I can use in developing algorithms which will detect faults/transients in less than 0.01 seconds in a 50 Hz power system
Hi there,
I think detection time is related to the limitation of measurement circuits. Most measurement units (CT and PT) typically takes atleast 1 cycle to detect and the fastest breaker around 2 cycles to operate.
I hope it will answer some part of your query.
We are talking about a power line (120/240 VAC), right?
Just some ideas:
1) Use a power semiconductor device and monitor dI/dt as well as the gate voltage on the psd. (Probably still needs one full cycle to detect). Maybe look up Stefan Hain, he works on super fast short circuit detection;
2) feed a phase shifted (90°) signal with at least 100 Hz to your AC line and monitor the signal of both, the signal generator and the line. The higher frequency would allow for 0.01s detection, other than the 50 Hz mains, because one cycle takes 0.02 s. You could set a threshold for d(Amplitude)/dt, or d(frequency)/dt depending on the modulation you apply and the phase and frequency you choose.
In any case, software based short circuit detection in these timeframes is a very hard thing to achieve and most likely damage would have happened before the software could break the circuit. Most known solutions take at least one full cycle, if not more under real-world conditions.
Samundra Gurung and Patrick Guth , thanks a lot for your responses. Patrick Guth , your number 2 suggestion is highly recommended and I am already working on it. Muchas Gracias
PMUs come to mind, that is, Phase Measurement Units, which unfortunately are not deployed widely in Power grids outside the United States etc due to their being very expensive. Your choice of a detection system also relates to the value of your infrastructure. Never allow the cost of treating a disease exceed the cost of the disease itself. Mathematical analysis is one thing but engineers are focussed on network solution value and improvement on power quality. Researchers need to gain greater exposure to actually power networks to have a better grasp of the lesser value of mathematical algorithms compared to practical testing to establish operating limits. Practicing engineers will continue to compare these theoretical and experimental innovations with skepticism simply field to the cost/benefit analysis. Good work.
- Badges
- Science topic
- Similar topics
- Publishing