Voltage Sag Compensation with DVR based on Machine Learning Controller

Machines and automation are becoming commonly used in multiple fields like automobiles, healthcare, finance, information technology, media etc. These machines are sensitive and fragile. Any disturbances such as unprecedented interruptions, sags, swells, sudden overvoltages and harmonics in the electrical distribution systems can permanently damage the equipment or automation circuits connected to it. In fields like healthcare and automobiles where lives are at stake, the damage will not only be limited to the machine getting damaged but directly affects the financial stability also. The issue in recent days is the negligence of these kinds of hazards. In developing countries like India, these situations are very common. If these issues are managed properly, a lot of machinery and automation can be saved, which leads to a lot of savings for the company or industry. The UPS (Uninterruptible Power System) is more commonly used. Due to its high steady-state loss, the applications in which it can be used are very finite. This article describes the working of a system that protects the machinery from momentary sags using an ANN (Artificial Neural Network) controller-based DVR and compares it with a PI controller-based DVR. This proposed system can restore the rated voltage during voltage sag conditions in a more efficient way than the already existing methods.