Analysis of Microprocessor Based Protective Re-lays (MBPR) Differential Equation Algorithms

This paper analyses and explains from the systems point of view, microprocessor based protective relay (MBPR) systems with emphasis on differential equation algorithms. Presently, the application of protective relaying in power systems, using MBPR systems, based on the differential equation algorithm is valued more than the protection relaying based on any other type of algorithm, because of advantages in accuracy and implementation. MBPR differential equation approach can tolerate some errors caused by power system abnormality such as DC offset. This paper shows that the algorithm is a system description based and it is immune from distortions such as DC-offset. Differential equation algorithms implemented in MBPR are widely used in the protection of transmission and distribution lines, transformers, buses, motors, etc. The parameters from the system, utilized in these algorithms, are obtained from the power system current i(t) or voltage v(t), which are abnormal values under fault or distortion situations. So, an error study for the algorithm is considered necessary.

💡 Research Summary

The paper presents a comprehensive systems‑level examination of microprocessor‑based protective relays (MBPR) that employ differential‑equation algorithms for fault detection and isolation in power networks. It begins by contrasting traditional protection schemes—such as over‑current, voltage‑current ratio, and differential current methods—with the differential‑equation approach, highlighting the latter’s inherent immunity to waveform distortions like DC offset and high‑frequency harmonics. By directly modeling the fundamental electrical relationship v(t)=R·i(t)+L·di(t)/dt, the algorithm avoids reliance on magnitude thresholds that can be compromised by transient phenomena.

The authors first derive the discrete‑time form of the continuous differential equation for typical protection elements (transmission line sections and transformer windings) modeled as first‑order RL circuits. They show that with a sufficiently small sampling interval Δt, the difference equation v(k)=R·i(k)+L·