電力潮流問題と解決策

The power flow problem calculates voltage magnitude, phase angle, real power, and reactive power flows in a balanced three-phase steady-state power system. Each bus has four variables: two are provided as input data, while the power flow program computes the remaining two. These calculations determine complex power delivered to each bus based on the system's single-line diagram and network configuration.

Buses are classified as swing bus, load (PQ) bus, or voltage-controlled bus. The swing bus maintains a voltage magnitude near 1.0 per unit and zero phase angle. Load buses have specified real and reactive power with unknown voltage magnitude and phase angle. Voltage-controlled buses have specified real power and voltage magnitude, with reactive power calculated by the power flow program.

Transmission lines are represented by equivalent pi circuits requiring four input data elements: series impedance, shunt admittance, connected buses, and maximum megavolt-ampere rating. This data constructs the bus admittance matrix, which leads to nodal equations for the network and determines complex power delivered to each bus in the system.

The Gauss-Seidel method iteratively solves nodal equations by recalculating current for load buses using known power values and adjusting reactive power for voltage-controlled buses until convergence. This approach uses the admittance matrix and complex power relationships to progressively refine voltage estimates across the network until the solution stabilizes.

The Newton-Raphson method linearizes power flow equations and uses the Jacobian matrix for voltage corrections, generally converging faster and more reliably than Gauss-Seidel. This superior convergence makes it more suitable for large systems where computational efficiency and accuracy are critical for maintaining stability and efficiency across the network.

Power delivered to a bus is split into generator and load terms, expressed through complex power relationships derived from nodal equations. By taking real and imaginary parts of these equations, power balance equations are formed for each bus, enabling the power flow program to calculate how real and reactive power distribute throughout the system.

The bus admittance matrix is constructed from transmission line input data and forms the foundation for nodal equations representing the entire network. This matrix relates source currents injected at each bus to bus voltages, enabling both Gauss-Seidel and Newton-Raphson iterative methods to calculate complex power delivered and determine maximum power flow and line loadability constraints.