This patent describes a method for efficiently managing inverse admittance matrices in real-time circuit simulations with multiple switches. It involves storing initial matrices based on user input and using a control variable to minimize computational load. When a switch changes state, the method quickly updates the matrices using pre-stored data and the Woodbury formula, which reduces the need for complex calculations. By maintaining high differences (Hamming distance) between stored matrices, the method ensures fast and accurate updates. This technique is demonstrated to improve simulation performance, especially in power electronic circuits, by reducing computation time and maintaining accurate real-time results.
Type of IP
Faculty
Department
- Increased Renewable Energy Integration: Efficient and accurate simulations support the integration of renewable energy sources into the power grid, promoting sustainable energy practices and reducing reliance on fossil fuels.
- Enhanced Safety in Electrical Systems: By accurately simulating circuit behaviors, potential faults and hazardous conditions can be predicted and mitigated, leading to safer electrical systems in both residential and industrial settings.
- Pre-computed Inverse Admittance Matrices: Pre-computing and storing inverse admittance matrices based on initial switch states reduces the need for real-time calculations, enabling faster simulation processes.
- Hamming Distance-based Matrix Selection: Storing matrices with high Hamming distances ensures that any computed matrix remains close to a stored one, which minimizes real-time computational effort.
- Dynamic Matrix Updates Using Woodbury Formula: The dynamic update of inverse admittance matrices using the Woodbury formula allows for quick and efficient adjustments during switch state changes, maintaining simulation accuracy and speed.
- Control Variable for Matrix Storage: Utilizing a user-defined control variable to determine the number of matrices stored allows customization of the balance between memory usage and computational load, enhancing the simulation's adaptability.
- Efficient Memory Usage: Limiting the number of stored matrices based on strategic selection ensures that memory resources are used efficiently, avoiding unnecessary storage and related computational delays.
- Real-time Simulation Efficiency: By minimizing the computational cost of matrix inversions during state transitions, the method ensures efficient real-time simulation, crucial for accurate and timely results in power electronics.
Power Electronics, Renewable Energy, Simulation Software, Semiconductor Manufacturing
Power Electronics, Real-Time Simulation, Renewable Energy Systems, Electrical Grid Management, Circuit Design, Smart Grids, Computational Efficiency