Distribution Line Parameter Estimation Using iPINN Voltage Dynamics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing joint parameter-state estimation techniques for electrical distribution networks face challenges due to insufficient instrumentation, uncertainty in line parameters, and high computational complexity, making accurate real-time monitoring and efficient network management difficult.
Innovation Solution
A method and system utilizing lossy Distflow equations and an inverse Physics-Informed Neural Network (iPINN) architecture to learn network voltage dynamics and estimate line parameters, reducing computational intensity by determining line resistance, reactance, and power flows without relying on initial values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If joint parameter-state estimation techniques are used to estimate line parameters and network states, then measurement accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The joint estimation problem is segmented into two separate sequential steps: first performing state estimation using weighted least square method to obtain network states, then using these states to estimate line parameters through a dedicated parameter estimation algorithm. This segmentation reduces the computational complexity of the overall system while maintaining estimation accuracy by breaking down the complex joint estimation into manageable independent tasks.
Solution Approach 2:
The state estimation is performed as a preliminary action before line parameter estimation. By first obtaining accurate network states (voltages, power flows) through state estimation using available measurements, these states are then used as inputs for the subsequent parameter estimation process. This preliminary action enables the parameter estimation to proceed with reduced computational burden compared to simultaneous joint estimation.
2Reliability
If Phasor Measurement Units are deployed for state and parameter estimation, then estimation reliability is improved, but device cost and complexity increase due to insufficient instrumentation in distribution systems
Solution Approach 1:
The system uses self-service by leveraging the existing measurement infrastructure already present in distribution networks (such as smart meters, voltage sensors, and power flow measurements) rather than requiring additional Phasor Measurement Units. The estimation algorithms are designed to extract reliable state and parameter information from these readily available measurements, making the system self-sufficient without needing expensive additional instrumentation.
Solution Approach 2:
The estimation system is designed to be universal by working with multiple types of measurements that are commonly available in distribution networks (voltage magnitudes, active and reactive power injections, power flows). The algorithm can handle different measurement configurations and types, making it adaptable to various distribution network setups without requiring specialized Phasor Measurement Units, thus achieving multi-functionality with existing infrastructure.
3Speed
If Gauss-Newton solver is used for parameter estimation, then convergence speed is improved, but sensitivity to initial point increases making the system less robust
Solution Approach 1:
The parameter estimation algorithm incorporates feedback mechanisms where the estimated states from the state estimation step are fed into the parameter estimation process. Additionally, the algorithm uses iterative refinement where initial guesses for line parameters are updated based on measurement residuals and state information, improving both convergence speed and robustness by continuously adjusting estimates based on feedback from the measurement system.
Solution Approach 2:
The algorithm employs parameter changes by transforming the parameter estimation problem into a form that is less sensitive to initial conditions. By using the previously estimated states and incorporating measurement information, the algorithm dynamically adjusts the estimation parameters and search directions, making the convergence process more robust to initial parameter guesses while maintaining fast convergence through efficient numerical methods.
Data Source
Figure 1~2
Figure 3
Figure 4A
AI summary
Accurate real-time monitoring of distribution networks becomes increasingly pivotal to ensure reliable and optimal operation of grid. Existing joint parameter-state estimation techniques predominantly focus on transmission systems, making their applicability to distribution systems challenging due to variations in available input data. Furthermore, the joint estimation techniques significantly suffer from computational complexity. Present disclosure provides a method and a system for personalized multi-subject text to image generation. The system first uses lossy Distflow equations and an inverse Physics-Informed Neural Network (iPINN) architecture for learning network voltage dynamics, drift matrix and network bus voltages from real-time bus power injection data and historical voltage data. Thereafter, the system determines a plurality of line parameters from the learnt drift matrix using a least-square optimization technique. Finally, the system calculates line power flows and line currents using the lossy Distflow equations.