⚡ electrical engineering

A Constrained Formulation for Simultaneous Line Parameter Estimation and Instrument Transformer Calibration

This paper proposes a novel constrained framework that simultaneously estimates line parameters and calibrates instrument transformers using PMU data, effectively resolving their mutual dependency without requiring additional hardware or offline testing.

Original authors: Antos Cheeramban Varghese, Rajasekhar Anguluri, Anamitra Pal

Published 2026-04-22

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Antos Cheeramban Varghese, Rajasekhar Anguluri, Anamitra Pal

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Problem: The "Broken Ruler" Dilemma

Imagine you are trying to measure the length of a room. You have a tape measure (the Instrument Transformer), but over time, the markings on the tape have stretched or shrunk due to heat and age. Now, your tape measure is slightly "broken."

If you try to use this broken tape to measure the room, you get the wrong length. But here is the catch: You don't know how long the room actually is.

In the power grid world, this is a massive chicken-and-egg problem:

To calibrate the "tape measure" (the Instrument Transformer), you need to know the exact physics of the power lines (the room).
To know the physics of the power lines, you need a perfectly calibrated "tape measure."

For years, engineers had to choose between taking the equipment offline (stopping power flow) to fix it, or guessing. This paper proposes a clever way to fix both problems at the same time without stopping the power.

The Solution: The "Smart Detective" Approach

The authors created a new method called SLIC (Simultaneous Line Parameter Estimation and Instrument Transformer Calibration). Think of it as a detective solving a crime where the witness (the meter) is lying, and the crime scene (the power line) is shifting.

Here is how they solved it, broken down into three simple steps:

1. The "Gold Standard" Anchor

The researchers realized that while most meters in the grid are "drifting" (getting inaccurate), there is usually at least one pair of Revenue Quality Meters (RQMs) at the very beginning of a line. These are the "Gold Standard" meters used for billing; they are incredibly accurate and rarely drift.

The Analogy: Imagine you are trying to calibrate a whole classroom of faulty rulers. You know that the teacher's ruler is perfect. You don't need to know the exact length of the desk to start; you just need to compare every student's ruler to the teacher's ruler.
The Paper's Move: They use the "Gold Standard" meter as a reference point. They assume the other meters are slightly off compared to this one, and they calculate the "correction factor" (how much to stretch or shrink the reading) relative to the perfect one.

2. The "Physics Rules" (The Constraints)

Even with a reference point, the math is tricky because there are too many unknowns. To fix this, the authors used the Laws of Physics as "rules" that the solution must follow.

The Analogy: Imagine you are trying to guess the weights of three people standing on a seesaw. You don't know their individual weights, but you know the Law of the Seesaw: If the seesaw is balanced, the total weight on the left must equal the total weight on the right.
The Paper's Move: They applied Kirchhoff's Laws (the rules of electricity).
- Voltage Rule: The voltage at a junction (bus) must be the same, no matter which wire you measure it from.
- Current Rule: The electricity flowing into a junction must equal the electricity flowing out.
- By forcing their math to obey these physical rules, they eliminated the "impossible" answers and narrowed it down to the one true solution.

3. The "Soft Nudge" (Regularization)

Sometimes, the math can get stuck or give weird answers. The authors added a "soft nudge" to the calculation.

The Analogy: Imagine you are trying to balance a stack of plates. You know the stack should be perfectly straight, but the wind is blowing. You don't glue the plates (which would be too rigid), but you gently hold them upright so they don't fall over.
The Paper's Move: They added a "regularization term." This is a mathematical penalty that says, "Hey, the perfect meter should be very close to 1.0 (perfect). If your answer drifts too far from that, we'll add a little penalty." This keeps the solution realistic without forcing it to be perfect.

How They Tested It

The team didn't just write equations; they tested it in two ways:

Simulation: They created a fake power grid (the IEEE 118-bus system) and injected "noise" and "broken meters" to see if their algorithm could find the truth. It worked with incredible precision (errors less than 0.1%).
Real World: They took data from a real power utility in the US. Since they didn't know the "true" answer for the real world, they checked for consistency.
- Did the measurements make sense? Yes.
- Did two different meters measuring the same spot agree after calibration? Yes. Before calibration, they disagreed; after, they matched perfectly.

Why Does This Matter? (The "So What?")

Why should a regular person care? Because this makes the power grid smarter and safer.

Better State Estimation: Power grids use "State Estimation" to know exactly what is happening in real-time (like a weather forecast for electricity). If the inputs (meters) are wrong, the forecast is wrong.
The Result: By fixing the meters and the line parameters simultaneously, the authors showed that the accuracy of the grid's "weather forecast" improved by 24% to 37%.

Summary in One Sentence

This paper teaches computers how to fix broken measuring tools and map the power lines at the same time by using a few perfect reference meters and the unbreakable laws of physics as a guide, making the entire power grid more accurate and reliable without ever having to shut it down.

1. Problem Statement

The paper addresses a critical interdependency problem in power system monitoring using Phasor Measurement Units (PMUs):

The Dilemma: Accurate Line Parameter Estimation (LPE) requires calibrated Instrument Transformers (ITs: Voltage Transformers/VTs and Current Transformers/CTs). Conversely, IT Calibration requires accurate knowledge of line parameters.
Current Limitations:
- Traditional "hard" calibration requires taking ITs offline, which is costly and impractical.
- Existing "online" methods usually assume either that line parameters are known (which is rarely true due to aging and environmental changes) or that ITs are ideal/perfectly calibrated.
- Previous joint estimation attempts (e.g., by Wu et al., Wang et al.) have suffered from sensitivity to noise, large estimation errors, or reliance on unrealistic assumptions (e.g., error-free ITs at one end).
Goal: To develop a framework that simultaneously estimates transmission line parameters ( $r, x, b$ ) and IT correction factors (scaling errors) using PMU data, without requiring offline calibration or assuming perfect instrumentation.

2. Methodology

The authors propose a novel framework called Simultaneous Line Parameter Estimation and IT Calibration (SLIC), solved via a constrained optimization approach.

A. Mathematical Modeling

Noise Model: The paper models PMU measurements as a composite of:
1. Multiplicative Error ( $\eta$ ): Caused by IT scaling deviations (ratio errors), which are slow-varying constants.
2. Additive Noise ( $\Delta$ ): Caused by PMU measurement noise, which is fast-varying and Gaussian.
Non-Linear Formulation (NL-SLIC):
- The authors derive a system of equations based on Kirchhoff's laws for a transmission line.
- Instead of solving a linear system with unknown ratios, they formulate a non-linear optimization problem where the unknowns are the physical line parameters and the IT correction factor ratios (CFRs) relative to a reference.
- The objective function minimizes the residual between the model predictions and the noisy PMU measurements.

B. Constrained Optimization Strategy

To solve the under-determined nature of the problem (more unknowns than equations), the authors introduce two types of constraints based on power system domain knowledge:

Regularization (Soft Constraints):
- The framework assumes the presence of at least one pair of Revenue Quality Meters (RQMs) (high-accuracy ITs) at one end of a "connected tree" of lines.
- A regularization term is added to the objective function to softly penalize deviations of the RQM correction factors from unity (ideal values). This relaxes the "perfect meter" assumption, acknowledging that RQMs have small but non-zero errors.
Equality Constraints (Hard Constraints):
- Voltage Consistency: At any bus connecting two lines, the true voltage is identical regardless of which VT measures it. This creates a constraint linking the VT correction factors of adjacent branches.
- Current Consistency (KCL): The sum of currents at a bus must be zero. This creates a constraint linking the CT correction factors of adjacent branches.
- These constraints allow the calibration of non-RQM branches by propagating information from the RQM branch through the network.

C. Solution Algorithm (NET-SLIC)

The authors employ a Newton's method enhanced with a Trust Region approach (dubbed NET-SLIC).
The algorithm operates in two stages:
1. RQM Branch: Solves the regularized non-linear problem for the branch containing the RQMs.
2. Network Propagation: Iteratively solves for adjacent branch pairs using the equality constraints derived from KCL and KVL, updating the estimates and propagating them across the entire connected tree.
Finally, the Correction Factor Ratios (CFRs) are converted back into absolute IT correction factors.

3. Key Contributions

Novel Formulation: A constrained non-linear optimization framework that simultaneously solves for line parameters and IT calibration without assuming ideal instrumentation.
Domain-Knowledge Integration: The innovative use of physical laws (KVL/KCL) as equality constraints and the statistical properties of RQMs as regularization terms to resolve the interdependency problem.
Robustness to Noise: The method explicitly accounts for both multiplicative (IT) and additive (PMU) noise, demonstrating stability even with realistic noise levels (up to 0.1% TVE).
Scalability: The approach can calibrate an entire network of lines (a "connected tree") starting from a single location with high-quality RQMs.

4. Results and Evaluation

The method was evaluated using the IEEE 118-bus system (345 kV network) and real PMU data from a U.S. utility.

Synthetic Data (Ideal & Noisy):
- Line Parameters: Achieved extremely low estimation errors. In noisy scenarios, Reactance and Susceptance errors were <1%, and Resistance errors were <5% (often <3%).
- IT Calibration: Achieved high accuracy. VT magnitude errors were <0.002%, and CT magnitude errors were <0.02%.
- Comparison: The proposed NET-SLIC significantly outperformed the state-of-the-art method by Wang et al. [17]. While Wang's method showed Resistance errors up to 12.89%, NET-SLIC kept all errors below 2%.
Real PMU Data:
- Applied to 230 kV substation data.
- Consistency: Line parameter estimates were consistent across weekdays, tracking physical changes (e.g., temperature effects).
- Calibration Validation: Before calibration, voltage measurements from different PMUs at the same bus showed significant scaling discrepancies. After SLIC, the measurements aligned closely, validating the calibration accuracy.
Downstream Impact (Linear State Estimation - LSE):
- Performing SLIC improved the accuracy of Linear State Estimation by 24% in magnitude and 37% in angle for the 345 kV network.

5. Significance

This paper provides a practical solution to a long-standing bottleneck in smart grid monitoring. By eliminating the need for offline calibration and removing the circular dependency between line parameters and IT calibration, the proposed NET-SLIC framework enables:

Continuous, real-time calibration of instrument transformers without service interruption.
Higher fidelity data for critical applications like state estimation, fault detection, and dynamic security assessment.
Cost reduction by leveraging existing PMU infrastructure and high-quality revenue meters rather than requiring expensive, frequent manual calibration campaigns.

The work demonstrates that integrating physical domain constraints into data-driven optimization is a powerful strategy for solving complex, interdependent estimation problems in power systems.