⚡ electrical engineering

Transferring the driveshaft inertia to the grid via the DC-link in MV drive systems

This paper proposes a control strategy for medium-voltage drive systems that synchronously couples driveshaft rotational inertia to the grid to enhance fault ride-through capabilities, while theoretically unifying standard phase-locked loop and matching control approaches as distinct feedback optimization schemes.

Original authors: Catalin Arghir, Pieder Jörg, Silvia Mastellone

Published 2026-06-01

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Catalin Arghir, Pieder Jörg, Silvia Mastellone

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

Imagine a massive industrial machine, like a giant pump or a wind turbine, connected to the electrical grid. Inside this machine, there is a heavy spinning shaft. Because it's heavy and spinning, it has inertia—a physical "stubbornness" that resists changes in speed. If the grid suddenly wobbles, this heavy shaft naturally wants to keep spinning at the same speed, acting like a flywheel to smooth things out.

Usually, this helpful "stubbornness" stays trapped inside the machine. The electrical grid, which is like a nervous system for the whole country, doesn't feel it. The grid has to rely on giant, dedicated power plants to provide that smoothing effect.

This paper proposes a clever trick to teleport that physical stubbornness from the machine's spinning shaft directly to the electrical grid, without building any new heavy equipment.

Here is how they do it, using simple analogies:

1. The "Elastic Band" Connection

Think of the machine's spinning shaft and the electrical system's "DC-link" (a temporary energy storage capacitor) as two separate wheels. Normally, they are just connected by a rigid axle.

The authors propose replacing that rigid axle with a stretchy rubber band.

They tweak the software controlling the machine so that the electrical voltage acts like a second, invisible spinning wheel.
When the grid wobbles, the "rubber band" stretches and pulls on the heavy physical shaft.
Suddenly, the grid "feels" the weight of that massive shaft. The heavy shaft starts helping the grid stabilize itself, just as if the shaft were physically bolted to the power plant.

2. The Two Ways to Drive the Car

To make this rubber band work, the authors tested two different "drivers" (control strategies) to steer the machine:

Driver A (The Traditional Driver): This is like a standard car with a speedometer and a gas pedal. It uses a chain of commands: "Check speed, adjust gas, check voltage, adjust current." It works, but it's a bit slow to react because it has to go through all those steps.
Driver B (The "Matching" Driver): This is a smarter, more intuitive driver. Instead of following a chain of commands, it looks at the road (the grid) and the car's position and instantly adjusts the steering and gas to "match" the rhythm of the road.
- The paper introduces a new way for this driver to "see" the road. Instead of just looking at the angle of the road (like a standard compass), it looks at both the angle and the brightness (voltage magnitude) simultaneously.
- This allows the driver to react much faster and smoother, especially when the road is bumpy or the weather is bad (weak grid conditions).

3. The "Ray-Circle" Trick

The paper uses a mathematical concept called "ray-circle complementarity" to explain how the new driver works.

Imagine trying to walk in a straight line toward a moving target. A standard approach is to just look at the angle.
The new approach is like walking toward a target while also adjusting your stride length based on how far away it is. It treats the movement as a combination of a ray (direction) and a circle (distance/magnitude).
This allows the system to naturally "lock on" to the grid's rhythm, even if the grid is noisy or unstable, without needing a separate, clunky tool (like a traditional Phase-Locked Loop) to find the rhythm.

The Results: A Tougher Machine

The authors tested this on a simulated industrial drive system (like a pumped-hydro power station) under two conditions:

Stiff Grid: A strong, stable electrical network.
Weak Grid: A fragile, noisy network (like a remote area with few power plants).

They subjected the system to sudden shocks:

Phase Jumps: The grid suddenly shifted its timing.
Short Circuits: The grid voltage dropped to zero for a few seconds.
Frequency Steps: The grid speed suddenly sped up or slowed down.

The Findings:

Both methods successfully transferred the heavy shaft's inertia to the grid, helping the grid stay stable.
The new "Matching" driver (Driver B) was faster and smoother than the traditional driver.
Crucially, the new driver didn't get confused or crash when the grid was weak and noisy. It kept the machine running and supported the grid, whereas traditional methods often struggle in these conditions.

In Summary

This paper shows how to turn a heavy, spinning industrial machine into a "super-helper" for the electrical grid. By using a clever software trick to connect the machine's physical weight to the electrical system, and by using a smarter, more intuitive control method, the machine can help stabilize the power grid during emergencies, acting like a giant, invisible flywheel for the entire city.

Technical Summary: Transferring Driveshaft Inertia to the Grid via the DC-Link in MV Drive Systems

Problem Statement
Medium-voltage (MV) drive systems are critical components in high-power industrial processes, electric propulsion, and energy generation. As grid regulations evolve (e.g., IEEE 1547.3, IEC 61400), there is an increasing expectation for these loads to provide ancillary services such as inertia support, phase jump support, and reactive power compensation, traditionally provided by synchronous generators. While significant research exists on motor-side control and grid-side compliance, the dynamic interaction between the mechanical energy stored in the drivetrain and the electrical grid remains underexplored. Conventional grid-forming converters typically emulate inertia using virtual synchronous machine (VSM) models or synthetic inertia loops. However, this paper addresses the challenge of utilizing the physical rotational inertia already present in the driveline shaft, transferring it to the grid interface without introducing virtual inertia, thereby enhancing fault ride-through capabilities and grid stabilization.

Methodology
The authors propose a two-step control approach to organically couple the mechanical driveshaft inertia to the grid through the DC-link capacitor.

Elastic Coupling of Driveshaft to DC-Link:
The first step involves modifying the speed control reference to create a virtual spring-damper element between the driveshaft and the DC-link. By applying a coordinate transformation where the DC-link voltage ( $v_{dc}$ ) is interpreted as the angular velocity of an additional mass element ( $w_{dc}$ ), the system effectively couples the mechanical inertia ( $M_1, M_2$ ) to the electrical capacitance ( $C_{dc}$ ). This transformation renders the total system passive with respect to the grid voltage and load torque, creating a "large" equivalent capacitance at the grid interface within the speed control bandwidth.
Grid-Side Control Strategies:
The paper evaluates two control architectures to induce synchronous machine (SM) behavior on the grid side:
- Modified Cascaded PI Control: The standard cascaded control structure is adapted by drooping the DC-bus voltage reference based on the grid frequency (output of a Phase-Locked Loop, PLL). This creates a coupling where the DC-link voltage tracks the grid frequency, effectively transferring the combined mechanical and electrical inertia to the grid.
- Matching Control: This approach utilizes a model-matching strategy based on the "ray-circle" complementarity. It interprets both standard PLLs and matching control as feedback optimization schemes. The controller directly acts on the modulation vector to minimize the energy distance between the system state and a desired steady-state map (defined by active/reactive power setpoints). This method inherently incorporates synchronization (replacing the need for a separate PLL), DC-link regulation, and current tracking within a single gradient-based controller.

Key Contributions

Physical Inertia Transfer: The paper demonstrates that by elastically coupling the driveshaft to the DC-link and utilizing specific control laws, the physical rotational inertia of the drivetrain becomes dynamically visible at the grid interface. This avoids the need for virtual inertia emulation.
Unified Control Perspective: The authors provide a theoretical framework interpreting standard PLLs and matching control as gradient-based algorithms (Online Feedback Optimization). This perspective reveals that matching control embeds a PLL mechanism and offers inherent current-limiting and tracking capabilities.
Novel PLL Design: A new PLL design is proposed based on decomposing the gradient into tangential and transverse components relative to a circular path (ray-circle decomposition). This allows for the reconstruction of grid voltage magnitude and angle via gradient descent, improving robustness in weak grid scenarios.
Comparative Validation: The study presents a comprehensive simulation benchmark using a high-fidelity model of a pumped-hydro drive system (7 MVA grid-side, 5 kV DC-link, 6 MW driveline). The proposed methods are tested under both stiff (SCR=49) and weak (SCR=1) grid conditions.

Results
The simulation results, covering phase jumps, three-phase short circuits, frequency steps, and voltage dips, indicate the following:

Fault Ride-Through: Both the modified cascaded control and the matching control successfully maintain stability and support the grid during severe disturbances, including phase jumps and short circuits.
Inertia Response: The coupled system exhibits a significantly larger total inertia at low frequencies (within the speed control bandwidth), effectively damping grid frequency oscillations.
Matching Control Performance: The matching control approach demonstrates a faster frequency-to-power response compared to the cascaded approach, suggesting a higher effective DC-link regulation bandwidth. It achieves grid-forming behavior (frequency and voltage support) without a separate outer DC-bus loop or distinct PLL, while maintaining stability in weak grid conditions where cascaded systems might struggle.
Current Limiting: Both approaches successfully adhere to current limits during transients, with the matching control offering a unified mechanism for synchronization and current tracking.

Significance and Claims
The paper claims that this approach offers a "natural, passivity-based" method for interconnected drive systems. By transferring existing mechanical inertia to the electrical domain, the method enhances the grid-supporting capabilities of MV drives without compromising their primary industrial functions. The authors highlight that the matching control framework allows for the complete replacement of the traditional cascaded structure, eliminating the need for separate PLL mechanisms and outer DC-bus loops while avoiding instabilities common in weak grids. The work suggests that this methodology is particularly relevant for complex shaft topologies and future drive system designs requiring robust ancillary service provision. The stability of the proposed systems is supported by time-scale separation arguments and Lyapunov-based analysis in the appendix.