⚡ electrical engineering

Design, Modelling, and Control of Magnetic Ball Suspension System

This paper presents the modeling, stability analysis, and comparative evaluation of pole placement, observer-based, and LQR control strategies for a nonlinear Magnetic Ball Suspension System, demonstrating through Simulink simulations that while linearized models achieve stable performance with minimal oscillations, observer-based and LQR approaches effectively manage the system's inherent instability and transient dynamics.

Original authors: Sampson E. Nwachukwu

Published 2026-01-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Sampson E. Nwachukwu

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 you are trying to balance a steel ball in mid-air, hovering just below a powerful magnet, without letting it touch anything. This is the Magnetic Ball Suspension System (MBSS). It's like trying to keep a feather floating in a strong wind tunnel; if you let go for a split second, the ball either crashes into the magnet or drops to the floor. The goal is to keep it perfectly still in the middle, using only invisible magnetic forces, which means no friction and no wear and tear.

This paper is essentially a guidebook on how to build the "brain" (the controller) that keeps this ball from falling. Here is the breakdown of their journey:

1. The Problem: A Wobbly, Unpredictable Ride

The system is tricky because it's nonlinear. Think of it like driving a car where the steering wheel gets heavier or lighter depending on how fast you go, and the brakes work differently if you turn left versus right. Because of this, the math gets messy. The ball wants to fall due to gravity, but the magnet pulls it up. If the magnet pulls too hard, the ball snaps up; too weak, and it drops.

2. The Map: Drawing the Rules

First, the authors built a mathematical map (a state-space model) of the system. They broke it down into two parts:

The Mechanical Part: How the ball moves (gravity pulling down, magnet pulling up).
The Electrical Part: How the electricity flows through the magnet's coil to create that pull.

They calculated the "sweet spot" (equilibrium point) where the ball should hang. For their specific setup, this was about 6 centimeters below the magnet.

3. The Strategy: Three Ways to Steer the Ship

Since the system is unstable, they needed a controller to constantly nudge the ball back to the center. They tested three different "drivers":

Driver A: The Pole Placer (State Feedback)
Imagine you have a car with a steering wheel that you can lock into specific angles to make the car drive straight. This method calculates exactly how much to push the magnet to keep the ball steady. They "placed the poles" (a math term for setting the system's stability speed) to make the ball settle down quickly without wobbling too much.
- The Result: It worked great on the "ideal" math version of the system. On the real, messy nonlinear version, the ball wobbled a bit more before settling, but it still worked.
Driver B: The Guess-Work Expert (Full-Order Observer)
In the real world, you can't always measure every single thing happening inside the machine (like the exact speed of the ball at every millisecond). It's like trying to drive a car with a foggy windshield where you can only see the road directly in front of you.
This controller acts like a super-smart guesser. It uses the limited data it can see (the ball's position) to estimate the rest (how fast it's moving, how much current is flowing). It then uses those guesses to steer the ball.
- The Result: It was very accurate. Even when the system was nonlinear, the "guessing" helped stabilize the ball, though it sometimes took a little longer to settle than the perfect math model.
Driver C: The Efficiency Expert (LQR - Linear Quadratic Regulator)
This driver cares about two things: keeping the ball in the center and not wasting energy. It's like a driver who wants to get to the destination smoothly but also wants to save gas. It uses a complex formula to find the "perfect balance" between how hard it pushes the magnet and how well the ball stays put.
- The Result: It performed very similarly to Driver A. It was robust and efficient, offering a great balance of stability and energy use.

4. The Race Results: Simulations

The authors ran these three drivers through a computer simulation (a virtual test track) to see how they performed.

The "Perfect World" (Linearized Model): In the simplified math version, all three drivers were champions. The ball stabilized quickly with very little shaking.
The "Real World" (Nonlinear Model): When they tested the actual, messy physics, the ball had to do a little dance (oscillate) before it finally stopped moving. It took a bit longer to calm down, but all three methods eventually got the job done.

The Bottom Line

The paper concludes that while the "perfect world" math models are easy to control, the real, wobbly system needs a bit more patience. The Observer method is crucial because, in real life, we can't measure everything perfectly, so we need a way to "guess" the missing pieces to keep the ball safe.

Ultimately, they found that if you want the most stable control, you should aim for the smallest possible equilibrium point (keeping the ball closer to the magnet), as this makes the system easier to manage. The study proves that with the right mathematical "brain," you can keep a steel ball floating in mid-air without it ever touching a thing.

Technical Summary: Design, Modelling, and Control of a Magnetic Ball Suspension System (MBSS)

Problem Statement
The paper addresses the challenges associated with the Magnetic Ball Suspension System (MBSS), a nonlinear and inherently unstable electromechanical system used in precision applications such as wind turbines and micro-robot actuation. The primary objective of the MBSS is to levitate a steel ball using electromagnetic force without physical contact, thereby eliminating frictional losses. However, the system's design and analysis are complicated by its innate nonlinearities and the volatility of input signals, which render standard linear control concepts effective only within limited operating ranges. The core problem is to develop a robust control strategy that can stabilize the ball's position despite these nonlinear dynamics and the practical inability to measure all state variables directly.

Methodology
The study employs a systematic approach involving mathematical modeling, linearization, and the design of three distinct control strategies:

Mathematical Modeling: A comprehensive state-space model was developed to capture both the mechanical and electrical dynamics. The mechanical part is modeled using Newton's second law, while the electrical part utilizes Kirchhoff's Voltage Law (KVL). The system is defined by three state variables: ball position ( $x_1$ ), ball velocity ( $x_2$ ), and coil current ( $x_3$ ).
Linearization and Analysis: Equilibrium points were determined using feedback linearization based on the Jacobian matrix. The system was linearized around a specific operating point (calculated with $E=8V$ , $R=10\Omega$ , $K=0.01$ , $M=0.2kg$, $L=0.5H$ ). Controllability and observability analyses were conducted, confirming that the system is both controllable and observable, which validates the feasibility of state feedback and observer-based control.
Control Design: Three control methods were implemented and simulated in Simulink for both linearized and nonlinear models:
- Pole Placement State Feedback: Designed to move system poles to desired locations (-5, -10, -20) to ensure stability.
- Full-Order Observer: Developed to estimate state variables when direct measurement is impractical, utilizing the duality principle between controllability and observability.
- Linear Quadratic Regulator (LQR): An optimal control strategy designed to minimize a quadratic cost function, balancing state regulation and control effort.

Key Results
Simulation results over a 50-second duration revealed distinct performance characteristics between linearized and nonlinear models:

Linearized Systems: All three control strategies (state feedback, observer-based, and LQR) consistently achieved the desired performance, stabilizing the system to the equilibrium point with minimal oscillations.
Nonlinear Systems: The nonlinear models exhibited significant transient oscillations before stabilization.
- The State Feedback control required more time and exhibited several oscillations due to the higher control effort needed to drive the ball to equilibrium.
- The Full-Order Observer successfully estimated states, enabling effective control. However, performance varied with nominal values; using a nominal value of 0.06 resulted in minor oscillations, while a value of 0.8 led to increased oscillations in the output waveforms.
- The LQR-based control offered improved robustness and minimized control effort. Its performance was found to be comparable to standard state feedback in several cases.

Significance and Claims
The paper claims that the proposed methodology successfully addresses the instability of the MBSS by accurately determining equilibrium points and applying robust control techniques. The study demonstrates that while linearized systems perform satisfactorily regardless of initial conditions, the inherent nonlinearity of the MBSS necessitates careful handling, as it leads to oscillations before settling.

The authors conclude that for more stable control of the system, it is recommended to utilize the smallest equilibrium point. The work validates that state-space modeling, combined with feedback linearization and advanced control strategies like LQR and full-order observers, provides a viable framework for managing the complex dynamics of magnetic levitation systems.

1. The Problem: A Wobbly, Unpredictable Ride

2. The Map: Drawing the Rules

3. The Strategy: Three Ways to Steer the Ship

4. The Race Results: Simulations

The Bottom Line

More like this