An Explicit Formulation for Mobility Analysis and Dynamics of General Multibody Systems
This paper proposes a universal multibody system description method based on four marker points to derive an explicit formulation for mobility analysis and Lagrangian dynamics, enabling a general dynamic simulation algorithm applicable to arbitrary serial, parallel, and hybrid mechanisms.
Original paper licensed under CC BY 4.0 (https://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
Robots are everywhere, from the arms assembling cars on factory floors to the delicate instruments helping surgeons operate. At their core, these machines are collections of solid parts connected by joints, a field of engineering known as multibody dynamics. To make a robot move correctly, engineers must first understand how many ways it can move, a concept called mobility, and then calculate exactly how forces like gravity or motors will affect that motion over time. For decades, creating these calculations has been a fragmented process. Engineers often had to build a unique mathematical model for every new type of robot, whether it was a simple arm, a complex parallel structure, or a hybrid mix. This meant that inventing a new machine required starting the math from scratch, slowing down the development of new technologies.
A team of researchers at HiSilicon in Shanghai has proposed a new way to handle these calculations that treats all robot mechanisms as variations of a single, universal system. Instead of trying to fit every new machine into a specific category, they developed a method that describes any collection of moving parts using a single, consistent language. Their work, published as a research article, offers a unified algorithm capable of analyzing the movement and forces of any robot, no matter how complex or unusual its design. By standardizing how the computer sees the machine, they have created a tool that can simulate the behavior of almost any multibody system without needing a custom formula for each one.
The heart of this new approach is a clever way of describing the position and orientation of a solid object. In traditional methods, engineers often use a mix of coordinates and angles to track where a part is and how it is turned. The researchers instead attached four imaginary points to every single moving part of the robot. These points act like markers that define the shape and location of the object in space. By tracking the movement of these four points, the computer can determine the exact position and speed of the entire body. This method turns a complex problem of rotation and translation into a straightforward calculation of how these points move relative to one another. It is similar to how a photographer might track a moving car by following four specific lights on its bumper rather than trying to calculate the car's center of gravity and wheel angles simultaneously.
Using this four-point system, the researchers built a framework that can describe any connection between two parts, whether it is a hinge, a sliding rail, or a ball joint. In the past, different types of joints required different mathematical rules, making it difficult to combine them in a single simulation. Here, the team showed that every constraint can be written in the same format. They demonstrated that by adding these constraints one by one to a set of free-floating parts, the computer can automatically figure out which movements are possible and which are blocked. This process generates a "mobility matrix," a mathematical map that reveals exactly how the entire machine can move. The beauty of this formulation is that it works for serial robots, parallel robots, and complex hybrids without changing the underlying code.
Once the movement possibilities are mapped, the team applied the same logic to the forces acting on the robot. They translated the laws of physics, specifically the equations that govern how energy and motion interact, into this new four-point language. This allowed them to create a general algorithm that can simulate the dynamic behavior of a robot over time. The computer can now predict how a robot will swing, fall, or lift a load, accounting for gravity, inertia, and external pushes. To prove this works, the researchers built a software tool in the Python programming language. This software allows users to define a robot, set its physical properties, and run a simulation without writing complex mathematical scripts.
The team tested their software with a classic physics problem: a double pendulum. This device consists of two rods connected end-to-end, with the top rod fixed to a wall. When released, the rods swing in a chaotic and unpredictable pattern that is notoriously difficult to calculate. The researchers simulated this motion using their new method and compared the results with those from a well-known, open-source simulation package called Exudyn. The two sets of results matched almost perfectly. The trajectory of the pendulum's end point, the speed of its swing, and the forces involved were identical in both simulations. This successful test confirmed that their universal formulation is not just a theoretical idea but a practical tool that produces accurate, reliable data.
The implications of this work extend beyond just making simulations faster. By providing a single, explicit formula for analyzing mobility and dynamics, the researchers have removed a major barrier to designing new types of robots. Engineers can now experiment with novel mechanisms, such as robots that combine sliding and rotating parts in unusual ways, without worrying that their mathematical models will fail. The software they developed serves as a verification platform, allowing designers to build and test virtual prototypes before ever cutting metal. While the current version focuses on accuracy and universality, the authors note that future work will aim to make the calculations even faster, particularly by streamlining the steps needed to find the allowed movements. For now, this new formulation offers a clear, unified path forward for understanding the motion of machines, turning a fragmented landscape of specialized math into a single, coherent field of study.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.