Abstract
In circular $k$-out-of-$n$: G systems, such as multi-rotor drones, operational success depends not only on the number of functioning components but also on the spatial balance of the system. Traditional reliability models rely on restrictive geometric symmetry or proportionality to define balance. This paper proposes a physics-based approach using the Center of Gravity (CoG) as the criterion for balance (Balance Condition III). By demonstrating that BC3 is a generalization of previous conditions, the study shows that systems can remain stable in a wider variety of failure states, thereby significantly increasing the calculated system reliability.
Introduction
The reliability of complex systems like hexacopters or planetary descent vehicles is often modeled using $k$-out-of-$n$: G configurations. In these systems, at least $k$ out of $n$ components must function for the system to succeed. However, for circular systems where units are arranged around a central axis, a simple count of functioning units is insufficient. If the remaining active units create an off-center force, the system may experience catastrophic torque, leading to failure regardless of total thrust.
Historically, "balance" has been defined through strict geometric rules: Balance Condition 1 (BC1), based on rotational symmetry, and Balance Condition 2 (BC2), based on proportionality. This research introduces Balance Condition 3 (BC3), which utilizes the physical principle of the Center of Gravity to identify stable operational states that traditional geometric models overlook.
Defining Balance: From Geometry to Physics
The core contribution of this work is the shift from rigid geometric patterns to a fundamental physical constraint.
- BC1 (Symmetry): Requires functioning units to be arranged in a perfectly symmetrical pattern.
- BC2 (Proportionality): A slightly more flexible rule where units must be balanced proportionally across the circle.
- BC3 (Center of Gravity): A system is considered balanced if the CoG of all functioning units coincides with the geometric origin of the system.
Mathematically, if each unit $i$ is located at $(x_i, y_i)$ on a unit circle, a subset of functioning units is balanced under BC3 if the average of their coordinates is $(0,0)$. The researchers prove an "Inclusion Hierarchy": $BC1 \subset BC2 \subset BC3$. This means any state considered balanced under the old rules is also balanced under BC3, but BC3 identifies additional stable configurations that were previously classified as failures.
Methodology
The study evaluates system reliability using the minimum tie-set method. A tie-set is a group of units whose operation guarantees system success. A minimum tie-set is a configuration where the failure of any single unit within the set would cause the system to fail.
To compare the three balance conditions, the authors performed extensive combinatorial enumerations and numerical studies on systems with $n$ (total units) ranging from 6 to 14. By identifying all possible minimum tie-sets for each condition, they calculated the overall system reliability ($R_s$) based on the reliability of individual units ($r$).
Key Results and Insights
The adoption of the CoG-based definition leads to a marked increase in the number of identified operational states. For example, in a 4-out-of-12 system:
- BC1 identifies 15 minimum tie-sets.
- BC2 identifies 19 minimum tie-sets.
- BC3 identifies 31 minimum tie-sets.
This expansion of "success" states translates directly into higher system reliability. The numerical analysis confirmed that $R_s^{BC3} \geq R_s^{BC2} \geq R_s^{BC1}$ across all tested scenarios.
The "Even" Advantage
A notable finding is the sensitivity of balance to the value of $k$. Systems are significantly easier to balance when $k$ is an even number. When $k$ increases from an even to an odd number, there is a much sharper drop in reliability compared to an increase from odd to even. This is because odd numbers of units are physically less likely to result in a CoG at the origin when arranged on a circular grid.
Reliability Sensitivity
The improvement offered by BC3 is most significant when individual unit reliability ($r$) is in the moderate range (0.5 to 0.7). In scenarios where units are nearly perfect or highly prone to failure, the specific definition of balance has a diminishing impact on the final reliability calculation.
Limitations and Future Work
While BC3 provides a more accurate physical model, it introduces computational challenges. Enumerating all possible states for large $n$ (e.g., swarms) becomes computationally intractable. Future research may focus on Monte Carlo simulations to approximate reliability in larger systems.
Additionally, the current model assumes homogeneity (all units provide identical thrust) and static balance. Real-world applications often involve units with varying performance levels and flight controllers that can dynamically adjust thrust to compensate for minor imbalances. Integrating these dynamic factors into the reliability model remains a key area for future exploration.
Conclusion
By moving from geometric symmetry to a Center of Gravity-based definition of balance, engineers can more accurately model the reliability of circular $k$-out-of-$n$: G systems. This physics-based approach reveals that these systems are more robust than previously estimated, providing a more realistic framework for the design and operation of multi-rotor aerial vehicles and similar circular arrays.
Summary updated