torquesharing

Torquesharing is a control and mechanical concept in which an overall torque demand is distributed among multiple actuators or drive sources to produce a single, coordinated output. In systems with more than one torque source—such as multi-motor robots, distributed drivetrains, or modular actuation platforms—the goal is to allocate the required torque across available actuators while respecting each device’s limits and capabilities.

The core idea involves solving a sharing problem: given a torque demand T and a set of actuators i with maximum torque capacities τi,max, determine the individual torques τi that sum to T and satisfy 0 ≤ τi ≤ τi,max. A common approach is to introduce sharing coefficients or a cost function that reflects objectives such as minimizing wear, balancing utilization, or maximizing efficiency. In centralized implementations, a solver computes τi in real time; in decentralized schemes, local controllers or simple rules approximate the distribution while maintaining coordination through a common reference or communication.

Control strategies for torquesharing include centralized optimization (often using model predictive control or quadratic programming), gradient-based sharing, and priority-based schemes where certain actuators bear more load under fault or saturation conditions. Robust implementations account for actuator nonlinearities, friction, delays, and measurement noise to maintain stable torque sharing.

Applications appear in electric and hybrid vehicles with multiple drive motors, collaborative robotic systems with several actuators sharing a joint, and modular machinery that benefits from redundancy. Benefits of torquesharing include improved fault tolerance, smoother torque delivery, reduced peak loads, and extended actuator life. Challenges involve coordinating multiple devices, ensuring fast convergence, and managing sensing and communication requirements.