Robust equalization of multi-lane electromechanical actuators: Comparative analysis of torque-summed and velocity-summed architectures

The increasing demand for energy-efficient, low-emission aircraft has accelerated the transition toward all-electric aircraft (AEA), driving the need for highly reliable electromechanical actuators (EMAs) in flight-critical systems. A major challenge in multi-lane EMA configurations lies in torque and velocity disparities caused by component mismatches, sensor drift, and mechanical tolerances. These discrepancies can result in force-fight, increased mechanical stress, and reduced control accuracy, necessitating robust equalisation strategies. This study investigates force equalisation and lane equalisation in torque-summed and velocity-summed EMA architectures to mitigate lane-torque disparities (ΔT) and enhance system robustness. A four-lane redundant EMA system, originally designed to actuate the inboard aileron of the Sea Harrier aircraft, was modelled in MATLAB–Simulink with a proportional–integral–derivative (PID) controller and Monitoring–Voting–Averaging Devices (MVADs) for feedback processing. Three-phase motor models were included to capture torque-ripple effects, and simulations were conducted across a range of inertial and aerodynamic load conditions. The results show that force equalisation effectively reduces lane-torque disparities in torque-summed architectures, improving load sharing and resilience under tachometer and potentiometer drift. In contrast, the velocity-summed architecture is inherently torque-balanced (lane-torque disparity ΔT is approximately zero), so lane equalisation has limited effect on ΔT. A potentiometer bias appears as a steady tracking offset that equalisation alone does not remove. These findings highlight critical trade-offs between architectures and underscore the importance of advanced equalisation and adaptive control strategies to further optimise EMA performance in next-generation all-electric aircraft. Although framed by an aerospace EMA, the methods and conclusions are not domain specific. The closed-form equalisation laws, cross-monitoring, and analysis of sensor-bias mechanisms generalise to multi-lane system architectures in robotics, automotive drive-by-wire, industrial automation, wind-turbine blade-pitch control, marine actuation, and medical and assistive devices.