A systematic review published in Discover Vehicles examines the control strategies developed for Series Active Variable Geometry Suspension, a mechatronic suspension architecture in which an electric actuator sits in series with a passive spring-damper assembly. The review, authored by Masoud Masih-Tehrani and covering peer-reviewed literature from 2014 to 2025, categorizes the full range of control approaches investigated for the technology, from robust H-infinity synthesis to sliding mode, optimal, and predictive methods, and traces how the field has progressed from simulation studies to full-vehicle prototype road testing.
The SAVGS concept differs fundamentally from conventional active suspensions. Instead of an actuator working in parallel with the spring to generate suspension force directly, the series arrangement modulates the effective motion ratio between wheel travel and spring-damper compression through geometric variation of the suspension linkage. Because the actuator does not bear the static vehicle weight, its sizing requirements shrink considerably, and the passive spring continues to support the load even if the actuator fails, providing inherent fail-safe behavior. The review reports typical actuator power consumption of roughly 134 to 200 watts during normal driving, in contrast to the multi-kilowatt draw described for conventional active systems.
According to the review, H-infinity robust control has emerged as the predominant methodology. The framework minimizes the worst-case gain from road disturbance inputs to performance outputs such as sprung mass acceleration, suspension deflection, and tire deflection, using frequency-dependent weighting functions that encode the trade-offs between ride comfort and road holding.
The strongest experimental evidence comes from a full-car SAVGS prototype developed at Imperial College London, which combined an outer-loop H-infinity controller with an inner-loop velocity tracking controller for actuator command following. Road testing on ISO Class D random road profiles demonstrated an average 3 dB attenuation, approximately a 30 percent reduction, in RMS sprung mass acceleration relative to the passive baseline. The review emphasizes that this figure is an individual-study result rather than a pooled estimate, and that quarter-car experiments provide intermediate validation under controlled inputs.
Beyond robust control, the review catalogs genetic algorithm-tuned H-infinity controllers, which automate the selection of weighting function parameters and have shown reported improvements of 15 to 25 percent in roll acceleration reduction in simulation studies focused on cornering on rough roads. Sliding mode control variants, including fuzzy and proportional-integral formulations, offer competitive performance with simpler implementation, though chattering remains the classical challenge. A triple fuzzy sliding mode controller achieved the best performance in 53 percent of on-road tests and 73 percent of off-road tests in the cited simulation work, improving vehicle angular accelerations by roughly 30 percent relative to the passive state.
Linear quadratic regulator approaches and model predictive control round out the methodology landscape. LQR provides an analytically tractable state-feedback solution but lacks the explicit robustness guarantees of H-infinity methods, and its weight selection requires iterative tuning. MPC is attractive because it can enforce suspension-travel, rocker-angle, torque, and rate constraints while incorporating road-preview information, and an LMI-based robust MPC architecture has been reported for a quarter-car SAVGS model, but its evidence remains simulation-based, with rig validation and deterministic execution time still open requirements.
The review is explicit about validation gaps. Linear parameter-varying control, adaptive approaches, reinforcement learning, adaptive dynamic programming, learning-based MPC, physics-informed machine learning, digital twins, integrated chassis control, and cloud-connected suspension concepts are all positioned as prospective directions rather than established SAVGS results, since none has been validated on SAVGS hardware as of the July 2025 targeted literature update. Transferable evidence from semi-active suspension studies, including cloud-scheduled LPV controllers and preview control with actuator delay, motivates these directions but cannot be interpreted as SAVGS validation.
Applications beyond passenger vehicles receive similarly cautious treatment. Military-vehicle testing of magnetorheological semi-active suspension on the Stryker platform and railway secondary-suspension research with semi-active dampers provide transferable evidence that controllable suspension can improve ride quality in those domains, but no military or railway SAVGS prototype validation was identified. The review also notes that the most mature full-vehicle evidence originates from a single research program, and that independent replication, common benchmark models, and standardized road profiles are needed before the strongest reported results can be generalized.
To address comparability problems, the review proposes a standardized benchmarking framework specifying vehicle and suspension parameters including actuator limits, standardized road and maneuver inputs, performance indices covering body acceleration, dynamic tire load, suspension travel, attitude response, control effort, and energy use, and an evaluation procedure addressing sampling rate, estimator assumptions, and uncertainty reporting. The framework is offered as a reporting structure rather than a validated dataset, intended to let simulation and hardware studies be compared without reconstructing undocumented assumptions.
The review identifies its own limitations, including English-language restrictions, single-author screening without inter-rater statistics, heterogeneous performance reporting across studies, and possible publication bias favoring positive results. Its central conclusions are that H-infinity control has reached full-vehicle prototype demonstration, that GA-tuned and sliding mode methods remain primarily at simulation proof-of-concept level, and that adaptive control mechanisms, machine learning integration, sensor fusion for autonomous vehicles, connected vehicle applications, and standardized benchmarking constitute the principal research gaps that must be closed before SAVGS technology can move toward production implementation.
Subject of Research: Control strategies for series active variable geometry suspension systems
Article Title: Control strategies for series active variable geometry suspension systems: a systematic review
Article References: Masih-Tehrani, M. (2026). Control strategies for series active variable geometry suspension systems: a systematic review. Discover Vehicles, 2(1), Article 42. https://doi.org/10.1007/s44465-026-00045-w
Image Credits: AI Generated
DOI: 10.1007/s44465-026-00045-w
Keywords: SAVGS, active suspension, H-infinity control, sliding mode control, vehicle dynamics, model predictive control, ride comfort, road holding, systematic review, automotive engineering, Control, strategies
News Source: Denise Maddox. (October 29, 2026). Suspension control review maps a decade of SAVGS research. Scienmag.
