25 February 2025 in IEEE Transactions on Transportation Electrification
To achieve precise lateral dynamics control for electric and autonomous vehicles under system model uncertainties and unknown external disturbances, this article introduces a novel barrier function (BF)-based neural network (NN) adaptive integral sliding mode control (ISMC) for multiaxle independent steering vehicles (MISVs). Initially, NNs are designed to approximate the unknown parameters associated with uncertainties in the MISV dynamics model. Subsequently, an adaptive ISMC based on BF is developed to ensure robust vehicle state tracking. Unlike traditional adaptive sliding mode controllers, the proposed approach guarantees rapid convergence of variables within a finite time, even without prior knowledge of the upper bounds of unknown external disturbances, thereby significantly reducing vibration and oscillation phenomena. Finally, the Lyapunov stability theory is applied to rigorously demonstrate that the closed-loop system of the MISV remains stable within finite time. The efficacy of the proposed controller is validated through hardware-in-the-loop (HIL) experiments, with results indicating a reduction in maximum tracking errors for the desired yaw rate and sideslip angle by 8.24% and 40.44%, respectively, compared to the conventional control method.
22 January 2025 in IEEE Transactions on Industrial Electronics
The driving range of multiaxle steer-by-wire vehicles is significantly affected by the consumption of vehicle cornering resistance power (VCRP). This article addresses this issue by proposing a novel energy-saving steering control strategy to effectively reduce VCRP by optimizing slip angle. Initially, a general VCRP model is developed to establish a relationship between VCRP and the slip angles of each tire. Later, an adaptive nonsingular terminal sliding mode control method is proposed to balance control energy and state errors. In addition, an energy-saving steering angle optimization method is proposed based on the VCRP model, which achieves an optimal steering angle configuration by optimizing the slip angle. The efficacy of the proposed control strategy in energy-saving is evaluated by hardware-in-the-loop tests. The experimental results exhibit noteworthy reductions of more than 21% and 45% in VCRP under both high-friction and low-friction conditions of high speed, respectively, compared to the traditional Ackerman steering control strategies while maintaining excellent steering stability.
23 January 2025 in IEEE/ASME Transactions on Mechatronics
A pump-controlled electrohydraulic servo steering system (PC-EHSSS) is proposed, to accommodate high energy-efficiency steering control of heavy vehicles; whereas, it's prone to steering-angle overshoot under abrupt signals, resulting from systematic low damping, poor dynamic performance, and unknown disturbances. For overshoot inhibition, barrier Lyapunov functions (BLF), were utilized to strictly restrict high-order steering-angle errors within boundaries. Once contravened, the BLF controller would be invalid, especially in complex road conditions with erratic and variable initial steering angles. For faster and better error convergence under any initial angle, the integral sliding mode control-BLF switching strategy was proposed. ISMC quickly constrained state variables to BLF's control domain before BLF took over. Still, based on the backward-iteration approach, BLF triggered “differential explosion” for its virtual control rates iteration when applied in high-order systems. To enhance its engineering applicability, dynamic surface control (DSC) was incorporated, by converting differentiation into algebraic operations. A closed PC-EHSSS experiment platform for heavy vehicles was constructed. The steering-angle overshoot is restrained within 0.73° under a trapezoidal signal, with 31.68% higher steady-state accuracy than DSC, and 28.04% greater energy efficiency than pump-valve-combined-controlled steering at 0.15 Hz, 3T. A low-overshoot, high-precision, and high-efficiency variable-speed PC-EHSSS for heavy vehicles has been developed.
26 April 2024 in Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering
Heavy-duty vehicles with long bodies, a large number of axles and large loads are subject to increasingly high requirements for precise steering technology due to the increasing trend toward energy conservation and intelligent assisted driving as well as variable driving conditions. In this paper, an energy-efficient open circuit variable-speed pump-controlled steering system (OPCEHSSS) adapted for heavy loads is used, but its strong flow output nonlinearity and system nonlinear dynamic behavior greatly impede the steering performance. Therefore, in order to reduce the influence of the flow leakage of the fixed-displacement pump on the system and to ensure that the flow output of the system matches the control model, a mapping model based on the fitting of a two-layer neural network algorithm with a dynamic real-time compensation strategy (FNC) is proposed. In addition, considering the strong robustness of the system even under parameter uncertainty and unknown disturbance, a complex nonlinear mathematical model is established based on OPCEHSSS physical characteristics, and a dual-objective control strategy of steering angle and pressure based on sliding mode control (SMC) is proposed. However, in order to reduce the influence of high-order switching discontinuity on the steering and ensure the fast convergence of the control system, a fast super twisting algorithm (STA) based on double saturation function of the boundary layer is proposed. The experimental results show that the three different controllers can effectively reduce the steering angle error after the introduction of FNC. And in the case of a single axle loaded with 6 tons, the improved new FNC+STA integrated dual-objective control strategy improves the accuracy by 53.16% compared with PID and 40.67% compared with SMC. The steady-state error is maintained within 0.9°, realizing the high-performance steering tracking control of OPCEHSSS for heavy vehicles.
11 January 2026 in Measurement
16 April 2026 in IEEE Sensors Journal
Accurately estimating the slip angle under dynamic conditions is critical for enhancing vehicle stability and safety. Intelligent tire systems and data-driven deep learning methods have been adopted to slip angle estimation. However, existing approaches often suffer from inadequate temporal feature extraction and limited model–data compatibility. To address these limitations, we perform a temporal dependency analysis of tire acceleration data, revealing strong local and global correlations. Based on this, we propose a novel feature extraction algorithm and a Hierarchical Temporal Fusion Network to fully capture these multiscale temporal correlations. Additionally, an Intelligent Linkage mechanism is designed to enhance model convergence speed and generalization capabilities. Linear probing and principal component analysis further confirm that the network learns highly discriminative and physically continuous representations. Experiments demonstrate that the proposed method achieves 99.38% prediction accuracy with only 11,842 parameters and processes more than 96,000 samples/s, outperforming baseline networks in both accuracy and efficiency. On-road validation confirms its applicability, reaching over 92% agreement on both dry and wet pavements against a kinematic reference.