Jun-Cheng Wang

For compound braking systems integrating electro-mechanical brakes (EMBs) and regenerative braking, conventional torque optimization strategies may not fully account for the stage-dependent dynamic characteristics of EMBs during braking. This limitation can lead to suboptimal synchronization between torque allocation and EMB operational dynamics, leading to compromised braking stability and reduced energy recovery efficiency. To address these considerations, this study proposes an optimized torque allocation control framework for distributed-drive electric vehicles (EVs) equipped with EMBs. The framework consists of two core stages, aiming to balance braking performance and energy recovery efficiency. In the primary stage, an interval type-2 fuzzy controller is employed. By adjusting the uncertainty intervals between upper and lower boundaries. This enhancement enables effective adaptation to parameter disturbances under various operating conditions, while improving braking energy recovery efficiency. In the secondary stage, the four-stage torque output characteristics of EMBs are incorporated to dynamically correct the residual braking torque of regenerative braking system in real time. This correction compensates for deviations between the actual EMB braking torque and the target braking torque, ensuring braking stability while maintaining high energy recuperation rates. Simulation results show that under low-velocity/low-intensity and high-velocity/high-intensity braking scenarios, the root mean square error of longitudinal acceleration is reduced by 68.5% and 61.4%, respectively, compared with the strategy without torque optimization. Additionally, under the urban dynamometer driving schedule (UDDS) and worldwide harmonized light vehicle test cycle (WLTC), the average energy recuperation rate is improved by 39.2% and 46.8%, respectively.

The control precision of semi-active suspension systems equipped with magnetorheological (MR) dampers is significantly affected by time-delay effects and the accuracy of mechanical characteristic modeling. To address this, a double-loop control architecture is developed, integrating an outer-loop predictive force control module for time-delay compensation and an inner-loop precise excitation current calculation module. A novel Gaussian process regression (GPR)-based mechanical characterization model is constructed by designing composite covariance functions and optimizing hyperparameters through data-driven methods. For the outer-loop control, an interval type-2 fuzzy Smith time-delay compensation strategy is proposed, which utilizes vehicle body vertical acceleration, suspension deflection, and tire displacement as state inputs to dynamically capture and respond to vehicle motion variations, thereby enhancing system anti-interference capability. The inner-loop control leverages the GPR model for accurate excitation current computation. Simulation results validate that the proposed double-loop controlled MR semi-active suspension system demonstrates superior control efficacy and robustness, effectively adapting to complex and variable driving conditions.

Precisely estimating road adhesion coefficients is essential for enhancing the control of wheel slip ratios. However, traditional neural networks exhibit instability in weight vector updates, resulting in suboptimal estimations that are inadequate for real-time traffic control. In this study, a novel method for predicting road adhesion coefficients utilizing the MFO-Elman neural network is devised to enhance the prediction accuracy. First, the neural network model is constructed based on a vehicle dynamic model. Subsequently, the initial moths are introduced to increase the diversity of the initial population, accelerate the convergence speed, and reduce the number of calculations. Combining the moth flight update mechanism can effectively mitigate the risk of the algorithm converging to a local optimum. Then, the Fourier transform method is employed to estimate the correlation between the optimal slip ratio and the tire-road friction coefficient. Both simulation and experimental findings corroborate the reliability of the newly devised MFO-Elman neural network method for predicting road adhesion coefficients, which yields a higher estimation precision for the road adhesion coefficient.

To effectively suppress the body’s high-frequency vibration and reduce the enormous demanded control force of the electro-hydraulic actuator, a serialized isolator, absorber, and actuator (SIAA) based active suspension (AS) is firstly proposed. The proposed AS comprises a hydro-pneumatic spring, a suspension third mass, and an electro-hydraulic actuator in series from bottom to up. The former two components serve as the isolator and absorber, respectively. Besides revealing influences of the suspension third mass on the high-frequency vibration reduction effect and demanded control force, an extended sliding mode (ESM) observer based extended optimal sliding mode (EOSM) controller is developed. The ESM observer is responsible for the observed values related to the suspension comprehensive performance, and the EOSM controller takes charge of the true nominal optimal performance and robustness. Consequently, by using the SIAA based AS regulated by the specially designed nonlinear optimal robust controller, the peak of the sprung mass acceleration’s power spectral density in the high-frequency region and demanded control force’s root mean square are respectively reduced by 84.68% and 61.71% compared to the single-staged electro-hydraulic AS.

This study presents an innovative paralleled spring and three-function (vibration isolation, absorption, and active reduction) integrated actuator (TFIA)-based active suspension system. This novel design significantly enhanced the high-frequency vibration reduction effect and minimized the demanded control force of rotary-motor-type active suspension. TFIA comprised a bushing spring, a suspension third mass, and a ball screw-motor assembly (BSMA) in series from bottom to up. Besides using the bushing spring to primarily suppress the input of wheel acceleration, TFIA adopted the suspension third mass to further absorb high-frequency vibration to BSMA. Based on the feedforward feedback linearization method, the full-information optimal sliding mode controller was designed by establishing the sliding mode manifold function to include the suspension structure and expected performance information. To balance performance improvement and easy installation, three schemes of structure and control parameters were proposed and compared. Results illustrated that the BSMA self-mass and visual mass scheme is more feasible. The effectiveness of TFIA in suppressing the body's high-frequency vibration and reducing the demanded control force was corroborated by simulation and hardware-in-the-loop experiments.

This research presents an augmented Kalman estimator and an equivalent replacement-based Taylor series (ERBTS)-linear quadratic Gaussian (LQG) control strategy to cope with the control accuracy and response delay of magnetorheological (MR) dampers for vehicle semi-active suspensions. The parameters in the MR model are identified from experimental measurements. Then, two main sources of control error, namely, modelling error and real-time variety of the MR damper output force, are defined as an integrated compound real-time variety. Subsequently, they are written into a differential equation with characteristics of the minimum system to augment the state equation of the semi-active suspension system. The augmented Kalman estimator is constructed to estimate the abovementioned compound real-time variety. To calculate an acceptable time-delay compensation predictive control force, an equivalent operation is implemented beforehand in the suspension comprehensive performance index by replacing a part of the squared time-delay control force with the corresponding predictive control force. Simulation results verify the effectiveness of the proposed augmented Kalman estimator, and the newly developed ERBTS-LQG controller almost achieves control effectiveness of the ideal time delay free semi-active suspension.

Accurate road irregularity model is very important to ensure reliability of vehicle dynamics simulation. Besides that modeling accuracy of power spectral density (PSD) around the lower cut-off frequency (LCOF) is poor, the method real-time generating non-stationary two wheel road irregularities still absents. To solve the above problems, a novel superposition of high-order filtering (HOF) white noises modulated by coherent and incoherent functions based reconstruction method is presented. The HOF white noise model is responsible for modeling accuracy of PSD related to the LCOF and different road irregularity's frequency exponents. The HOF, coherent and incoherent transfer functions are newly developed to modulate three independent standard white noises to generate non-stationary road irregularities online. Three reconstruction applications verify that the PSD and coherent function curves of the reconstructed road irregularities agree with those of the actual ones well respectively.

Control precision and robustness can be regarded as essential challenges of slip ratio control because of the external uncertainties of road conditions and the internal delay response of braking actuators. In addition, it is difficult to obtain a trade-off between braking energy recovery efficiency and braking safety under some extreme braking conditions. To synchronously improve both slip ratio control and braking energy recovery efficiency for different road conditions, an anti-lock braking system (ABS) using a novel interval type-2 fuzzy neural network (IT2FNN) control scheme is proposed for electrohydraulic braking systems. The novel IT2FNN control scheme with five layers is designed to calculate the commanded braking torque. The membership function layer utilizes type-2 fuzzy sets to describe the slip ratio error degree, which enhances the scheme’s anti-interference ability, and the enhanced Karnik-Mendel (EKM) algorithm is used in the type-reduction layer to accelerate computation. Additionally, the network adjusts the parameters of the membership function and rules by decreasing the performance function value corresponding to the braking torque error to improve slip ratio control and enhance the self-adaptation capacity. Simulations showed that IT2FNN control can not only improve slip ratio control performance and decrease the braking torque error but also enhance the energy recovery efficiency of regenerative braking on different road surfaces.