Zhenwei Cao

Robust sliding mode control for Steer-by-Wire systems with AC motors in road vehicles

In this paper, the modeling of steer-by-wire (SbW) systems is further studied, and a sliding mode control scheme for the SbW systems with uncertain dynamics is developed. It is shown that an SbW system, from the steering motor to the steered front wheels, is equivalent to a second-order system. A sliding mode controller can then be designed based on the bound information of uncertain system parameters, uncertain self-aligning torque, and uncertain torque pulsation disturbances, in the sense that not only the strong robustness with respect to large and nonlinear system uncertainties can be obtained but also the front-wheel steering angle can converge to the handwheel reference angle asymptotically. Both the simulation and experimental results are presented in support of the excellent performance and effectiveness of the proposed scheme.

A new robust training algorithm for a class of single-hidden layer feedforward neural networks

Abstract A robust training algorithm for a class of single-hidden layer feedforward neural networks (SLFNs) with linear nodes and an input tapped-delay-line memory is developed in this paper. It is seen that, in order to remove the effects of the input disturbances and reduce both the structural and empirical risks of the SLFN, the input weights of the SLFN are assigned such that the hidden layer of the SLFN performs as a pre-processor, and the output weights are then trained to minimize the weighted sum of the output error squares as well as the weighted sum of the output weight squares. The performance of an SLFN-based signal classifier trained with the proposed robust algorithm is studied in the simulation section to show the effectiveness and efficiency of the new scheme.

Design of Robust Repetitive Control With Time-Varying Sampling Periods

This paper proposes the design of robust repetitive control with time-varying sampling periods. First, it develops a new frequency domain method to design a low-order, stable, robust, and causal IIR repetitive compensator using an optimization method to achieve fast convergence and high tracking accuracy. As such, a new stable and causal repetitive controller can be implemented independently to reduce the design complexity. The comprehensive analysis and comparison study are presented. Then, this paper extends the method to design a robust repetitive controller, which compensates time-varying periodic signals in a known range. A complete series of experiments is successfully carried out to demonstrate the effectiveness of the proposed algorithms.

Adaptive gain sliding mode observer for state of charge estimation based on combined battery equivalent circuit model

An adaptive gain sliding mode observer (AGSMO) for battery state of charge (SOC) estimation based on a combined battery equivalent circuit model (CBECM) is presented. The errors convergences of the AGSMO for the SOC estimation are proved by Lyapunov stability theory. The AGSMO has a capability of compensating modeling errors caused by the parameters variation of the CBECM and minimizing chattering level in SOC estimation. The lithium-polymer battery (LiPB) is used to conduct experiments for extracting the parameters of the CBECM and verifying the effectiveness of the proposed AGSMO for the SOC estimation.

Intelligent Sensorless ABS for In-Wheel Electric Vehicles

The design of electric vehicles (EVs) is increasingly based upon using the in-wheel technology. In this design, the use of a separate electric machine at each corner of the vehicle provides unique opportunities for innovative vehicle control strategies. In this paper, a sensorless antilock braking system (ABS) is proposed that eliminates the need for the installation of separate conventional ABS sensors and saves the costs associated with the installation and maintenance of those sensors for in-wheel EVs. The proposed ABS exploits the information carried by the back electromotive force (EMF) of the electric machines of the in-wheel vehicle to obtain accurate wheel speed estimation at each wheel and conduct road identification simultaneously. A wavelet-packet denoising method is used to maintain the accuracy of wheel speed estimation in the presence of noise. In addition to sensorless wheel speed estimation, the proposed ABS is capable of road identification by analyzing the back EMF signal using discrete wavelet transforms. The design was realized and fully tested using actual ABS hardware. The results of the sensorless technique were compared with a commercial ABS sensor. The experimental results showed that the sensorless ABS can adequately replace the conventional ABS sensor in in-wheel EVs and significantly improve the performance of the ABS.

Robust Single-Hidden Layer Feedforward Network-Based Pattern Classifier

In this paper, a new robust single-hidden layer feedforward network (SLFN)-based pattern classifier is developed. It is shown that the frequency spectrums of the desired feature vectors can be specified in terms of the discrete Fourier transform (DFT) technique. The input weights of the SLFN are then optimized with the regularization theory such that the error between the frequency components of the desired feature vectors and the ones of the feature vectors extracted from the outputs of the hidden layer is minimized. For the linearly separable input patterns, the hidden layer of the SLFN plays the role of removing the effects of the disturbance from the noisy input data and providing the linearly separable feature vectors for the accurate classification. However, for the nonlinearly separable input patterns, the hidden layer is capable of assigning the DFTs of all feature vectors to the desired positions in the frequency-domain such that the separability of all nonlinearly separable patterns are maximized. In addition, the output weights of the SLFN are also optimally designed so that both the empirical and the structural risks are well balanced and minimized in a noisy environment. Two simulation examples are presented to show the excellent performance and effectiveness of the proposed classification scheme.

Robust control for steer-by-wire systems with partially known dynamics

In this paper, a robust control scheme (RCS) for Steer-by-Wire (SbW) systems with partially known dynamics is proposed. It is shown that an SbW system can be represented by a nominal model and an unknown portion. A nominal feedback controller can then be used to stabilize the nominal model and a sliding mode compensator (SMC) is designed to remove the effects of both the unknown system dynamics and uncertain road conditions on the steering performance. For practical consideration, robust exact differentiator (RED) technique is utilized to estimate the derivatives of the position signals for controller design. It is further shown that the designed RCS is able to guarantee a robust steering performance against system and road uncertainties. The comparative experimental studies are given to verify the excellent performance of the proposed RCS for SbW systems.

Adaptive Repetitive Control of System Subject to Periodic Disturbance with Time-Varying Frequency

Repetitive Control (RC) has been widely used to track/reject periodic signal. However, RC alone fails to track any non-periodic reference signal. Another control scheme such as Model Reference Control (MRC) or Model Reference Adaptive Control (MRAC) is required to do such task. MRC is employed when the plant parameters are known, while MRAC is used when the plant parameters are unknown. Therefore, MRC/MRAC needs to be combined with RC in order to simultaneously track any reference signal (not necessarily periodic) and reject the periodic disturbance. The design of RC mostly assumes the constant frequency of disturbance which leads to the selection of a fixed sampling period. In practical, disturbance is possibly time-varying in frequency. The sampling period has to be carefully adjusted in order to keep the number of samples per period remains constant. This sampling period adjustments change the plant parameters. This paper proposes the design of MRAC combined with RC for system subject to periodic disturbance with time-varying frequency. As a preliminary, the design of MRC combined with RC is also discussed here.

An optimal weight learning machine for handwritten digit image recognition

An optimal weight learning machine for a single-hidden layer feedforward network (SLFN) with the application to handwritten digit image recognition is developed in this paper. It is seen that both the input weights and the output weights of the SLFN are globally optimized with the batch learning type of least squares. All feature vectors of the classifier can then be placed at the prescribed positions in the feature space in the sense that the separability of all nonlinearly separable patterns can be maximized, and a high degree of recognition accuracy can be achieved with a small number of hidden nodes in the SLFN. An experiment for the recognition of the handwritten digit image from both the MNIST database and the USPS database is performed to show the excellent performance and effectiveness of the proposed methodology.

Robust Adaptive Sliding-Mode Observer Using RBF Neural Network for Lithium-Ion Battery State of Charge Estimation in Electric Vehicles

This paper presents a robust sliding-mode observer (RSMO) for state-of-charge (SOC) estimation of a lithium-polymer battery (LiPB) in electric vehicles (EVs). A radial basis function (RBF) neural network (NN) is employed to adaptively learn an upper bound of system uncertainty. The switching gain of the RSMO is adjusted based on the learned upper bound to achieve asymptotic error convergence of the SOC estimation. A battery equivalent circuit model (BECM) is constructed for battery modeling, and its BECM is identified in real time by using a forgetting-factor recursive least squares (FFRLS) algorithm. The experiments under the discharge current profiles based on EV driving cycles are conducted on the LiPB to validate the effectiveness and accuracy of the proposed framework for the SOC estimation.