Study on Four Disturbance Observers for FO-LTI Systems

This paper addresses the problem of designing disturbance observer for fractional order linear time invariant U+0028 FO-LTI U+0029 systems, where the disturbance includes time series expansion disturbance and sinusoidal disturbance. On one hand, the reduced order extended state observer U+0028 ROESO U+0029 and reduced order cascade extended state observer U+0028 ROCESO U+0029 are proposed for the case that the system state can be measured directly. On the other hand, the extended state observer U+0028 ESO U+0029 and the cascade extended state observer U+0028 CESO U+0029 are presented for another case when the system state cannot be measured directly. It is shown that combination of ROCESO and CESO can achieve a highly effective observation result. In addition, the way how to tune observer parameters to ensure the stability of the observers and reduce the observation error is presented in this paper. Finally, numerical simulations are given to illustrate the effectiveness of the proposed methods. © 2014 Chinese Association of Automation.

Feature-Oriented Nonfunctional Requirement Analysis for Software Product Line

Domain analysis in software product line （SPL） development provides a basis for core assets design and imple- mentation by a systematic and comprehensive commonality/variability analysis. In feature-oriented SPL methods, products of the domain analysis are domain feature models and corresponding feature decision models to facilitate application-oriented customization. As in requirement analysis for a single system, the domain analysis in the SPL development should con- sider both flmctional and nonfunctional domain requirements. However, the nonfunctional requirements （NFRs） are often neglected in the existing domain analysis methods. In this paper, we propose a context-based method of the NFR analysis for the SPL development. In the method, NFRs are materialized by connecting nonfunctional goals with real-world context, thus NFR elicitation and variability analysis can be performed by context analysis for the whole domain with the assistance of NFR templates and NFR graphs. After the variability analysis, our method integrates both functional and nonfunc- tional perspectives by incorporating the nonfunctional goals and operationalizations into an initial functional feature model. NFR-related constraints are also elicited and integrated. Finally, a decision model with both functional and nonfunctional perspectives is constructed to facilitate application-oriented feature model customization. A computer-aided grading system （CAGS） product line is employed to demonstrate the method throughout the paper.

Robust sliding mode control of general time-varying delay stochastic systems with structural uncertainties

This paper presents a new robust sliding mode control （SMC） method with well-developed theoretical proof for general uncertain time-varying delay stochastic systems with structural uncertainties and the Brownian noise （Wiener process）. The key features of the proposed method are to apply singular value decomposition （SVD） to all structural uncertainties and to introduce adjustable parameters for control design along with the SMC method. It leads to a less-conservative condition for robust stability and a new robust controller for the general uncertain stochastic systems via linear matrix inequality （LMI） forms. The system states are able to reach the SMC switching surface as guaranteed in probability 1. Furthermore, it is theoretically proved that the proposed method with the SVD and adjustable parameters is less conservatism than the method without the SVD. The paper is mainly to provide all strict theoretical proofs for the method and results.

Effective LiDAR Damage Detection:Comparing Two Detection Algorithms

The health conditions of highway bridges is critical for sustained transportation operations.US federal government mandates that all bridges built with public funds are to be inspected visually every two years. There is a growing consensus that additional rapid and non-intrusive methods for bridge damage evaluation are needed.This paper explores the potential of applying ground-based laser scanners for bridge damage evaluation. LiDAR has the potential of providing high-density,full-field surface static imaging.Hence,it can generate volumetric quantification of concrete corrosion or steel erosion.By recording object surface topology,LiDAR can detect different damages on the bridge structure and differentiate damage types according to the surface flatness and smoothness.To determine the effectiveness of LiDAR damage detection,two damage detection algorithms are presented and compared using scans on actual bridge damages.The results demonstrate and validate LiDAR damage quantification,which can be a powerful tool for bridge condition evaluation.

IPatch:a remote adversarial patch

Applications such as autonomous vehicles and medical screening use deep learning models to localize and identify hundreds of objects in a single frame.In the past,it has been shown how an attacker can fool these models by placing an adversarial patch within a scene.However,these patches must be placed in the target location and do not explicitly alter the semantics elsewhere in the image.In this paper,we introduce a new type of adversarial patch which alters a model’s perception of an image’s semantics.These patches can be placed anywhere within an image to change the classification or semantics of locations far from the patch.We call this new class of adversarial examples‘remote adversarial patches’(RAP).We implement our own RAP called IPatch and perform an in-depth analysis on without pixel clipping on image segmentation RAP attacks using five state-of-the-art architectures with eight different encoders on the CamVid street view dataset.Moreover,we demonstrate that the attack can be extended to object recognition models with preliminary results on the popular YOLOv3 model.We found that the patch can change the classification of a remote target region with a success rate of up to 93%on average.