No abstract

X-ray computed tomography (CT) is a powerful technique for non-destructive volumetric inspection of objects and is widely used for studying internal structures of a large variety of sample types. The raw data obtained through an X-ray CT practice is a gray-scale 3D array of voxels. This data must undergo a geometric feature extraction process before it can be used for interpretation purposes. Such feature extraction process is conventionally done manually, but with the ever-increasing trend of image data sizes and the interest in identifying more miniature features, automated feature extraction methods are sought. Given the fact that conventional computer-vision-based methods, which attempt to segment images into partitions using techniques such as thresholding, are often only useful for aiding the manual feature extraction process, machine-learning based algorithms are becoming popular to develop fully automated feature extraction processes. Nevertheless, the machine-learning algorithms require a huge pool of labeled data for proper training, which is often unavailable. We propose to address this shortage, through a data synthesis procedure. We will do so by fabricating miniature features, with known geometry, position and orientation on thin silicon wafer layers using a femtosecond laser machining system, followed by stacking these layers to construct a 3D object with internal features, and finally obtaining the X-ray CT image of the resulting 3D object. Given that the exact geometry, position and orientation of the fabricated features are known, the X-ray CT image is inherently labeled and is ready to be used for training the machine learning algorithms for automated feature extraction. Through several examples, we will showcase: (1) the capability of synthesizing features of arbitrary geometries and their corresponding labeled images; and (2) use of the synthesized data for training machine-learning based shape classifiers and features parameter extractors.

In this paper, two active control schemes are presented to improve simultaneously vehicle ride comfort and steady-state handling performance. First, adaptive H∞ controller is designed for nonlinear vehicle suspension systems with actuator time delay based on Genetic Algorithm Wavelet Support Vector Machines and then adaptive H∞ controller is designed based on Genetic Algorithm Mixed Wavelet and RBF Support Vector Machines. The varying sprung and unsprung masses and the suspension performances with actuator delay are taken into account simultaneously, and the corresponding mathematical model is established. The most important feature of the proposed control strategy is its inherent robustness and its ability to handle the nonlinear behaviour of the system. Simulation results show that the designed controllers can achieve good active suspension performance regardless of the variation on the sprung mass in the presence of actuator time delay.

This paper is concerned with adaptive integral sliding mode control (AISMC) based on a wavelet kernel support vector machine for offshore steel jacket platform subject to nonlinear wave-excited force and parameter perturbations. The sliding mode control technique is combined with adaptive control algorithm and wavelet support vector machine to achieve the desired attenuation on the wave-induced vibration and to limit the displacement of offshore platforms. In this method, wavelet kernel support vector machine is used to establish the adaptive controller and an on-line learning rule for the weighting vector and bias is derived. The most important feature of the proposed control strategy is its inherent robustness and its ability to handle the nonlinear behavior of the offshore platform. By means of this control scheme the performance of the AISMC has been improved. The performance of the proposed control strategy is compared with some existing control schemes. It is demonstrated that the proposed control scheme in this paper is more effective in improving the control performance of the offshore platform. This controller is designed based on solving a set of linear matrix inequalities. It has been illustrated through simulation results that the proposed control scheme is effective in improving the control performance of offshore platforms.

No abstract

No abstract

No abstract

Nonspecific high‐energy radiation for treatment of metastatic ovarian cancer is limited by damage to healthy organs, which can be mitigated by the use of radiosensitizers and image‐guided radiotherapy. Gold (Au) and tantalum oxide (TaOx) nanoparticles (NPs), by virtue of their high atomic numbers, find utility in the design of bimetallic NP systems capable of high‐contrast computed tomography (CT) imaging as well as a potential radiosensitizing effect. These two radio‐dense metals are integrated into dendritic mesoporous silica NPs (dMSNs) with radial porous channels for high surface‐area loading of therapeutic agents. This approach results in stable, monodispersed dMSNs with a uniform distribution of Au on the surface and TaOx in the core that exhibits CT attenuation up to seven times greater than iodine or monometallic dMSNs without either TaOx or Au. Tumor targeting is assessed in a metastatic ovarian cancer mouse model. Ex vivo micro‐CT imaging of collected tumors shows that these NPs not only accumulate at tumor sites but also penetrate inside tumor tissues. This study demonstrates that after intraperitoneal administration, rationally designed bimetallic NPs can simultaneously serve as targeted contrast agents for imaging tumors and to enhance radiation therapy in metastatic ovarian cancer.