Learning mixtures of point distribution models with the EM algorithm

This paper demonstrates how the EM algorithm can be used for learning and matching mixtures of point distribution models. We make two contributions. First, we show how shape-classes can be learned in an unsupervised manner. We present a fast procedure for training point distribution models using the EM algorithm. Rather than estimating the class means and covariance matrices needed to construct the PDM, the method iteratively refines the eigenvectors of the covariance matrix using a gradient ascent technique. Second, we show how recognition by alignment can be realised by fitting a mixture of linear shape deformations. We evaluate the method on the problem of learning the class-structure and recognising Arabic characters.     

Preparation of an exponentially rising optical pulse for efficient excitation of single atoms in free space

We report on a simple method to prepare optical pulses with exponentially rising envelope on the time scale of a few ns. The scheme is based on the exponential transfer function of a fast transistor, which generates an exponentially rising envelope that is transferred first on a radio frequency carrier, and then on a coherent cw laser beam with an electro-optical phase modulator. The temporally shaped sideband is then extracted with an optical resonator and can be used to efficiently excite a single (87)Rb atom.     

Discretized Markov chain in damage assessment using Rainflow cycle with effects of mean stress on an automobile crankshaft

We studied the effect of mean stress correction factor using the Rainflow counting technique to assess the fatigue damage of an automobile crankshaft under service loading by considering the stochastic process of the Markov chain. The failure of the crankshaft will cause serious damage to the engine and also to other connecting subcomponents. The service loading is computationally generated from the Discrete Markov chain model and the fatigue cycle is counted using the Rainflow counting technique with the consideration of the local minima and maxima load. To quantify the fatigue damage, the strain-life curve using the fatigue mean stresses was used to model the fatigue failure of the material used in for the crankshaft at N _f = 10 ⁶ . The fatigue mean stresses were used to estimate the effects of the mean stress on the fatigue strength of the component under service loading condition. Statistical verification with the boundary condition of the 90% confidence level was performed to observe the difference between the stochastic algorithms when compared towards the fatigue life behavior of the ductile cast iron material. We concluded that for the practical application, the proposed stochastic model provides a highly accurate assessment of fatigue damage prediction for improving the safety and controlling the risk factors in terms of structural health monitoring.     

A non-contact method for part-based process performance monitoring in end milling operations

Surface response to excitation (SuRE) method was originally developed for structural health monitoring (SHM) applications. SuRE was used to evaluate the performance of completed milling operations. The method generates surface waves on the plate and studies the spectrum changes at selected points to detect defects and change of compressive forces. In this study, the length, depth, and width of a slot were changed step by step. The surface of the aluminum plate was excited in the 20–400 kHz range with a piezoelectric element. A laser scanning vibrometer was used to monitor the vibrations at the predetermined grid points after the dimensions of the slot were changed methodically. The frequency spectrums of measured vibrations were calculated by using the Fast Fourier Transformation (FFT). The sums of the squares of the differences (SSD) of the spectrums were calculated to evaluate the change of the spectrums. The SuRE method was able to determine if the dimensions were changed in each case at all the selected points. The scanning laser vibrometer is not feasible to be used at the shop floor. However, the study demonstrated that a piezoelectric element attached to any of the grid points would be able to evaluate the completed machining process.     

Predicting surface quality of γ-TiAl produced by additive manufacturing process using response surface method

Electron beam melting (EBM) has been found to be a promising technology for producing complex shaped parts from gamma titanium aluminide alloys (γ-TiAl). The parts produced by this process are projected to have dimensions very close to the desired final shapes. However, the surface roughness of the parts produced by EBM is excessively rough. In many applications, it is necessary to improve the quality of manufactured parts using a convenient post process. This paper determines process parameters of end milling when it is used as a post process for the parts produced by EBM. Design of experiments has been used to study the effect of the selected input parameters of end milling (spindle speed, feed rate, depth of cut and coolant type) on the surface roughness of γ-TiAl parts. Response surface methodology is used to develop a predictive model for surface roughness. Effects of the selected milling process are investigated. This paper also optimizes the selected process parameters to minimize the value of the obtained surface roughness.     

A new design of magnetorheological fluid based braking system using genetic algorithm optimization

This paper proposes a new design for an automotive magnetorheological (MR) braking system using multiple rotary disks. We develop an analytical model to calculate the torque and validate our results using finite element analysis considering a non-linear relationship between magnetic field and magnetic flux. Using genetic algorithm, we optimize the system’s dimensions to generate maximum torque. The optimized geometry shows an improvement in output torque compared to existing systems. Moreover, our design directs higher flux onto the MR fluid and subsequently generates greater shear frictions. This design can be used in applications such as rehabilitation devices.     

Identifying critical autonomous systems in the Internet

The Internet not only facilitates our daily activities, such as communication, entertainment and shopping but also serves as the enabling technology for many critical services, including finance, manufacturing, healthcare and transportation. On the other hand, a wide spectrum of attacks targets its communication infrastructure to disable or disrupt the network connectivity and traffic flow until recovery processes take place. Attacking all autonomous systems (ASes) in the Internet is typically beyond the capability of an adversary. Therefore, targeting a small number of ASes which results in the highest impact is the best strategy for attackers. Similarly, it is important for network practitioners to identify, fortify and secure those critical ASes to mitigate the impact of the attacks. In this study we introduce an intuitive and effective measure, IP address spatial path stress centrality, to assess and identify the critical ASes in the Internet. We compare IP address spatial path stress centrality to the three well-known and widely used centrality measures, namely customer-cone size, node degree and betweenness. We demonstrate that the proposed measure incorporates business relations and IP address spaces to achieve a better measure for identifying the critical ASes in the Internet.