SiC单晶片切割力建模与自适应控制

Si C单晶因优良的物理和机械性能而大量用于大功率器件和IC行业。但由于材料的高硬度和高脆性,使其加工过程变得很困难。为此,分析了Si C单晶片切割过程,建立切割过程模型,通过F检验法进行系统阶次辨识,采用遗忘因子递推最小二乘算法在线估计模型参数,建立进给量与切割力的差分方程,设计基于最小方差自校正的切割力控制器,并进行实验验证。结果表明:控制器能够很好的跟踪不同信号,具有良好的鲁棒性,提高了Si C单晶片的加工效率和表面质量。     

A Comprehensive Review of Genetically Engineered Mouse Models for Prader-Willi Syndrome Research

Prader-Willi syndrome (PWS) is a neurogenetic multifactorial disorder caused by the deletion or inactivation of paternally imprinted genes on human chromosome 15q11-q13. The affected homologous locus is on mouse chromosome 7C. The positional conservation and organization of genes including the imprinting pattern between mice and men implies similar physiological functions of this locus. Therefore, considerable efforts to recreate the pathogenesis of PWS have been accomplished in mouse models. We provide a summary of different mouse models that were generated for the analysis of PWS and discuss their impact on our current understanding of corresponding genes, their putative functions and the pathogenesis of PWS. Murine models of PWS unveiled the contribution of each affected gene to this multi-facetted disease, and also enabled the establishment of the minimal critical genomic region (PWScr) responsible for core symptoms, highlighting the importance of non-protein coding genes in the PWS locus. Although the underlying disease-causing mechanisms of PWS remain widely unresolved and existing mouse models do not fully capture the entire spectrum of the human PWS disorder, continuous improvements of genetically engineered mouse models have proven to be very powerful and valuable tools in PWS research.     

Determination and correlation of temperature and pH value dependent solubility of dl-methionine

In order to provide basic data for the crystallization of dl -methionine and understand the thermodynamic properties of dissolution, the solubility of dl -methionine in water was measured by gravimetric method from 294.05 to 347.05 K and the solubility at different pH values under temperature 302.95 to 342.05 K was determined by elemental analysis in this work. The experimental data show that the solubility of dl -methionine increases with increasing temperature as well as more acidic or basic pH values. The van't Hoff equation, modified Apelblat equation, λh equation, Wilson model, NRTL model, and UNIQUAC model were used to correlate the solubility data versus temperature. The melting temperature, enthalpy of fusion, enthalpy and entropy of dissolution were estimated from solubility data and it shows that the expression of solubility by enthalpy and entropy of dissolution gives more reasonable estimation and Wilson model gives a better correlation for these thermodynamic properties.     

Soft sensing of water depth in combined sewers using LSTM neural networks with missing observations

Information and communication technologies combined with in-situ sensors are increasingly being used in the management of urban drainage systems. The large amount of data collected in these systems can be used to train a data-driven soft sensor, which can supplement the physical sensor. Artificial Neural Networks have long been used for time series forecasting given their ability to recognize patterns in the data. Long Short-Term Memory (LSTM) neural networks are equipped with memory gates to help them learn time dependencies in a data series and have been proven to outperform other type of networks in predicting water levels in urban drainage systems. When used for soft sensing, neural networks typically receive antecedent observations as input, as these are good predictors of the current value. However, the antecedent observations may be missing due to transmission errors or deemed anomalous due to errors that are not easily explained. This study quantifies and compares the predictive accuracy of LSTM networks in scenarios of limited or missing antecedent observations. We applied these scenarios to an 11-month observation series from a combined sewer overflow chamber in Copenhagen, Denmark. We observed that i) LSTM predictions generally displayed large variability across training runs, which may be reduced by improving the selection of hyperparameters (non-trainable parameters); ii) when the most recent observations were known, adding information on the past did not improve the prediction accuracy; iii) when gaps were introduced in the antecedent water depth observations, LSTM networks were capable of compensating for the missing information with the other available input features (time of the day and rainfall intensity); iv) LSTM networks trained without antecedent water depth observations yielded larger prediction errors, but still comparable with other scenarios and captured both dry and wet weather behaviors. Therefore, we concluded that LSTM neural network may be trained to act as soft sensors in urban drainage systems even when observations from the physical sensors are missing.     

Style-Controllable Speech-Driven Gesture Synthesis Using Normalising Flows

Automatic synthesis of realistic gestures promises to transform the fields of animation, avatars and communicative agents. In off-line applications, novel tools can alter the role of an animator to that of a director, who provides only high-level input for the desired animation; a learned network then translates these instructions into an appropriate sequence of body poses. In interactive scenarios, systems for generating natural animations on the fly are key to achieving believable and relatable characters. In this paper we address some of the core issues towards these ends. By adapting a deep learning-based motion synthesis method called MoGlow, we propose a new generative model for generating state-of-the-art realistic speech-driven gesticulation. Owing to the probabilistic nature of the approach, our model can produce a battery of different, yet plausible, gestures given the same input speech signal. Just like humans, this gives a rich natural variation of motion. We additionally demonstrate the ability to exert directorial control over the output style, such as gesture level, speed, symmetry and spacial extent. Such control can be leveraged to convey a desired character personality or mood. We achieve all this without any manual annotation of the data. User studies evaluating upper-body gesticulation confirm that the generated motions are natural and well match the input speech. Our method scores above all prior systems and baselines on these measures, and comes close to the ratings of the original recorded motions. We furthermore find that we can accurately control gesticulation styles without unnecessarily compromising perceived naturalness. Finally, we also demonstrate an application of the same method to full-body gesticulation, including the synthesis of stepping motion and stance.     

A numerical filtering method for linear state-space models with Markov switching

A class of discrete-time random processes arising in engineering and econometrics applications consists of a linear state-space model whose parameters are modulated by the state of a finite-state Markov chain. Typical filtering approaches are collapsing methods, which approximate filtered distributions by mixtures of Gaussians, each Gaussian corresponding to one possibility of the recent history of the Markov chain, and particle methods. This article presents an alternative approach to filtering these processes based on keeping track of the values of the underlying filtered density and its characteristic function on grids. We prove that it has favorable convergence properties under certain assumptions. On the other hand, as a grid method, it suffers from the curse of dimensionality, and so is only suitable for low-dimensional systems. We compare our method to collapsing filters and a particle filter with examples, and find that it can outperform them on 1- and 2-dimensional problems, but loses its speed advantage on 3-dimensional systems. Meanwhile, our method has a proven theoretical convergence rate that is probably not achieved by collapsing and particle methods.     

Large lakes in climate models: A Great Lakes case study on the usability of CMIP5

Large lakes have an impact on regional weather. In addition, they can be both sensitive to and influence regional climate changes. In the climate models that are used to investigate future climate changes, lakes are greatly simplified and sometimes absent. At the regional scale, this can have strong implications for the quality of the model information about the future. Through our work with climate information users in the Laurentian Great Lakes region, we have found that basic credibility of the information requires the underlying climate models simulate lake-atmosphere-land interactions. We are not aware of efforts within the scientific community to make known how individual large lakes are represented in models and how those representations translate to the quality of the data for particular regions. We share our framework for identifying how the Laurentian Great Lakes are represented in the Coupled Model Intercomparison Project (CMIP) version 5 climate models. We found that most CMIP5 models do not simulate the Great Lakes in a way that captures their impact on the regional climate, which is a credibility issue for their projections. We provide a perspective on the usability of CMIP5 for practitioners in the Great Lakes region and offer recommendations for alternative options.     

Accuracy of cohesive laws with different shape for the shear behaviour prediction of bonded joints

The use of adhesive bonding as a joining technique is increasingly being used in many industries because of its convenience and high efficiency. Cohesive Zone Models (CZM) are a powerful tool for the strength prediction of bonded joints, but they require an accurate estimation of the tensile and shear cohesive laws of the adhesive layer. This work evaluated the shear fracture toughness (J_IIC) and CZM laws of bonded joints for three adhesives with distinct ductility. The End-Notched Flexure (ENF) test geometry was used. The experimental work consisted of the shear fracture characterization of the bond by the J-integral. Additionally, by this technique, the precise shape of the cohesive law was defined. For the J-integral, digital image correlation was used for the evaluation of the adhesive layer shear displacement at the crack tip during the test, coupled to a Matlab sub-routine for extraction of this parameter automatically. Finite Element Method (FEM) simulations were carried out in Abaqus® to assess the accuracy of triangular, trapezoidal and linear-exponential CZM laws in predicting the experimental behaviour of the ENF tests. As output of this work, fracture data is provided in shear for the selected adhesives, allowing the subsequent strength prediction of bonded joints.     

Improving the performance of asphalt mixture by addition of short-thin wheat straw pieces

To reduce costs and prevent environmental pollution in the fabrication process of straws fibrosis, a simple and efficient method of straw-reinforced asphalt concrete is proposed, which adds short-thin wheat straw pieces (SWSPs) into asphalt mixture directly. The road performances of wheat straw asphalt mixture (WSAM) are evaluated by means of three-point flexural test, rutting test, freeze thaw split test and the test of fracture toughness. The results show that when the SWSPs content is between 0.1% and 0.2%, the flexural strength of asphalt mixture increases by 23%, the dynamic stability by 48% (reaching 1592 times), fracture toughness by 27% while the rut depth decreases by 44%. It is indicated that high temperature stability of the asphalt mixture and low temperature crack resistance are intensified by adding an appropriate amount of SWSPs, and the results of freeze thaw split test manifest that the moisture susceptibility of WSAM is weaker.     

阜阳市SBS改性沥青混凝土道路工程实践探讨

结合阜城SBS改性沥青混凝土道路工程实践,从不同侧面分析探讨影响工程质量的诸多因素,并提出针对性的防控措施,以便在今后类似工程中引以为鉴。