基于BP神经网络的高等级公路交织区安全评价

分析了交通冲突数与其影响因素之间及交通冲突数与安全性之间的关系;以交通冲突数作为其影响因素与道路安全性之间的"转换器",提出了基于BP神经网络的高等级公路交织区安全评价模型;为高等级公路交织区安全评价开辟了新的途径。     

论神经元再生现象的科学认识论意蕴

本从神经元再生这一新的医学发现出发，阐述了神经元再生现象对以稳定神经结构为基础的科学认识论带来的挑战，并就它对人的大脑可塑性、人工智能的拓展及传统认知观念的更新所具有的意义进行了多角度的分析。论力图表明：神经科学的新发现必然会不断地粉碎陈旧的思维定势所设置的障碍，改变我们对世界的看法，推动哲学观念的变革。     

ARTIFICIAL IMMUNE ALGORITHM OF MULTICELLULAR GROUP AND ITS CONVERGENCE

Objective To find out more extrema simultaneously including global optimum and multiple local optima existed in multi-modal functions. Methods Germinal center is the generator and selector of high-affinity B cells, a multicellular group＇s artificial immune algorithm was proposed based on the germinal center reaction mechanism of natural immune systems. Main steps of the algorithm were given, including hyper-mutation, selection, memory, similarity suppression and recruitment of B cells and the convergence of it was proved. Results The algorithm has been tested to optimize various multi-modal functions, and the simulation results show that the artificial immune algorithm proposed here can find multiple extremum of these functions with lower computational cost. Conclusion The algorithm is valid and can converge on the satisfactory solution set D with probability 1 and approach to global solution and many local optimal solutions existed.     

化LF(X)中任一公式为可归约形式的算法

针对基于归结方法的自动推理过程中的化任一公式为可归约形式这一问题进行研究．简述了LF(X)及其中公式的相关概念，证明了对LF(X)中公式进行转化的若干理论结果，并由此给出化LF(X)中任一公式为可归约形式的理论基础和基本算法．     

Artificial anal sphincter with sensor feedback powered by transcutaneous energy transmission

This artificial anal sphincter (AAS) is used to help patients who are in severe incontinence. It consists of two parts. The first part is taken outside with patients. The second is implanted in the bodies to provide independent evacuation for patients. It is powered through transcutaneous energy transmission by the first part. There are two sensors in the second part. One sensor measures the pressure in the cuff-clamp around the recta. The other measures the pressure of the recta. As soon as the pressure measured by the later reaches the threshold, a signal is sent out to patients through wireless communication to alarm patients preparing defecation. Patients can give a start signal to begin defecating after they are ready. And they can also give a signal to stop defecating like a healthy person. With the aids of the novel medical device, patients can defecate as a healthy person. The first generation of this AAS is presented.     

动车组盥洗室水盆台面裂纹分析研究

针对动车组盥洗室水盆人造石台面裂纹,调研发现,台面裂纹数量较多并且主要集中在背部接缝和台盆四角处。根据调研结果,对人造石台面进行仿真分析、拉伸强度测试、弯曲强度测试、落球冲击试验。试验结果表明,人造石台面符合标准要求,但是台面中的亚克力含量对台面强度有一定影响。根据分析和试验结果,提出了改进接缝粘接结构、加强支撑、选择高含量的亚克力台面等优化措施。     

Neural network-based fuel consumption estimation for container ships in Korea

ABSTRACT

Due to the outstanding strength of advanced machine-learning techniques, they have become increasingly common in predictive studies in recent years, particularly in predicting ship energy performance. In constructing predictive models, prior studies have mostly employed vessels’ technical parameters to establish machine-learning algorithms. To bridge this research gap and enable wider applications, this paper presents the design of a multilayer perceptron artificial neural network (MLP ANN) as a machine-learning technique to estimate ship fuel consumption. We utilized the real operational data from 100–143 container ships to estimate fuel consumption for five different container ships grouped by size. We compared the performance of two ANN models and two multiple-regression models. Four input parameters (sailing time, speed, cargo weight, and capacity) were included in the first ANN and the first regression model, while the other two models only consider two inputs from physical function. The mean absolute percentage error of the ANN models with four inputs was the smallest and less than those in extended statistical models, demonstrating the MLP’s superiority over the statistical model. The MLP ANN model can thus be applied to confirm the effectiveness of the slow-steaming method for achieving energy efficiency.     

多年冻土区灌注桩的人工冷却试验研究

多年冻土区灌注桩修筑过程中，混凝土携带的热量和水化热会对周围冻土产生强烈的热扰动，由于基础在回冻前承载力非常小，因此桩基的快速回冻已成为该地区基础设施建设中备受关注的问题。为解决该问题，创新性地提出人工冷却法在多年冻土区灌注桩中的应用，以实现基础的快速降温和回冻，并在青藏高原北麓河进行现场人工冷却灌注桩试验。试验结果表明：人工冷却降温效果显著，试验进行到第2天，基础周围土体温度已经降至天然场温度以下；随着冷却试验的继续进行，基础和周围土体的温度会进一步降低；人工冷却有效降低了土体温度，并增加了桩基周围土体的冷储量，冷却试验结束后的第7个月试验桩的平均界面温度为-0.6℃，无冷却措施桩对应的平均界面温度为-0.37℃。承载力计算结果表明：人工冷却可以快速、大幅提高桩基的承载力，通过该措施的应用，试验桩的承载力可提高至2 231 kN，而无冷却措施桩对应时刻的承载力仅为549 kN；该措施有效缩短了冻土区灌注桩施工的后续等待时间，经过人工冷却的桩基承载力很高，冷却结束后即可进行上部荷载的施工。人工冷却可作为冻土地区桩基快速施工中的一种重要方法和有效途径，其不仅能够解决灌注桩对冻土产生热扰动这一难题，还能够有效缩短工程的建设周期，具有很高的工程应用价值。     

地铁隧道半断面水平冻结施工的水热力耦合特性研究

地层冻结是一个水、热、力三场耦合问题,基于考虑相变的水热力耦合理论,分析地层温度场、水分场随时间的变化规律,对比冻结前后隧道开挖的位移场分布特点。研究表明:随着冻结时间的增加,冻结管周边土体的温度将逐渐降低并发生冻结,出现冻结区域扩大、交圈、冻结帷幕增厚等变化过程;冻结帷幕内的冻土存在未冻结水,随着冻结时间的增加,大多数位置的未冻结水体积含量出现先增大后减小的变化趋势;建议在冻结帷幕两侧下方追加冻结管以保证实际的冻结帷幕达到设计范围,并对冻结管距开挖轮廓线的距离进行差异化设定。     

基于人工神经网络的快艇总振动预报

快艇因其重量轻、航速高，因而振动比较难控制，且一旦船体结构基本确定之后，要改善其总振动状况将更加困难。由于目前的规范计算法和有限元法需进行大量计算，费时费力，这对快艇设计的初期阶段很不利，因而在已有的设计经验的基础上，建立了快艇总振动预报的人工神经网络模型，为快艇设计的初期阶段提供参考。样本检验表明，该网络能较好地预报快艇的总振动。