基于带漂移的布朗运动的民用航空发动机实时性能可靠性预测

以民用航空发动机为研究对象,运用性能退化可靠性理论和随机过程方法,对发动机的可靠性进行了研究。通过分析发动机性能退化过程,建立了基于带漂移的布朗运动的可靠性模型,利用布朗运动特性研究了基于当前状态的退化时间预测方法,实例证明,该方法可操作性强,工程实现,为航空公司进行发动机机队科学管理提供了基础。     

基于随机松弛时间的行程时间可靠性计算模型

提出了基于随机松弛时间的行程时间可靠性动态计算模型。研究了交通系统内部车辆间相互干扰作用对行程时间可靠性的影响。运用交通流动力学理论建立模型,证明了模型的各项异性,并基于蒙特卡洛和欧拉折现法设计相应的求解算法。模型能够较好地反映给定交通供需条件下的行程时间可靠性退化规律,量化交通流内部车辆间相互干扰对行程时间可靠性的影响。算例分析表明车辆间相互干扰程度的加剧会引起行程时间可靠性退化轨迹的明显改变,干扰越剧烈,动态行程时间可靠性的退化速度越快、达到稳定状态时的行程时间可靠性水平越低。拓展了传统行程时间可靠性计算模型仅考虑外部供需的局限性,探讨了交通流松弛时间的随机波动对行程时间可靠性的影响。所建模型和方法可以为高速公路、城市快速路、主干道等大型设施的动态行程时间可靠性预测提供理论依据和实践参考。     

单兵筒式武器浮动发动机系统内弹道过程数值仿真研究

为研究浮动发动机的内弹道问题,该文对其发射过程进行了建模及计算分析。首先探讨了浮动发射系统的结构设计与工作原理,运用经典内弹道理论建立了发射系统内弹道过程的数学模型,并运用龙格-库塔公式进行数值计算,最后给出某组数值仿真及地面试验的研究结果。试验验证了模型的合理性,为单兵筒式武器浮动发动机发射系统的设计提供了理论指导。     

民用发动机状态混沌预测算法研究

摘 要：本文首先应用Haar和Db16小波对航空发动机排气温度的原始数据序列进行去噪处理,并且证明了处理后的数据序列具有混沌特征。其次应用混沌理论建立发动机状态预测算法,实现对排气温度的预测。通过检验排气温度预测值是否超过所规定的红线,以及该曲线是否平稳,从而进行发动机的健康状态排查。作为验证实例,文中使用一组某机型发动机实际飞行数据对预测算法进行了验证,并与加权一阶局域法和自回归滑动平均模型预测法进行了对比。结果表明,该组合模型算法精度优于加权一阶局域法和自回归滑动平均模型预测法,该方法可以为这种机型发动机故障预测提供决策依据。     

结合统计的模型辨识在发动机故障诊断中应用

以一个双轴分排涡扇发动机为例,研究了数理统计和模型辨识组合方法诊断发动机双故障的可行性和效果。模型辨识法基于发动机性能仿真模型,在辨识的过程中利用了发动机平衡技术,考虑了模型的非线性。运用数理统计方法对辨识法得到的结果进行再处理,弥补了单纯以拟合程度为标准进行故障隔离的不足,定义和选择合适的故障指标,可以在多重故障模型下得到良好的故障隔离效果。     

小型航空二冲程风冷发动机缸体流固耦合传热的仿真

以某小型航空二冲程风冷发动机为研究对象,结合风冷发动机缸体的传热特点和汽油机的燃烧规律并基于流固耦合理论,对风冷发动机缸体传热进行了模拟计算。根据耦合传热理论,将缸体与缸套的导热以及风冷翅片与冷却气流的对流换热转化为系统内部边界条件进行简化处理,并将模拟计算的结果与地面试验结果相比较,证明了耦合传热方法在研究缸体热负荷方面的可靠性。在考虑飞行器整个飞行包线下的外界大气环境下,对风冷发动机在整个工作高度范围内缸体的传热仿真进行了模拟计算,为发动机缸体传热的进一步研究开发奠定了基础。     

可拓综合评判在流脑发病率预测中的应用

本文运用物元分析理论，首先介绍了可拓综合评判在流脑发病率预测中的应用，然后在对所给多指标数据样本进行统计和分组的基础上，确定了各指标的权重；根据关联度的大小，确定了流脑发病率的范围。对１９７０～１９８０年回代预测，预测结果的符合率为９０．９１％；对１９８１、１９８２年外延预测，均与实际情况相符，比用模糊综合评判方法预测的总符合率提高了７．６９％。     

基于符号序列联合熵的发动机健康监控方法

本文引入信息熵理论,结合对多参数时间序列的符号化处理,探讨其在航空发动机健康监控中的应用。利用符号序列联合熵对航空发动机性能参数无序程度的描述,分析航空发动机健康状态及其演化方向。对两组实际飞行数据的实验分析表明,该方法描述的航空发动机的健康发展趋势与实际情况相吻合,能够较好反映航空发动机的健康状态,可以为航空发动机的健康监控提供一种新的方法。     

基于神经网络和模糊逻辑的航空发动机状态监视

摘要：设计了基于神经网络和模糊逻辑的航空发动机状态监视系统,它包含数据有效性检查和发动机健康评估。系统利用BP神经网络对发动机进行数据有效性检查,根据发动机的状态确定测量参数的有效边界以检测发动机的测量参数是否超限;利用模糊逻辑对发动机健康状况进行实时的评估,监控发动机因性能蜕化引起的异常。通过对某型涡扇发动机仿真,验证了基于神经网络和模糊逻辑的状态监视系统的有效性。     

基于定量构效关系的碳青霉烯类抗生素的分子设计

目的通过对碳青霉烯类抗生素定量构效关系（QSAR）的研究，进行新的碳青霉烯类化合物的分子设计。方法利用量子化学方法计算87种碳青霉烯类抗生素的18个结构参数，利用神经网络方法建立了碳青霉烯类抗生素的QSAR模型，根据药物分子设计理论，设计了一系列新的碳青霉烯类化合物。结果利用建立的碳青霉烯类抗生素的QSAR模型，对所设计的一系列新的碳青霉烯类化合物进行了MIC值预测，筛选出了有较高MIC值的5个碳青霉烯类化合物。结论所建立的QSAR模型能有效地进行碳青霉烯类化合物的MIC值的预测和化合物筛选，为这类新药开发提供新的参考和思路。     

Predicting drug metabolism and pharmacokinetics features of in-house compounds by a hybrid machine-learning model

We constructed machine learning-based pharmacokinetic prediction models with very high performance. The models were trained on 26138 and 16613 compounds involved in metabolic stability and cytochrome P450 inhibition, respectively. Because the compound features largely differed between the publicly available and in-house compounds, the models learned on the public data could not predict the in-house compounds, suggesting that outside of a certain applicability domain (AD), the prediction results are unreliable and can mislead the design of novel compounds. To exclude the uncertain prediction results, we constructed another machine learning model that determines whether the newly designed compound is inside or outside the AD. The AD was evaluated multi-dimensionally with some explanatory variables: The structural similarities and the probability obtained from the pharmacokinetic prediction model. The accuracy of predicting metabolic stability was 79.9% on the test set, increasing significantly to 93.6% after excluding the low-reliability compounds. The model properly classified the reliability of the compounds. After learning on the in-house compounds, the reliability model classified almost all (90%) of the public compounds as low reliability, indicating that the AD was properly determined by the model.     

Assessing performance of prediction rules in machine learning

INTRODUCTION: An important goal in machine learning is to assess the degree to which prediction rules are robust and replicable, since these rules are used for decision making and for planning follow-up studies. This requires an estimate of a prediction rule's true error rate, a statistic that can be estimated by resampling data. However, there are many possible approaches depending upon whether we draw observations with or without replacement, or sample once, repeatedly, or not at all, and the pros and cons of each are often unclear. This study illustrates and compares different methods for estimating true error with the aim of providing practical guidance to users of machine learning techniques. METHODS: We conducted Monte Carlo simulation studies using four different error estimators: bootstrap, split sample, resubstitution and a direct estimate of true error. Here, 'split sample' refers to a single random partition of the data into a pair of training and test samples, a popular scheme. We used stochastic gradient boosting as a learning algorithm, and considered data from two studies for which the underlying data mechanism was known to be complex: a library of 6000 tripeptide substrates collected for analysis of proteasome inhibition as part of anticancer drug design, and a cardiovascular study involving 600 subjects receiving antiplatelet treatment for acute coronary syndrome. RESULTS: There were important differences in the performance of the various error estimators examined. Error estimators for split sample and resubstitution, while being the most transparent in action and the simplest to apply, did not quantify the performance of prediction rules as accurately as the bootstrap. This was true for both types of study data, despite their highly different nature. CONCLUSIONS: The robustness and reliability of decisions based on analysis of genomics data could, in many cases, be improved by following best practices for prediction error estimation. For this, techniques such as bootstrap should be considered.     

Early Prediction of Pharmaceutical Oxidation Pathways by Computational Chemistry and Forced Degradation

PURPOSE: To show, using a model study, how electronic structure theory can be applied in combination with LC/UV/MS/MS for the prediction and identification of oxidative degradants. METHODS: The benzyloxazole 1, was used to represent an active pharmaceutical ingredient for oxidative forced degradation studies. Bond dissociation energies (BDEs) calculated at the B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level with isodesmic corrections were used to predict sites of autoxidation. In addition, frontier molecular orbital (FMO) theory at the Hartree-Fock level was used to predict sites of peroxide oxidation and electron transfer. Compound 1 was then subjected to autoxidation and H2O2 forced degradation as well as formal stability conditions. Samples were analyzed by LC/UV/MS/MS and degradation products proposed. RESULTS: The computational BDEs and FMO analysis of 1 was consistent with the LC/UV/MS/MS data and allowed for structural proposals, which were confirmed by LC/MS/NMR. The autoxidation conditions yielded a number of degradants not observed under peroxide degradation while formal stability conditions gave both peroxide and autoxidation degradants. CONCLUSIONS: Electronic structure methods were successfully applied in combination with LC/UV/MS/MS to predict degradation pathways and assist in spectral identification. The degradation and excipient stability studies highlight the importance of including both peroxide and autoxidation conditions in forced degradation studies.     

An organizational tool to describe and evaluate group performance in the health care setting

OBJECTIVE: To develop and test a tool for evaluating the performance of working groups in the institutional health care setting and to discuss its utility for pharmacists and other health care providers. DESIGN: A model for understanding how groups make decisions in health care settings was developed based on theories derived from organizational research. Our general group decision-making model was based on data collected from a national survey of members of hospital pharmacy and therapeutics (P&T) committees. Members were asked to describe how they interacted when making decisions regarding whether to add or delete drugs from their hospital's formulary. The relationships were examined using structural equation modeling. SETTING: Large teaching hospitals. PARTICIPANTS: Key members of P&T committees. MAIN OUTCOME MEASURES: Committee members' responses about committee structure, organizational environment, quality of their interactions with each other, and the committee's decision-making abilities. Individual responses were aggregated into group measures to improve reliability and validity. RESULTS: Responses were received from 114 of the 222 hospitals surveyed (51.3%), and analysis was conducted on 95 hospitals (42.8%) for which three or more group members responded. Positive interactions among group members had the greatest influence on the group's decision-making abilities (beta = 0.92). These interactions were facilitated by a well-designed group (gamma = 0.71) and the support of the environment within which the committee operates (gamma = 0.29). CONCLUSION: The general group decision-making model tested in this study may be useful in understanding how the internal processes of health care groups such as P&T committees influence group performance and, hence, group outputs. Health care providers who apply a systematic approach to assessing group performance will find this model useful because it provides a starting point for examining the process, a way to show interrelationships among various group activities, and a guide to which strategies may be most effective for improving the decision-making process.     

Using the Box–Behnken experimental design to optimise operating parameters in pulsed spray fluidised bed granulation

In this work, the influence factors of pulsed frequency, binder spray rate and atomisation pressure of a top-spray fluidised bed granulation process were studied using the Box–Behnken experimental design method. Different mathematical models were developed to predict the mean size of granules, yield, relative width of granule distribution, Hausner ratio and final granule moisture content. The study has supported the theory that the granule size can be controlled through the liquid feed pulsing. However, care has to be taken when the pulsed frequency is chosen for controlling the granule size due to the nonlinear quadratic relation in the regression model. The design space of the ranges of operating parameters has been determined based on constraints of the mean size of granules and granule yield. High degree of prediction obtained from validation experiments has shown the reliability and effectiveness of using the Box–Behnken experimental design method to study a fluidised bed granulation process.     

蚌壳式反推装置打开时反推性能参数计算

根据自由流线理论建立了蚌壳式反推装置二维流动数学模型,得到蚌壳式反推装置打开时反向气流偏转角 与蚌壳式反推装置几何参数间的关联关系。将该模型和涡扇发动机性能计算模型耦合,得到了蚌壳式反推装置打开时反推参数的计算模型。以某型涡扇发动机为例,计算了着陆时反推气流角和反推装置几何参数,以及反推力随马赫数、发动机转速的变化关系,并与测量数据进行了对比分析。表明本文给出的蚌壳式反推装置打开时反推性能计算模型具有较好的计算精度。     

A prediction system of oxidation reaction as a solid-state stress condition: Applied to a pyrrole-containing pharmaceutical compound

Stress conditions for predicting oxidative degradation products in solid-state pharmaceutical compounds were investigated. 4-Methyl-2-(3,4-dimethylphenyl)-1-(4-sulfamoylphenyl)pyrrole, Compound A, was used as the model compound for this study and its four main degradation products were due to oxidation, as identified by LC–MS and LC-¹H NMR. In order to develop a prediction system for the oxidation reaction, solid-state Compound A was stored under moisture-saturated conditions. Hydrogen peroxide was added to the solution used to saturate the headspace with moisture and oxygen was substituted for the headspace air, in order to stimulate the oxidation reaction. After optimizing the conditions, a similar degradation product profile to that actually observed in the stability studies was obtained in only 3 days under conditions using 3% hydrogen peroxide at 40 °C. The prediction of the oxidative degradation products in a solid-state pharmaceutical compound was successfully achieved in a short term utilizing this newly developed prediction system.