A model reference adaptive control(MRAC)with smooth switching scheme was proposed for piecewise linear systems,and the method was utilized in turbofan engine control to avoid the discontinuity of control input.In this...A model reference adaptive control(MRAC)with smooth switching scheme was proposed for piecewise linear systems,and the method was utilized in turbofan engine control to avoid the discontinuity of control input.In this scheme,each sub-region of the operating envelope had its own MRAC controller,and smooth indicator function based smooth switching scheme was introduced to switch multiple controllers smoothly at the boundary of adjacent sub-regions.The Lyapunov stability analysis indicated that the proposed smooth switching scheme can guarantee the convergence of the closed-loop system during the controllers switching.The tracking error system was converted into a switched system to analyze the global stability of the closed-loop system.The advantage of the method was that the chattering of system output and instability caused by asynchronous switching can be eliminated.The simulation illustrates the effectiveness of the proposed control scheme in comparison with the existing MRAC controller with gain scheduling for turbofan engine.展开更多
文摘目的探讨卒中相关性肺炎(stroke-associated pneumonia,SAP)患者发生神经功能恶化(neurological deterioration,ND)的影像标志物,并对其预测价值进行分析。方法回顾性收集2020年1月至12月急诊或神经内、外科收治的SAP患者,根据美国国立卫生研究院卒中量表(National Institutes of Health Stroke Scale,NIHSS)是否增加≥4分,分为ND组与非ND组。首先,通过单因素分析比较两组患者临床特征、实验室检查、脑部磁共振成像(magnetic resonance imaging,MRI)及肺部计算机断层扫描(computed tomography,CT)影像学指标差异;其次,将单变量分析P<0.05的指标纳入多因素Logistic回归模型进行分析。结果本研究共纳入SAP患者73例,其中有15例发生ND(20.54%)。单因素结果显示两组患者在枕叶梗死、丘脑梗死、大面积脑梗死和肺部CT评分等方面差异有统计学意义,Logistic回归分析显示,肺部CT评分(OR=1.218,95%CI:1.010~1.469)、是否为大面积脑梗死(OR=9.033,95%CI:1.746~46.742)、丘脑梗死(OR=10.016,95%CI:1.523~65.862)是SAP发生ND的独立危险因素,受试者工作特征(receiver operating characteristic,ROC)曲线分析显示,曲线下面积(area under curve,AUC)为0.849(95%CI:0.752~0.947),表明模型拟合效果较好;将多因素Logistic回归分析中OR值最大的丘脑梗死进行ROC分析,AUC为0.624(95%CI:0.450~0.797),其预测ND的敏感度为33.33%,特异度为91.38%,阳性预测值为50.00%,阴性预测值为84.13%,准确率为79.45%;大面积脑梗死及肺部CT评分联合分析AUC为0.799(95%CI:0.690~0.909),预测ND的敏感度为100.00%,特异度为48.28%,阳性预测值为33.33%,阴性预测值为100%,准确率为58.90%;肺部CT评分AUC为0.744(95%CI:0.611~0.878),取肺部CT评分截断值为4.5时,其预测ND的敏感度为80.00%,特异度为63.80%,阳性预测值为36.36%,阴性预测值为92.50%,准确率为67.12%。结论肺部CT评分、大面积脑梗死、丘脑梗死可能是SAP患者发生ND的危险因素。肺部CT评分敏感度及诊断价值高,且肺部CT评分越高(>4.5分),ND转化风险越高。丘脑梗死的特异度且诊断价值较高,可以预测SAP发生ND。脑梗死范围及肺部CT评分联合预测ND的敏感度高,漏诊率低,可以提示临床对于可能发生ND的患者重点关注。
文摘A model reference adaptive control(MRAC)with smooth switching scheme was proposed for piecewise linear systems,and the method was utilized in turbofan engine control to avoid the discontinuity of control input.In this scheme,each sub-region of the operating envelope had its own MRAC controller,and smooth indicator function based smooth switching scheme was introduced to switch multiple controllers smoothly at the boundary of adjacent sub-regions.The Lyapunov stability analysis indicated that the proposed smooth switching scheme can guarantee the convergence of the closed-loop system during the controllers switching.The tracking error system was converted into a switched system to analyze the global stability of the closed-loop system.The advantage of the method was that the chattering of system output and instability caused by asynchronous switching can be eliminated.The simulation illustrates the effectiveness of the proposed control scheme in comparison with the existing MRAC controller with gain scheduling for turbofan engine.
