Abstract: After analyzing and screening 7 significantly related indicators affecting the stress behavior of controllers, and classifying the stress degree of controllers in an orderly way, a prediction model of the stress degree of controllers based on cumulative logistic was established. Parallelism test was carried out on the prediction effect of the proposed model. The feasibility of the proposed prediction model was verified by using the quantitative measurement data of 74 other controllers in the actual data collected by an air traffic control authority. Results show that the significance level of the proposed models determining coefficient of goodness of fit (Pearson χ2 statistical magnitude and Deviance statistical magnitude) is 0.076 and 0.690 respectively, which are far greater than 0.05; the prediction accuracy of the measured data is 75.67%. The proposed model is more accurate in predicting the stress level of controllers.

Key words: traffic and transportation engineering, stress (source), cumulative logistic regression model, controller, correlation analysis, Wald test

摘要： 分析并筛选出影响管制员应激行为的7个显著相关指标，并将管制员应激程度进行有序多分类后，建立了基于累积Logistic的管制员应激程度预测模型，并对模型预测效果进行平行性检验。利用某空管局实际采集数据中的另外74名管制员的定量测量数据进行分析，验证预测模型的可行性。结果表明：该模型的拟合优度判定系数(Pearsonχ2统计量和Deviance统计量)的显著水平分别为0.076, 0.690，远大于0.05；对实测数据的预测准确率达到75.67%，该模型对管制员应激程度的预测较为精确。

关键词: 交通运输工程, 应激(源), 累积Logistic回归模型, 管制员, 相关性分析, Wald检验