非等时距Verhulst修正模型及其在直升机故障率预测中的应用

直升机累计故障率曲线与Verhulst模型曲线相近。但是，Verhulst模型要求初始数据具有等距、呈“S”型特点，且其模型终值趋于灰色发展系数a／b，无法实现对直升机故障率准确预测。针对这一问题，提出非等时距Verhulst模型，并应用该模型进行直升机累计故障率预测。同时根据直升机累计故障率终值趋于1的特点，为模型增加了预测误差修正项。通过实例验证表明，非等时距Verhulst修正模型预测精度达到了94．1％，对中长期预测具有重要的意义。

Non-Equal-Time-Interval Modified Verhulst Model and Its Application in Helicopters Failure Rates Prediction

Helicopter failure rates curve is similar with the Verhulst model curve. However, the Verhulst model application need equal-interval original data and the original data must fitting to ＂S＂ curve. Furthermore, the terminal value of the Verhulst model approaches grey developing factor a/ b. So, it is difficult to use the Verhulst model to predict the helicopter failure rates. Aiming at resol- ving this problem, the non-equal-time-interval modified Verhulst model was proposed, and it was applied to predict the helicopter failure rates. At the same time, a prediction error modified item was applied to the model based on the helicopter failure rates terminal value approaches 1. It was proved by the instance that the precision of the non-equal-interval Verhulst model could approaehe 94.1%.