摘要：研究数据拟合模型的可靠性问题对工程实践研究具有重要意义。针对这一问题，本文提出了使用蒙特卡洛方法来预测拟合模型的可靠性，并给出了离散数据拟合的拟合参数、可靠性指标和误差概率三个指标，最后结合电动机空载试验数据编写了相应的程序，同时进行了U-I性能曲线拟合。计算结果表明，该方法的引入使数据拟合模型与概率分析有效结合、互相补充，为离散数据拟合模型的可靠性估计提供了一种更科学，更准确的方法，可以用来估计曲线拟合的可靠性。
关键词：蒙特卡洛方法；拟合模型；拟合参数；可靠性指标；误差概率

Abstract: Research the reliability problem of discrete data fitting model is significant to the engineering practice research. In order to tackle this problem, the paper put forward the Monte Carlo method to predict the reliability of the fitting model, and gave the three indexes such as: fitting parameters, reliability index and error probability, finally did the UI curve fitting and compiles the corresponding programs using the data of motor test. The calculated result shows that this method can combine data fitting model with probability analysis effectively, making its complement each other, a more scientific and accuracy method is presented for the reliability evaluation of discrete data fitting model, and can assess the reliability of curve fitting.
Key words: Monte Carlo method; fitting model; fitting parameters; reliability index; error probability