考虑到大气温度、大气湿度和燃烧热值等因素的影响,在第三相似理论的基础上提出了一种基于气路参数偏差值的标准化修正模型,以发动机生产厂家的历史数据为学习样本,建立了多元非线性超定方程组,采用高斯牛顿迭代法进行模型中未知参数的回归求解.使用该模型对测试集数据进行偏差值求解,并把结果与其他模型求解的偏差值进行对比,结果表明:修正后的标准化模型具有更高的精度,更接近于厂家系统的原始模型,具有较高的创新性和实用价值. Considering the influence of atmospheric temperature, atmospheric humidity and calorific value, a standard correction model based on the third similarity theory is proposed based on the deviation of gas path parameters. The historical data of engine manufacturer is taken as the learning sample, A set of multivariate nonlinear overdetermined equations was established and the Gauss Newton iteration method was used to solve the regression of the unknown parameters in the model.Using the model to solve the deviation of the test set data and comparing the results with those of other models, It shows that the modified standardized model has higher accuracy and is closer to the original model of the manufacturer system, which has a higher innovation and practical value.