Design And Implementation Of Inlet Total Temperature Correction System

Data analysis is the core of the research and development process of aero-engine.At present,there is a lack of data analysis tools for aero-engine.In order to fully embody the value of existing and emerging data resources,it is necessary to establish an efficient data analysis and processing system for aero-engine test data.Inlet total temperature is an important parameter to measure engine performance.At present,it is very difficult and expensive to determine the recovery coefficient of total temperature in China,and empirical recovery coefficient is used.Therefore,the correction effect of this method still needs to be improved,and a new method is urgently needed to improve the performance of the correction.In this thesis,based on aero-engine test data and application scenarios of inlet total temperature correction,machine learning technology are used to propose a inlet total temperature correction model,and the inlet total temperature correction system is designed and implemented.（1）In this thesis,the inlet total temperature correction problem is classified as a nonlinear regression problem according to the problem scenario,and four model methods such as random forest and neural network are used to carry out experiments on this problem,which provides model support for the subsequent realization of the inlet total temperature correction function of the system.（2）The data processing and analysis function in the system is realized by the algorithm model.In order to continuously expand the new functions of the system,the structure of the algorithm library is designed in this thesis.The algorithm library includes an algorithm content module,call interface,and algorithm management module,covering the whole process of data processing and analysis.In addition,algorithm management can realize the addition of algorithms,so that the system has expansibility and provides convenience for enriching system functions.（3）System design and implementation.In this thesis,demand analysis and system design are carried out based on actual project requirements.The whole system includes two parts: mobile terminal and management terminal.It is divided into three layers: data storage layer,business logic layer and application layer.The management terminal mainly manages users,logs,experimental results and other aspects of the system.The mobile terminal is developed by Android Studio framework,which is mainly used for data processing and analysis of real-time data and offline data.In addition,Tensorflow Lite framework is adopted to realize the deployment of the model in mobile terminals,and tensorflow.js is used to realize simple model training.The system test proves that the system can basically meet the actual demand and run stably.