机场能源数据的采集与处理方法研究

发布时间：2018-08-07 21:05

【摘要】：随着科技的发展,机场在信息化数字化方面也有了长足的进步,机场的各个部门也都开发或者引进了不同类别的信息管理系统,改善了机场信息化应用的环境。但受制于当前的科技水平、机场各大能源站点复杂的监测环境、电子仪器的不稳定性等因素,机场能源数据的采集出现各种各样的困难,采集上来的能源数据广泛存在冗余、缺失、噪声等不良现象,所以研究新型的机场能源数据采集技术,以及研究新型的机场能源数据处理算法尤为重要。本论文首先针对机场能源数据采集过程中的各种问题,利用无线传感网络技术、自动化技术、数据库技术等,设计了机场能源数据实时采集平台,然后对采集到的机场能源数据进行处理与分析。根据机场能源数据在不同时间的不同特性提出特征权重的数据预处理方法,提出结合经验模式分解与最小二乘支持向量机的联合回归预测方法解决能源数据缺失的问题,同时利用果蝇算法改进最小二乘支持向量机的参数寻优过程。将常用的预测方法与本论文提出的方法进行对比验证实验,仿真结果表明采用本论文方法预测准确度有显著提高,能够胜任机场能源数据缺失的填补工作。最后利用已经建立的回归预测模型,提出基于无迹变换的机场能源数据的改进型卡尔曼滤波方法,在原有卡尔曼滤波方法的基础上,通过加入误差反馈提高滤波效果。通过对改进算法的实现,得到更加精准的机场能源数据。在相同站点条件下进行对比实验,对比不同滤波方法的滤波效果;在不同站点不同模型的条件下进行验证实验,验证本论文方法的适用性和有效性。结果表明,本论文方法在建立回归预测模型的基础上,通过闭环的误差反馈控制减少误差扩散的影响,对于未知的非线性系统,有很好的滤波效果,具有良好的发展与应用前景。
[Abstract]:With the development of science and technology, the airport has made great progress in the field of information digitization. Various departments of the airport have also developed or introduced different kinds of information management systems, which have improved the environment for the application of airport information. However, due to the current level of science and technology, the complex monitoring environment of the major energy stations in the airport, the instability of electronic instruments, and so on, there are various difficulties in the acquisition of airport energy data, and there is widespread redundancy in the energy data collected. Therefore, it is very important to study the new airport energy data acquisition technology and the new airport energy data processing algorithm. In this paper, firstly, aiming at various problems in the process of airport energy data acquisition, using wireless sensor network technology, automation technology, database technology and so on, the real-time acquisition platform of airport energy data is designed. Then the collected airport energy data processing and analysis. According to the different characteristics of airport energy data at different time, the data preprocessing method of feature weight is put forward, and the joint regression prediction method based on empirical mode decomposition and least squares support vector machine is proposed to solve the problem of missing energy data. At the same time, the algorithm of Drosophila was used to improve the parameter optimization process of least squares support vector machine (LS-SVM). The simulation results show that the prediction accuracy of this method is significantly improved and can be used to fill the lack of airport energy data. Finally, an improved Kalman filtering method based on unscented energy data is proposed by using the established regression prediction model. On the basis of the original Kalman filtering method, the filtering effect is improved by adding error feedback. Through the implementation of the improved algorithm, more accurate airport energy data can be obtained. Comparing the filtering effect of different filtering methods under the same site condition, and verifying the applicability and validity of this method under the condition of different stations and different models. The results show that this method can reduce the influence of error diffusion through closed-loop error feedback control on the basis of establishing regression prediction model, and has a good filtering effect for unknown nonlinear systems. It has a good prospect of development and application.
【学位授予单位】：中国民航大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP274.2

【参考文献】