The forecasting for indoor mobile robot position is a key factor affecting the safety of the robot navigation system and its operation. Combining with mobile robot position big data, it will be more reliable and robustness to make prediction of the mobile robot position. In this paper, a hybrid position big data forecasting model for indoor mobile robot is proposed: EWT-GB-DT-IEWT on Apache Spark platform. The proposed position big data forecasting model mainly consists of three parts, decomposition preprocessing, big data forecasting method and reconstruction post-processing. The EWT (Empirical Wavelet Transform) method is utilized to decompose the initial position data into subseries. The DT method (Decision Tree) enhanced by GB (Gradient Boosting) is applied to forecast each subseries on Apache Spark. The forecasting results of different series are reconstructed by IEWT (Inverse Empirical Wavelet Transform) post-processing. The results show that the proposed hybrid position big data forecasting method can predict the position trajectory accurately. The proposed position big data forecasting model on Spark can improve the reliability and stability of the indoor robot navigation system.
Clustering algorithm
• Hardware: Processor: i3 ,i5 RAM: 4GB Hard disk: 16 GB • Software: operating System : Windws2000/XP/7/8/10 Anaconda,jupyter,spyder,flask,hadoop Frontend :-python Backend:- MYSQL
₹10000 (INR)
2019