Research On Intelligent Processing Method Of Geo Robot Monitoring Data

Georobot are widely uesd for unmanned monitoring field due to its high precision,good stability,and strong development.The data collected by Georobot can be intelligently processed to realize dynamic and real-time intelligent monitoring,and provide reference and early warnings for emergent security issues.Due to the influence of external environment and other factors,Georobot has low precision in measurement,and the observation value contains singular values.Therefore,it is very necessary to perform gross error detection and observation model correction on the data obtained by the measurement robot.At this stage,there are few researches on intelligent data processing algorithms for Georobot.This article focuses on the observation data of Georobot,starting from real-time gross error detection,observation value model correction and triangular elevation correction model,to intelligently process the data,the main research content as follows:(1)Aiming at the problem of singular values in the observation data during the unmanned monitoring process of Georobot,a real-time gross error detection method based on historical data is proposed.Based on the real-time gross error detection method,the constraint value is added.This method improves the real-time gross error detection method,and solves the problem of identifying normal changes as singular values,and analyzes the measured data.This method can effectively identify the singular values in the observation sequence and ensure the observation The validity of the data.(2)Aiming at the problem of inaccurate measurement of the observation value of the measuring robot due to external influences,this paper analyzes the influencing factors of the observation value of the measuring robot and the error of the triangular elevation measurement,and introduces the reference self-correction model and the triangular elevation correction model based on the inverse spherical air difference.Correct the angle observation value,distance observation value and calculated triangle elevation value of the measuring robot,and analyze according to the measured data.(3)Combining the triangular elevation correction model with deep learning,A Trigonometric Height Correction Model Based on long-term and short-term memory artificial neural network(LSTM)is proposed in this paper.The vertical angle below uses weather and other factors as the characteristic value,and the vertical angle correction amount as the label.The label of the monitoring point is predicted through the data of the historical sequence of the reference point,and the model correction effect is analyzed through the measured data.