Font Size: a A A

Determination Method Of Optimal Threshold Based On The Bootstrapping Technology And Analysis Of The Extrema Influence Factors

Posted on:2019-04-25Degree:MasterType:Thesis
Country:ChinaCandidate:Y WangFull Text:PDF
GTID:2382330542482771Subject:Mechanical design and theory
Abstract/Summary:
The load spectrum plays an important role in guiding the design of controllable life.It has been a key research subject in many industries,including engineering vehicles,and is still a very active research field.During the normal service of engineering vehicles,the extreme load is restricted by many factors,such as the differences of operators,the change of operation mode and operation object.The effects of these disturbances on the load are unpredictable and difficult to repeat.Based on the national science and technology support program(project no.: 2015BAF07B00)variety of extrapolation methods and the effects of different operators and road conditions on load are studied.In this paper,the methods of POT extrapolation and parameter extrapolation are analyzed by using the data of half shaft torque of wheel loader under different working conditions and operators.Using non-parametric extrapolation method,the extrema of half shaft torque of nine groups of loader in the excavation section are obtained.The main contents of this paper are as follows.1.Contrapose the problems of singular value,noise and trend term in the load signal caused by the complex operation of wheeled vehicle,the diversity of operating conditions and the differences of operators,this paper introduces the least square method,amplitude threshold method,gradient threshold detection method,statistical variance smoothing method and wavelet fractal noise reduction method and uses the least square method and wavelet fractal noise reduction method to eliminate the singular value,trend term and noise.Wheel loader has obvious operation characteristics.Based on the data of speed and bucket cylinder displacement,the loader semi-axis torque load is divided into six sections,and the load data of mining section are studied emphatically.2.This paper introduces routine threshold determination methods,analyzes the shortcomings of these methods,and proposes the optimal threshold selection method based on bootstrapping technology.Based on QQ-plot method,average probability deviation method and parameter sensitivity method,this paper compares thresholds selected by Kurtosis method,MEF method and Bootstrapping technology method and verifies the rationality and advantages of Bootstrapping technology method.3.This paper introduces the key steps of POT extreme value extrapolation and parameter extrapolation.Based on the measured data and extrapolation theory,the extrapolated extrema of nine measured load data are obtained respectively,and two independent extrapolation extrema are compared.4.Statistical analysis is made on the nine groups of semi-axial torque data under different operators and conditions.Based on SPSS software,this paper analyses the operation characteristics of different operators and different working conditions with the method of single variable multivariate analysis.A non-parametric extrapolation method is used to extrapolate the excavation section of nine groups of semi-axial torsional data of loader.This paper compares the test extrema and POT extrapolation extrema with non-parametric extrapolation extrema.And compared with the extrapolation extrema in different drivers and working conditions,the influence of the operation difference of the drivers is explained.Based on the measured data of semi-axial torque of wheel loader under different operators and working conditions,this paper analyzes the rationality of the current extrapolation method,studies the influence of test load and its extrapolation extrema caused by driver’s operating characteristics and operation objects and provides reference for the interference of many external factors in the load spectrum.
Keywords/Search Tags:Engineering Vehicle, Load Spectra, Extrapolation, Bootstrapping technology, Wavelet
Related items