Font Size: a A A

Research On Driving Fatigue Model Based On The Physiological Signal

Posted on:2018-11-08Degree:MasterType:Thesis
Country:ChinaCandidate:Z XieFull Text:PDF
GTID:2348330542467145Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
In recent years,with the rapid development of rail transit industry and its convenience advantages,more and more people choose rail transportation as their first way to travel.Since the rail way is mostly the exclusive,people can be less worried about the road ahead,so the rail transit usually safer than the road traffic.But the rail transit has its own unique security issues.Such as rail transit has different driving conditions from highway driving conditions,the driver may run long time in the same orbit for monotone iterative driving operation,which causes lazy,fatigue and lack of concentration,and it easily makes the accidents happen.Research of rail transit driver fatigue recognition problems to prevent the occurrence of safety accidents;has important practical significance.In this paper,rail transit driver fatigue recognition is studied as the main goal.Three aspects of content are as the follows:1.Fatigue experiments.From a number of driver fatigue experiments,we record the driver's pulse,respiratory and skin electrical signals to provide data support for the next research.and also based on the self assessment of driver's fatigue and the distance the drivers find the obstacles,we identify whether the driver is in the state of fatigue assistantly.2.Signal preprocessing and feature extraction.We use Butterworth filter to filter collected three signals' noises,wavelet transform to remove the pulse' baseline drift,extract the feature points of the three signals,and then extract the various characteristics of the signals.Through validity analysis,we make feature selection more suitable for fatigue judgment.3.Fatigue state recognition.Since there are multiple experiment subjects,individual differences and the data from different participants are uncertainty and randomness,a fusion of k-means clustering algorithm was proposed to build fatigue driving model,and then analyze the results of the experiment to identify the driving state.Through the above research,this paper proposes a suitable multiple people fatigue detection model.The experimental results show that the model has good driver state recognition rate and the total recognition rate has reached 80%.
Keywords/Search Tags:Rail transport, Fatigue recognition, Physiological signals, Characteristic effectiveness analysis, K-means clustering
PDF Full Text Request
Related items