Recreation of a vehicles under programmed guiding control, from the tenant's perspective

2099 days ago, 747 views
PowerPoint PPT Presentation
What visual signals does the HO use in deciding how to control the vehicle? ... How might such a plan contrast and vehicle execution under full programmed control? ...

Presentation Transcript

Slide 1

´╗┐Blame Detection in Lateral Vehicle Control Simulation of a car under programmed guiding control, from the inhabitant's perspective Evaluation of a tenant's capacity to recognize and adjust for controller glitches that will bring about the vehicle to float out of its path Evaluation of a tenant's capacity to make up for a controller blame and keep the auto securely in its path

Slide 2

Test Setup PC Monitor Steering Wheel HO Controller Noise

Slide 3

Unique Properties of the Human Controller The human controller brings certain properties up 'til now unduplicated in machine controllers. These include: 1) Superior picture handling 2) Superior versatility based upon intellectual aptitudes 3) Superior capacity to foresee But slower reactions, poorer control execution, isolated consideration, fatigability, differing qualities of abilities, and distractibility are particularly human attributes too. What, then, is the right system to consolidate human and machine blame recognition abilities?

Slide 4

Roles of the Human Operator We see the scope of conceivable parts as lying on a continuum. Toward one side (H), we have the human completely in control, at the flip side, the controller (C) is in control. We have centered our consideration on three middle of the road focuses or agreeable procedures (CS), between these extremes: H C 1) as helper 2) as peer 3) as emergency handler We have built up a Scenario Evaluation System (SES) in which to investigate both the blame location and the blame taking care of execution of each of these three situations.

Slide 5

Purpose of the Scenario Evaluation System The SES is basically a pre-test system. It is a framework in which we disconnect numerous factors to focus on only a couple. Our primary intrigue is the part of vision in the execution of the blend shaped by the human administrator and the controller. We along these lines avoid imperative factors from thought. The favorable position is specificity concerning the part of vision. Impediments are overcome by moving the study to either a genuine test system or to the test track.

Slide 6

Detection Test Series How precisely and rapidly can the HO distinguish a controller blame? # issues # "hits" # "misses" # "false alerts" Lateral and rakish deviation at identification Point on track where blame happened/was distinguished With and without framework commotion As a component of street geometry (straight-a-way, bend, S-bend, move point) Physical characteristics, conduct, and location criteria utilized will differ from subject to subject How do recognitions (hits) and false cautions exchange off concerning The different situations Road geometry

Slide 7

Quality of Reaction Test Series Comparison of HO's and Controller's direction execution Comparison of Lateral and Angular Deviation How well can the HO make up for a controller blame and keep the auto securely in the path? Horizontal and rakish deviation As a component of CS (assistant, peer, emergency handler) As an element of street geometry How does execution change with assignment nature? To what extent does it take to take in a given CS?

Slide 8

Fault & Curvature Change Locations

Slide 9

Detection Test Series: Summary

Slide 10

Lateral Deviation

Slide 11

Learning Curve Effect

Slide 12

Placement of False Alarm Criterion

Slide 13

Hit Rate versus False Alarm Rate

Slide 14

Quality of Reaction Test Series: Part I

Slide 15

Additional Questions the SES Can Answer 1. How is the HO's location/nature of response execution influenced by: - Cooperative Strategy - Road Geometry - Atmospheric Conditions - Vision Defects - Presence of different vehicles (in front, to the sides) 2. How well can the HO "drive" the PC-model of the vehicle contrasted with the programmed controller? - What is the HO's system for arranging different street geometry's circumstances (bends, surpassing and passing slower vehicles)? - What visual signs does the HO use in deciding how to control the vehicle? How can he utilize these? - Can this be utilized to enhance the operation of the controller?

Slide 16

Additional Questions (Continued) 3. How well can the HO make up for a controller blame in the wake of taking (fractional or full) control of the vehicle subsequent to recognizing a blame? - What is the appropriation of ordinary human execution 4. How best can a HO impart the driving obligations to a programmed controller? - What sort of controller/helpful methodology works best in a common game plan? - How might such a plan contrast and vehicle execution under full programmed control?

Slide 17

Additional Questions (Continued) 5. How would such a mix (Item 4) be displayed? - HO liable to be nonstationary, nonhomogeneous, nonlinear, and so on? - Develop subjective model of HO? - Interface amongst HO and controller is basic, for model exactness, as well as for appropriate coordination and correspondence between the two. - How will controller plan be influenced?

Slide 18

Additional Uses Simulate new plans and driving methodology Develop particulars for Operator sight and aptitude prerequisites Road and vehicle/controller execution benchmarks Customize Cooperative Strategies for individual drivers Performance examinations

Slide 19

Additional Uses (proceeded with) Re-making of real driving situations for further investigation Identify and evaluate the imperative results that more intricate and costly test techniques must give First step towards advancement of a subjective model consolidating the HO and controller in a "total" framework Develop best helpful system Incorporate full scope of human abilities into model Tailor CS to individual drivers

Slide 20

Cooperative Strategies