Algorithm_InformationAcquisition

This module represents the information acquisition of the driver-model. Information from the sensor is collected and the real vision of the agent is defined. Currently this module includes a substitute model for gaze movement in different areas of interest and an avert-view model based on stochastic calculated time-spans.

The information acquisition in the information acquisition module is divided into areas of gaze movement, which are defined via the corresponding parameters and their fixation time-distributions. For the simulation on straight roads with different lanes, in analogy to the Sensor-Driver, the following areas (BEHAVIORNAME) can be defined:

  • ALL,

  • FRONT,

  • REAR,

  • RIGHT,

  • LEFT,

  • FRONTLEFT,

  • FRONTMIDDLE,

  • FRONTRIGHT,

  • REARLEFT,

  • REARMIDDLE,

  • REARRIGHT,

  • ONLYMIDDLE,

  • AVERT

The combination of them and the corresponding parameters define a gaze-pattern.

|op| Gaze-types

The parameters of the information acquisition process can be set through the static system configuration with the corresponding Ids. We use four possible parameters to set up four gaze-patterns, which are used within the simulation and can change the gaze-behavior depending on the current situation (e.g. lane-change demand). The following PATTERNNAMEs are currently used and lead to a top-down request of the lane-change model:

  • NONE,

  • NORMALDRIVE,

  • LCLEFT,

  • LCRIGHT,

  • ONRAMP,

  • COOPERATIVEFORONRAMPER

Parameter

Type

Unit

Id

Description

Example

NormalGazePattern

String

0

Describes the gaze-pattern which is used for the normal driving behavior on a straight road with separate lanes. It is defined with the following syntax: “PATTERNNAME:BEHAVIORNAME,minTime,meanTime,stdTime,priority;BEHAVIORNAME,…

NORMALDRIVE:ALL,0,0,100,1

ExtraGazePattern1

String

1

“”

“”

ExtraGazePattern2

String

2

“”

“”

ExtraGazePattern3

String

3

“”

“”