Decision making through stochastic maneuver validation for overtaking on country roads

Adelberger, Daniel; Wang, Meng; Re, Luigi del

doi:10.1109/cdc42340.2020.9303922

Cited by 3 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition to the prediction method based on a Bayesian network structure, the approach from [4] (from now on with superscript C) is utilized. As a base of the calculation, the time dependent spatial mean ŝC µ (k + n) as well as the spatial standard deviation ŝC σ (k + n) need to be calculated.…”

Section: Stochastical Approach Cmentioning

confidence: 99%

“…s(k ) and v(k ) represent the initial position and velocity. Based on the assumptions from [4], the stochastic distribution over all steps results in…”

Section: Stochastical Approach Cmentioning

confidence: 99%

“…where µ C a and σ C a represent the mean and standard deviation of an assumed distribution of acceleration. Based on [4], a C µ = 0.2 m/s 2 and a C σ = 1 m/s 2 are used.…”

Section: Stochastical Approach Cmentioning

confidence: 99%

“…Additionally, not only influence factors in the observable area, but also limited observability may play a role. Perception limitations [3] pose a problem for ADAS systems [4] and are especially due to safety reasons subconsciously considered by human drivers [5]. These limitations can be overcome through vehicle-to-everything (V2X) communication [6] which, however, cannot yet be assumed to be realistically available in the near future [7].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Probabilistic Prediction of Driving Behavior on Country Roads

Sovtic

Adelberger

Wang

2022

2022 European Control Conference (ECC)

Self Cite

View full text Add to dashboard Cite

Advanced Driver Assistance Systems (ADAS) are already an important contributor to safety in conventional vehicles and thus in road traffic. This is obviously necessary, as in Austria in 2020, almost 99% of all accidents with personal injury were presumably caused by misconduct of involved persons or impairments [1]. Furthermore, country roads seem to be a safety-relevant environment, as they account for more than 50% of the fatalities in road traffic in Austria.Im nächsten Schritt wird der Einfluss der Identifikationsstrategie auf die Prädiktionsergebnisse betrachtet, wobei für die Identifikation ein Maß für die "Reichhaltigkeit" eines Trainingsdatensatzes eingeführt wird. Es zeigt sich, dass die "Reichhaltigkeit" als auch der Umfang des Identifikationsdatensatzes die Qualität der Prädiktionsmodelle beeinflusst.Schließlich wird ein ausgewähltes Bayes'sches Netzwerk in Simulation für einen Abstandsregeltempomat (ACC) verwendet, um das Verhalten des vorausfahrenden Fahrzeugs vorherzusagen. Der ACC wird als optimal-control Problem formuliert und im Rahmen eines MPC gelöst, während die Validierung auf einem Maß für den Kraftstoffverbrauch und Fahrkomfort basiert. Es zeigt sich bei der Prädiktion mit dem Bayes'schen Netz eine Verbesserung des Kraftstoffverbrauchs und Fahrkomforts im Vergleich zu anderen Prädiktionsmethoden, die keine topologischen Eigenschaften berücksichtigen.

show abstract

Section: Stochastical Approach Cmentioning

confidence: 99%

“…s(k ) and v(k ) represent the initial position and velocity. Based on the assumptions from [4], the stochastic distribution over all steps results in…”

Section: Stochastical Approach Cmentioning

confidence: 99%

“…where µ C a and σ C a represent the mean and standard deviation of an assumed distribution of acceleration. Based on [4], a C µ = 0.2 m/s 2 and a C σ = 1 m/s 2 are used.…”

Section: Stochastical Approach Cmentioning

confidence: 99%