In this paper for an autonomous parking system two different approaches for fitness functions are presented, which evaluate the quality of parking maneuver automatically. The illustrations show, that both fitness functions have a sloping characteristic for collision maneuvers, where their fitness value is less than zero. The slope of the distance criterion function is more steep in that domain, compared to the area criterion function. Furthermore, the distance criterion function shows an edge at the crossing to negative values, where the area criterion function has a smooth transition for fitness values around the zero. Nevertheless, both fitness functions show the same crossover characteristic from positive to negative values, like the logarithmic absolute representation illustrates. From that follows, that both types of function do identify the same scenarios where a collision happens. The area criterion function provides a sloping characteristic for maneuvers with positive fitness values, in contrast to the distance criterion function, which returns constant fitness values for scenarios without clash and does not differentiate between good and better maneuvers. The consequence is, that the area criterion function can direct the search better towards collision maneuvers, than the distance criterion function. This supports a faster convergence for the optimization of the test cases to find faulty parking situations.
Automatic testing of an autonomous parking system using evolutionary computation
2004
8 Seiten, 20 Bilder, 10 Quellen
Conference paper
English
Autoindustrie , Positionsmessung , Fahrzeug , elektronische Steuerung und Regelung , Produktentwicklung , numerisches Verfahren , automatisches System , Manövrierfähigkeit , Analysis (Mathematik) , Logarithmus , Vergleich , Fehlerabschätzung , Optimierung , Kraftfahrzeugelektronik , evolutionärer Algorithmus
Automatic Testing of an Autonomous Parking System Using Evolutionary Computation
British Library Conference Proceedings | 2004
|Automatic testing of an autonomous parking system using evolutionary computation
Automotive engineering | 2004
|Automatic Testing of an Autonomous Parking System Using Evolutionary Computation
British Library Conference Proceedings | 2004
|Automatic Testing of an Autonomous Parking System Using Evolutionary Computation
SAE Technical Papers | 2004
|