Genetic Algorithm-Based Test Parameter Optimization for ADAS System Testing

Klück, Florian; Zimmermann, Martin; Wotawa, Franz; Nica, Mihai

doi:10.1109/qrs.2019.00058

Cited by 49 publications

(56 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results of the study show that using the genetic algorithm, safety-critical cases can be generated by performing a lower number of tests than when using simulated annealing. Furthermore, previously in reference [36], it has been shown that the genetic algorithm is also more effective than random selection in generating such cases.…”

Section: B Test Case Definition and Generationmentioning

confidence: 96%

Software Verification and Validation of Safe Autonomous Cars: A Systematic Literature Review

et al. 2021

View full text Add to dashboard Cite

Section: B Test Case Definition and Generationmentioning

confidence: 96%

Software Verification and Validation of Safe Autonomous Cars: A Systematic Literature Review

et al. 2021

View full text Add to dashboard Cite

“…Zhang et al describe a critical scenario as a dangerous road situation that may lead to an unsafe decision for the autonomous vehicle, and appropriate countermeasures must be taken immediately to avoid collision [15]. In contrast, Kluck et al focus more on the concrete scenarios level, and consider a scenario to be critical if underlying parameter values cause a malfunction of the autonomous driving system [7]. Hallerbach et al propose critical scenarios as the scenarios that need to be tested, which can be derived from both functional and non-functional requirements (e.g., traffic efficiency, driver comfort etc.)…”

Section: Critical Scenariomentioning

confidence: 99%

“…[16]. Herein we take the interpretation from Kluck et al, where critical scenarios are defined as the scenarios with parameter set that have high probability of revealing unintended and unsafe behavior of the systems, which may cause a collsion or near-collision consequence of the vehicle and other entities on the road traffic [7].…”

Section: Critical Scenariomentioning

confidence: 99%

“…Nevertheless, existing studies mostly present parts of a complete solution for critical test scenario identification, for example, focusing on either simulation or optimization of driving scenarios [7,8]. Also, the reported studies are in many cases function-specific, for example, by proposing interventions based on a particular function module, like motion-planning for the highway scenario [9].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Critical Scenario Identification for Realistic Testing of Autonomous Driving Systems

Song

Tan

Runeson

et al. 2022

Preprint

View full text Add to dashboard Cite

Autonomous driving has become an important research area for road traffic, whereas testing of autonomous driving systems to ensure a safe and reliable operation, remains an open challenge. Substantial real-world testing or massive driving data collection does not scale, as the potential test scenarios in real-world traffic are infinite and covering large shares of them in test is impractical, thus critical ones have to be prioritized. In this study, we establish a systematic approach for critical test scenario identification with integrated tools and a workflow, to explore the most critical test scenarios and facilitate testing of the autonomous driving functions. We also demonstrate the effectiveness of our approach by using two real autonomous driving systems from the industry by collaborating with Volvo Cars. Our main contribution in this work is a feasible and complete tool-chain for critical test scenario identification that is general for testing different autonomous driving systems.

show abstract

“…The genetic algorithm is efficient in optimization or parameter identification in the automotive industry. Several papers are discussing this method to find the optimal values related to suspension designs (Hemati and Shooshtari, 2019;Mitra et al, 2016) or to find optimal parameters for the Pacejka tire model (Vetturi et al, 1996) and it is effective even in the case of Advanced Driver Assistant Systems (ADAS) (Klück et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Estimating Vehicle Suspension Characteristics for Digital Twin Creation with Genetic Algorithm

Ormándi

Varga

Tettamanti

2021

Period. Polytech. Transp. Eng.

View full text Add to dashboard Cite

Usage of simulation techniques like Vehicle-in-the-Loop, Scenario-in-the-Loop, and other mixed-reality systems are becoming inevitable in autonomous vehicle development, particularly in testing and validation. These methods rely on using digital twins, realistic representations of real vehicles, and traffic in a carefully rebuilt virtual world. Recreating them precisely in a virtual ecosystem requires many parameters of real vehicles to follow their properties in a simulation. This is especially true for vehicle dynamics, where these parameters have high impact on the simulation results. The paper's objective is to provide a method that can help reverse engineering a real car's suspension characteristics with the help of a genetic algorithm. A detailed description of the method is presented, guiding the reader through the whole process, including the meta-heuristic function's settings and how it interfaces with IPG Carmaker. The paper also presents multiple measurements, which can be effortlessly recreated without expensive devices or the need to disassemble any vehicle parts. Measurements are reproduced in two separate simulation tools with special scenarios providing an efficient way to analyze and verify the results. The provided method creates vehicle suspension characteristics with adequate quality, opening up the possibility to use them in the creation of digital twins or creating virtual traffic with realistic vehicle dynamics for high-quality visualization. Results show satisfying accuracy when tested with OpenCRG.

show abstract

Genetic Algorithm-Based Test Parameter Optimization for ADAS System Testing

Cited by 49 publications

References 12 publications

Software Verification and Validation of Safe Autonomous Cars: A Systematic Literature Review

Software Verification and Validation of Safe Autonomous Cars: A Systematic Literature Review

Critical Scenario Identification for Realistic Testing of Autonomous Driving Systems

Estimating Vehicle Suspension Characteristics for Digital Twin Creation with Genetic Algorithm

Contact Info

Product

Resources

About