Preference-based cone contraction algorithms for interactive evolutionary multiple objective optimization

Kadziński, Miłosz; Tomczyk, Michał K.; Słowiński, Roman

doi:10.1016/j.swevo.2019.100602

Cited by 26 publications

(4 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The high performance of DE algorithm as one of the most powerful evolutionary algorithms has been proven in presence of dynamic constraints without getting stuck in local minima even in the time-variant problems. 41,42…”

Section: Adaptive Lane Changementioning

confidence: 99%

A novel adaptive lane change method in transient dynamic traffic conditions for highly automated vehicles

Rafat

Azadi

2021

Proceedings of the Institution of Mechanical Engineers, Part K:

View full text Add to dashboard Cite

Trajectory planning with consideration of surrounding vehicles and decision-making in the middle of complicated manoeuvres are some of the most important challenges regarding the implementation of automated driving. Since transient dynamic traffic conditions was limited to the start of the manoeuvre in previous research, no solution was provided for the surrounding vehicles’ transient changes during the manoeuvre. The algorithm presented in this paper is able to adapt to unstable variable traffic conditions and it is robust to changes in the surrounding vehicles’ conditions, even during the lane change manoeuvre. The Adaptive Lane Change algorithm provides all possible safe trajectories for any moment of manoeuvre via applying DE optimization method on a fifth-order polynomial equation. In this way, it is able to make a new decision and plan safe trajectories according to the new conditions of surrounding vehicles during the manoeuvre. Also, it guarantees collision avoidance at all-time via simultaneous longitudinal and lateral vehicle control. Improving the trajectory during a lane change manoeuvre regarding the surrounding vehicles’ conditions is considered as one of the main contributions of the presented algorithm. A second main contribution is the collision avoidance considering the vehicle’s dynamic via returning to the initial lane when there is no safe trajectory in the target lane, even during the lane change manoeuvre. The decision-making unit is evaluated by real driving tests. Then, the whole structure is simulated with MATLAB in complex transient dynamic traffic conditions via various scenarios and its performance is tested with IPG CarMaker in the presence of simulated surrounding vehicles.

show abstract

Section: Adaptive Lane Changementioning

confidence: 99%

A novel adaptive lane change method in transient dynamic traffic conditions for highly automated vehicles

Rafat

Azadi

2021

Proceedings of the Institution of Mechanical Engineers, Part K:

View full text Add to dashboard Cite

show abstract

“…Such a pattern includes setting which generation to begin interaction, the interval of subsequent consultations, the total number of consultations, and the number of PWCs to present to the DM at each consultation. Each of these tasks is non-trivial and depends on the problem type and the desires [34] provides an outline of several common patterns, while [40] and [51] (for example), use specific methods to determine when to interact during optimisation. We adopt a simple approach, and use consultations evenly spaced throughout the optimisation run (determined by a fixed number of generations), and the number of PWCs used for each consultation is limited to a maximum of four (see Table 1 for experimental details).…”

Section: Integration With Multi-objective Optimisationmentioning

confidence: 99%

Bayesian preference learning for interactive multi-objective optimisation

Taylor

et al. 2021

Proceedings of the Genetic and Evolutionary Computation Conference

View full text Add to dashboard Cite

This work proposes a Bayesian optimisation with Gaussian Process approach to learn decision maker (DM) preferences in the attribute search space of a multi-objective optimisation problem (MOP). The DM is consulted periodically during optimisation of the problem and asked to provide their preference over a series of pairwise comparisons of candidate solutions. After each consultation, the most preferred solution is used as the reference point in an appropriate multiobjective optimisation evolutionary algorithm (MOEA). The rationale for using Bayesian optimisation is to identify the most preferred location in the decision search space with the least number of DM queries, thereby minimising DM cognitive burden and fatigue. This enables non-expert DMs to be involved in the optimisation process and make more informed decisions. We further reduce the number of preference queries required, by progressively redefining the Bayesian search space to reflect the MOEA's decision bounds as it converges toward the Pareto Front. We demonstrate how this approach can locate a reference point close to an unknown preferred location on the Pareto Front, of both benchmark and real-world problems with relatively few pairwise comparisons. CCS CONCEPTS• Applied computing → Multi-criterion optimization and decision-making; • Theory of computation → Interactive computation; Active learning;

show abstract

“…On the one hand, the geometric progression allows constructing elicitation intervals of a similar size, which is desired [1]. On the other hand, depending on when the first interaction is performed, the method may suitably shift the interactions toward the early or late stages of the optimization process, adapting in this way to the difficulty of the underlying Multiple Objective Optimization (MOO) problem.…”

Section: The Implementation Details Of the ℎ Questioning Strategymentioning

confidence: 99%

On the elicitation of indirect preferences in interactive evolutionary multiple objective optimization

Tomczyk

Kadziński

2020

Proceedings of the 2020 Genetic and Evolutionary Computation Conference

Self Cite

View full text Add to dashboard Cite

Preference-based cone contraction algorithms for interactive evolutionary multiple objective optimization

Cited by 26 publications

References 52 publications

A novel adaptive lane change method in transient dynamic traffic conditions for highly automated vehicles

A novel adaptive lane change method in transient dynamic traffic conditions for highly automated vehicles

Bayesian preference learning for interactive multi-objective optimisation

On the elicitation of indirect preferences in interactive evolutionary multiple objective optimization

Contact Info

Product

Resources

About