UAV-UGV-UMV Multi-Swarms for Cooperative Surveillance

In this paper we present a surveillance system for early detection of escapers from a restricted area based on a new swarming mobility model called CROMM-MS (Chaotic Rössler Mobility Model for Multi-Swarms). CROMM-MS is designed for controlling the trajectories of heterogeneous multi-swarms of aerial, ground and marine unmanned vehicles with important features such as prioritising early detections and success rate. A new Competitive Coevolutionary Genetic Algorithm (CompCGA) is proposed to optimise the vehicles’ parameters and escapers’ evasion ability using a predator-prey approach. Our results show that CROMM-MS is not only viable for surveillance tasks but also that its results are competitive in regard to the state-of-the-art approaches.

Nesting Scheduling in Sheet Metal Processing Based on Coevolutionary Genetic Algorithm in Different Environments

In sheet metal processing, nesting and scheduling are important factors affecting the efficiency and agility of manufacturing. The objective of nesting is to minimize the waste of material, while that of scheduling is to optimize the processing sequence. As the relation between them often becomes a trade-off, they should be considered simultaneously for the efficiency of the total manufacturing process. In this study, we propose a co-evolutionary genetic algorithm-based nesting scheduling method. We first define a cost function as a fitness value, and then we propose a grouping method that forms gene groups based on the processing layout and processing time. Finally, we validate the effectiveness of the proposed method through computational experiments.

Cooperative Coevolution with Dynamic Species-Size Strategy for Vibration-Based Damage Detection in Plates

Vibration-based damage detection is based on the fact that vibration characteristics such as natural frequencies and mode shapes of structures are changed when the damage occurs. This paper proposes dynamic species-size strategy in cooperative coevolution concept. The resulting algorithm, cooperative coevolutionary genetic algorithm with dynamics species-size (CCGADSS), is used as the optimization algorithm in the vibration-based damage detection in plates. The objective function is numerically calculated from the difference between experimentally vibration characteristics and numerically evaluated vibration characteristics of the predicted damage. In finite element model for objective calculation, the plates are equally divided into 64 elements. There are 2 different cases with dissimilar occurred damage in plates are considered. In first case, the plate hase only one region consisting of 4 elements which are together connected and have same damage. In second case, there are 5 separated elements which are damaged differently. In order to demonstrate the performance of the dynamic species-size strategy, 3 optimization algorithms, which are genetic algorithm (GA), cooperative coevolutionary genetic algorithm (CCGA), and CCGADSS. The results indicate that CCGADSS is superior to GA and CCGA. Moreover solutions obtained using CCGADSS are quite close the actual damage. These results show that the dynamic species-size strategy can enhance performance of cooperative coevolution concept.