Motion Coordination for Construction Vehicles using Swarm Intelligence

In a variety of tasks performed in construction sites, coordinated operations of multi-vehicles are foreshadowed to outperform the deployment of a single vehicle in terms of increased capacity and flexibility. This paper presents the application of the particle swarm optimization (PSO) algorithm in deriving drive commands, speed and turning, for the vehicles such that they are steered into and maintained in desirable formations according to an assigned task. The PSO is adopted for its implementation simplicity and relaxing the need for analytical system models. To this end, the coordination of vehicles is posed as an optimization problem minimizing the translational and angular errors between the current vehicle positions and their corresponding targets. Inter-vehicle collisions are mitigated, in this work, by employing a behavioural-based reactive scheme together with a dynamical index rescheduling procedure. Simulation results for coordinated multi-vehicle motions, in benchmark formation patterns, are included to demonstrate the effectiveness of the proposed approach.

Bouc-Wen model parameter identification for a MR fluid damper using computationally efficient GA

A non-symmetrical Bouc-Wen model is proposed in this paper for magnetorheological (MR) fluid dampers. The model considers the effect of non-symmetrical hysteresis which has not been taken into account in the original Bouc-Wen model. The model parameters are identified with a Genetic Algorithm (GA) using its flexibility in identification of complex dynamics. The computational efficiency of the proposed GA is improved with the absorption of the selection stage into the crossover and mutation operations. Crossover and mutation are also made adaptive to the fitness values such that their probabilities need not be user-specified. Instead of using a sufficiently number of generations or a pre-determined fitness value, the algorithm termination criterion is formulated on the basis of a statistical hypothesis test, thus enhancing the performance of the parameter identification. Experimental test data of the damper displacement and force are used to verify the proposed approach with satisfactory parameter identification results.