چكيده به لاتين
The consumption of concrete pavements is dramatically raising and
its popularity is increasing. Civil engineers especially road experts recognize
that concrete pavements can be a suitable commercial rival for asphalt
pavements. In concrete pavements field, the ideal mixture proportioning ought
to be opted, because it affects strength, durability, and mechanical
characteristics of concrete. Also, mixture proportioning has the greatest impact
on the cost of concrete, so optimizing the mixture design should be considered
to reduce the amount of total cost in projects. Generally, objectives conflict with
each other and by improving performance of some of them, the performance of
others are deteriorated. Considering all the objectives above (cost, strength,
durability, and mechanical characteristics) creates a NP-hard problem and
makes the problem complicated, so multi-objective optimization is used to solve
all the problems. In this paper, it is tried to optimize multi variable functions
(total cost, flexural strength, abrasion strength, slump, drying shrinkage, and
freezing-thawing resistance) that affect the concrete features such as mechanical
characteristics, strength, durability, and cost. Materials and functions have
continues ranges, so algorithms which are suitable over continues spaces are
used in this optimization. Genetic algorithm (GA), Particle swarm optimization
(PSO), and Differential Evolution (DE) are used to optimize all functions and
trade of between them. Results show that Differential Evolution provides the
best mixture proportioning. Nonetheless, the optimal solutions which are
introduced by Genetic algorithm and Particle swarm optimization are much
more valuable than initial mixture proportions. To compare the result of
algorithms, time is one of the most important criteria. Differential Evolution is
the fastest algorithm, followed by Particle swarm optimization, and Genetic
algorithm.