A bio-retention system is an important measure for non-point source pollution
control. In order to improve the calculation precision for parameter
optimization of the moisture movement in a bio-retention system, a
real-encoded genetic algorithm based on the fractional-order operation is
proposed, in which initial populations are generated by random mapping, and
the searching range is automatically renewed with the excellent individuals
by fractional-order particle swarm optimization operation. Its efficiency is
verified experimentally. The results indicate that the absolute error by
the fractional-order operation decreases by 67.73%, 62.23%, and 4.16%, and
the relative error decreases by 42.88%, 35.76%, and 6.77%, respectively,
compared to those by the standard binary-encoded genetic algorithm, random
algorithm, and the particle swarm optimization algorithm. The
fractional-order operation has higher precision and it is good for the
practical parameter optimization in ecological environment systems.