We have recently developed a quantum genetic algorithm which has both quantum crossover and mutation operations [arXiv: 1202.2026 (cs.NE)]. Improvements have been made on each of genetic operations so that they are substantially different from those of existing quantum genetic algorithms. For instance, we have tailored the inversion-about-average operation of the Grover routine by introducing a pseudo-scrambler so as to reduce the internal cost. In this talk, we are going to present the details of improvements.