Indoor visible light communication (VLC) based on light-emitting diodes (LED) can meet the needs of illuminance and high-rate data communications. However, the optical power and illuminance intensity distributions for the downlink of a VLC system on the same receiving plane are nonuniform owing to the LED Lambertian radiation pattern, the inherent ray multipath effect, and so on. Thus, it is difficult to ensure the fairness of communication services. An improved genetic simulated annealing algorithm (IGSAA) is proposed to search a set of optimal power regulating factors and adjust the transmitting power of LED arrays. In the IGSAA, the adaptive function is designed in the view of the received power differences among different positions on the receiving plane. A two-point crossover operation mode is used and a adaptive mutation probability is designed based on the optimal chromosome information of the population and the evolutional generation to enhance the search ability of the algorithm. Besides, to maintain the diversity of population, the probability crossover and mutation methods based on subgroup division are used to increase the population diversity. The numerical results show that the proposed IGSAA can effectively improve the coverage uniformity of an indoor VLC system within limited evolutional generations.