The increased penetration of photovoltaic (PV) systems in low voltage (LV) distribution networks is causing power quality problems such as unbalanced voltage and phase's overvoltage. In the residential LV distribution networks, most of installed PV systems are single-phase rooftop PV systems. The increasing installation of single-phase rooftop PV systems at random locations with various ratings that are according to customer's request is causing PV systems unbalance distribution in the network. Usually, at noon hours that PV's active power production are in the peak, the customer's power demand are low, that is causing reverse power flow and overvoltage problems. Based on single-phase PV inverters control, three approaches are proposed to improve power quality problems in high PV penetration networks. PV's Active power production is lower than the inverters rated capacity even in PV's active power peak time at noon. So inverters always have the available capacity for reactive power control. In the first approach available capacity of inverters are used for PV inverters reactive power control and improve power quality, but in the second approach the inverters capacity increased, so reactive power control range is more than the first approach. Due to the high R/X value in LV networks, the effect of reactive power in voltage regulation and power quality improvement is limited. Therefore, the third approach proposed that by simultaneous control of reactive power and active power curtailment on PV inverters improve network power quality in such LV networks with high impedance. Centralized control method is used to control PV inverters and after optimization, optimal values of inverters active and reactive power send for PV inverters installed in the distribution network every 0.5 hour. All power quality problems caused by the increased penetration of PV systems which are voltage unbalance, voltage amplitude deviation and zero voltage sequence increase are chosen as objective functions .The multiobjective optimization problem is converted to a single-objective optimization problem using the weighting factors method, which can be effectively solved by Genetic Algorithm (GA) in Matlab optimization toolbox. The simulations are performed for a LV distribution network with high PV penetration during 24 hours and results for the three proposed approaches is analyzed and compared. Keywords : 1-Power Quality 2-Photovoltaic (PV) systems 3-Inverter Control 4-Unbalanced distribution network 5-Optimization 6-Genetic Algorithm (GA) 7-Reactive Power Control