The use of decentralized renewable energy systems will continue to play a significant role in electricity generation especially in developing countries where grid expansion to most remote areas is uneconomical. The income levels of these off-grid communities are often low, such that there is a need for the delivery of cost-effective energy solutions through optimum control and sizing of energy system components. This paper aims at minimizing the net present cost (NPC) and the levelised cost of energy (LCOE). The study presents a hybrid power system involving a hydroelectric, solar photovoltaic (PV), and battery system for a rural community in Cameroon. The optimization of the system was done using HOMER Pro and validated using a meta-heuristic algorithm known as genetic algorithm (GA). The GA approach was programmed using the MATLAB software. After the HOMER simulation, the optimal power capacity of 3 kW solar PV, 334.89 Ah battery, and 32.2 kW microhydropower was used to meet the load. The village load profile had a daily energy usage of 431.32 kWh/day and a peak power demand of 38.49 kW. The optimized results showed an NPC and LCOE of $90,469.16 and 0.0453 $/kWh, respectively. The system configuration was tested against an increase in hydropower capacity, and it was observed that increasing the hydropower capacity has the ability to significantly reduce the LCOE as well as the battery and solar PV size. A comparative analysis of the two approaches showed that the optimization using GA was more cost-effective than HOMER Pro with the least LCOE of 0.0344 $/kWh and NPC of $86,990.94 as well as a loss of power supply probability (LPSP) of 0.99%. In addition, the GA method gave more hydropower generation than HOMER Pro. This supports the fact that stochastic methods are more realistic and economically viable. They also accurately predict system operation than deterministic methods.