The paper presents a honey badger algorithm (HB) based on a modified backwardforward sweep power flow method to determine the optimal placement of droop-controlled dispatchable distributed generations (DDG) corresponding to their sizes in an autonomous microgrid (AMG). The objectives are to minimise active power loss while considering the reduction of reactive power loss and total bus voltage deviation, and the maximisation of the voltage stability index. The proposed HB algorithm has been tested on a modified IEEE 33-bus AMG under four scenarios of the load profile at 40%, 60%, 80%, and 100% of the rated load. The analysis of the results indicates that Scenario 4, where the HB algorithm is used to optimise droop gains, the positioning of DDGs, and their reference voltage magnitudes within a permissible range, is more effective in mitigating transmission line losses than the other scenarios. Specifically, the active and reactive power losses in Scenario 4 with the HB algorithm are only 0.184% and 0.271% of the total investigated load demands, respectively. Compared to the base scenario (rated load), Scenario 4 using the HB algorithm also reduces active and reactive power losses by 41.86% and 31.54%, respectively. Furthermore, the proposed HB algorithm outperforms the differential evolution algorithm when comparing power losses for scenarios at the total investigated load and the rated load. The results obtained demonstrate that the proposed algorithm is effective in reducing power losses for the problem of optimal placement and size of DDGs in the AMG.