A significant amount of interest from the scientific community has been focused on Direct Current (DC) microgrids in recent years, as a direct result of the proliferation of appliances that run on DC power. Nevertheless, the acceptability of DC microgrids by power utilities is still restricted owing to the challenges connected with the construction of a dependable protection system. This is because obtaining dependable protection for DC microgrids might be difficult due to the large size of DC fault current, its quick rate of rise, and the lack of zero crossing. In addition, the intermittent nature of the power produced by non-traditional distributed generators necessitates adaptability in response to a wide range of climatic circumstances. The method used in this study to solve the problems outlined above is developing an Ensemble of Decision Tree-based protection solutions for a multi-terminal DC microgrid. The plan developed accommodates the intermittent nature of renewable energy sources. It only uses information that is local to the current and voltage signals, which allows it to execute the duties of fault detection and classification, and prevents the disadvantages associated with communication networks, such as data loss and delay.