This paper proposes an architecture for a planning optimization platform called the Grid of Grids Optimal Designer (GGOD) to facilitate the optimization of cluster type micro-grid installation and operation. Micro-grid clusters are groups of micro-grids with renewable energy that operate autonomously without power supply from main grids. The GGOD simulates micro-grid optimizations based on a service-oriented architecture and comprises three layers, i.e., end user, service broker, and service supply layers. The enduser layer corresponds to the client system, and the GGOD corresponds to the service broker and service supply layers. The service broker layer is for service management, and the service supply layer comprises various optimization components. Combinable optimization components are activated based on request text in extensible markup language format from end users. By combining optimization components, different optimizations are executed for a single request, the results of the same type of optimization methods are compared, and sudden operation environment changes are considered. Geospatial integrated resource planning and micro-grid facility planning are demonstrated as examples of GGOD applications.