The development of a hybrid knowledge-based system with a coupling between knowledge-based and numerical methods for multiobjective optimization of power distribution operations is described. The advantages of a hybrid knowledge-based system are described followed by the system objectives, means of control, and constraints. A framework is provided that describes the necessary development stages of a commercial knowledge-based package. An overview of the utility knowledge acquisition procedure is provided to appreciate the complexity of defining the rule base. This is followed by a description of the flow of information in a three-level hierarchical rule base and a summary of network radiality, parameter, and performance rules employed in this rule base. After a heuristic preprocessor identifies a list of switch closures that would seem to reduce total system losses, network radiality rules assess if a particular search path has identified a switch that can be closed and a corresponding switch opened to maintain the radiality of the system or if the path is worth pursuing further. Network parameter rules ensure the system operates within original design parameters. Network performance rules assess the reduction in total system losses of each proposed switching operation. Where there is a coupling between knowledge-based and numerical methods, the integration of numerical methods is described. Finally, the validation and simulations as well as the benefits of this hybrid knowledge-based system are described.