Improved distribution system planning using computational evolution

This paper presents a new problem solving environment that can utilize different forms of resources to approach better solutions in the distribution network planning domain. To implement this concept, a design optimization framework, named design by expectation (DBE), is developed. DBE exploits the adaptation of genetic algorithms in searching the optimal solution, and releases the strict requirements for objective functions and constraints in conventional optimization techniques. One example, street lighting design, is used to demonstrate the applicability of the proposed approach.