Planning is a class of complex problem. It is a way to improve the efficiency of planning algorithm in extracting and using goal orderings. Because deciding goal orderings is also PSPACE-complete, it is necessary to extract goal orderings efficiently when using goal orderings. The paper presents a method, called GOWN (goal ordering with invariants) and uses state invariants to extract goal orderings. During the process of ordering, abstraction and unification are utilized to control the increase of problem size that improves the efficiency of ordering.