The Research and Application on Improved Intelligence Optimization Algorithm Based on Knowledge Base

摘要

The current intelligence optimization algorithm has the limitation of slow search, stagnation and easy falling into local optimum. So the algorithm characteristic was researched, and the improved intelligence optimization algorithm based on knowledge base was proposed. The cases, experiences and rules facing different kinds of model were stored in the knowledge base, which guided intelligence optimization algorithm to generate initial state and improve search strategy. The evaluation indexes of intelligence optimization algorithm were proposed, including optimization performance, time performance and robustness performance. The Chinese Traveling Salesman Problem "CTSP" was solved by improved ant colony algorithm based on knowledge base, the result shows that the improved algorithm could get better performances. The improved algorithm could solve the problem of design, decision and scheduling more effectively.