Solving constrained optimization problems based on rBOA

This paper combined real-coded Bayesian optimization algorithm (rBOA) with constraint-handling techniques to solve constrained optimization problems. The main idea was as follows: constraints were used to influence the selection of better individuals, which could guide the probability model to generate solutions satisfying constrained conditions indirectly. As a result, the optimal solution could be obtained when the evolution halts. The experimental results suggest that the method which integrates multi-objective optimization technique with rBOA shows the best performance among the compared methods. However, it is still difficult to find optimal solutions for the test functions. Hence, the capability of this method has plenty of space for improvement.