An Interval Optimization Algorithm with Embedded Point Evolutionary Strategy and Its Application to Bounded Error Modeling

Aiming at the problems of low efficiency and difficulty in constructing acceleration devices traditional interval optimization algorithms (IOAs), this paper constructs a valid device based on more concise point evolutionary strategy (ES), then proposes novel hybrid IOA (HIOA) with no requirement derivative objective function. The HIOA first divides initial search area into N equal parts, randomly selects multiple individuals each subinterval to represent their information, performs fewer iterations using ES for all make them closer optima; reliable subintervals containing split, deletes unreliable without any individuals; finally, provides upper bound direct pruning operation further improve efficiency. Furthermore, convergence property proposed algorithm is analyzed. Extensive numerical experiments several typical test functions application bounded error parameter estimation demonstrate superiority by comparing it existing conventional algorithms, which confirms effectiveness applicability suggested algorithm.