The modeling of dynamical systems from a time series implemented by our DSA program introduces binary trees of height D with all leaves on the same level, and the related subtrees of height L 〈 D. These are called e-...The modeling of dynamical systems from a time series implemented by our DSA program introduces binary trees of height D with all leaves on the same level, and the related subtrees of height L 〈 D. These are called e-trees and e-subtrees. The recursive and nonrecursive versions of the traversal algorithms for the trees with dynamically created nodes are discussed. The original nonrecursive algorithms that return the pointer to the next node in preorder, inorder and postorder traversals are presented. The space-time complexity analysis shows and the execution time measurements confirm that for these O（2D） algorithms, the recursive versions have approximately 10-25% better time constants. Still, the use of nonrecursive algorithms may be more appropriate in several occasions.展开更多
Due to the excellent dynamic performance,the Finite Control Set Model Predictive Control has been widely used in various types of converters.However,when Finite Control Set Model Predictive Control is adopted,the swit...Due to the excellent dynamic performance,the Finite Control Set Model Predictive Control has been widely used in various types of converters.However,when Finite Control Set Model Predictive Control is adopted,the switching frequency of converters varies significantly with system operating conditions.Consequently,constant-frequency predictive control strategy has been proposed.Two active voltage vectors and a zero voltage vector are selected within each sampling period.The action time sequence is then calculated.Due to the unsymmetrical distribution of current variation rates around zero,the calculated value of the voltage-vector action time will turn up negative.According to common sense,the voltage-vector action time is greater than or equal to zero.The action time is normally forced to zero whenever a negative value is predicted,resulting in the control failure and performance deterioration.To solve this problem,this paper proposes modified strategy.The modified strategy examines the action time calculated out.When negative action time comes out,the modified strategy reselects the active voltage vector accordingly,instead of forcing the action time to be zero.Optimized action time sequence is further determined by minimizing the cost function.The effectiveness of the modified strategy is clearly verified by experimental tests,and analytical remarks are all founded in practical results.展开更多
文摘The modeling of dynamical systems from a time series implemented by our DSA program introduces binary trees of height D with all leaves on the same level, and the related subtrees of height L 〈 D. These are called e-trees and e-subtrees. The recursive and nonrecursive versions of the traversal algorithms for the trees with dynamically created nodes are discussed. The original nonrecursive algorithms that return the pointer to the next node in preorder, inorder and postorder traversals are presented. The space-time complexity analysis shows and the execution time measurements confirm that for these O（2D） algorithms, the recursive versions have approximately 10-25% better time constants. Still, the use of nonrecursive algorithms may be more appropriate in several occasions.
基金supported in part by the National Key Basic Research Program of China ("973" Program) (Grant No. 2013CB035602)the National Natural Science Foundation of China (Grant No. 51107084)the Doctoral Program of Higher Education of China(Grant No. 20100032120081)
文摘Due to the excellent dynamic performance,the Finite Control Set Model Predictive Control has been widely used in various types of converters.However,when Finite Control Set Model Predictive Control is adopted,the switching frequency of converters varies significantly with system operating conditions.Consequently,constant-frequency predictive control strategy has been proposed.Two active voltage vectors and a zero voltage vector are selected within each sampling period.The action time sequence is then calculated.Due to the unsymmetrical distribution of current variation rates around zero,the calculated value of the voltage-vector action time will turn up negative.According to common sense,the voltage-vector action time is greater than or equal to zero.The action time is normally forced to zero whenever a negative value is predicted,resulting in the control failure and performance deterioration.To solve this problem,this paper proposes modified strategy.The modified strategy examines the action time calculated out.When negative action time comes out,the modified strategy reselects the active voltage vector accordingly,instead of forcing the action time to be zero.Optimized action time sequence is further determined by minimizing the cost function.The effectiveness of the modified strategy is clearly verified by experimental tests,and analytical remarks are all founded in practical results.
