Impulse noise removal is an important task in image restoration.In this paper,we introduce a general nonsmooth nonconvex model for recovering images degraded by blur and impulsive noise,which can easily include some p...Impulse noise removal is an important task in image restoration.In this paper,we introduce a general nonsmooth nonconvex model for recovering images degraded by blur and impulsive noise,which can easily include some prior information,such as box constraint or low rank,etc.To deal with the nonconvex problem,we employ the proximal linearized minimization algorithm.For the subproblem,we use the alternating direction method of multipliers to solve it.Furthermore,based on the assumption that the objective function satisfies the KurdykaLojasiewicz property,we prove the global convergence of the proposed algorithm.Numerical experiments demonstrate that our method outperforms both the l1TV and Nonconvex TV models in terms of subjective and objective quality measurements.展开更多
Digital twin is considered the key technique for real-time monitoring and life-cycle management of electric equipment.To construct the digital twin model of electric equipment,a multi-parameter electromagnetic analysi...Digital twin is considered the key technique for real-time monitoring and life-cycle management of electric equipment.To construct the digital twin model of electric equipment,a multi-parameter electromagnetic analysis is needed to generate a large amount of high fidelity data under various working condition.However,repeated solving such multi-parameter electromagnetic problems based on full order finite element method may lead to extreme scale calculations.To address this issue,a hybrid approach that combines tensor decomposition and proper orthogonal decomposition(POD)is introduced,which can effectively establish a reduced order model for multi-parameter electromagnetic field problems.The performance of the proposed approach is illustrated through three numerical examples,namely an electrical motor,a transformer and a voice coil actuator including parameter variations of operating conditions,geometric parameters,and material parameters.The numerical results show that the proposed hybrid approach has significant advantage compared to conventional reduced order model in situation where the solution changes dramatically within the parameter variation range and even more apparent when the parameter dimension is high.展开更多
基金Supported by the National Natural Science Foundations of China(Grant No.12061045,12031003)the Guangzhou Education Scientific Research Project 2024(Grant No.202315829)the Natural Science Foundation of Jiangxi Province(Grant No.20224ACB211004)。
文摘Impulse noise removal is an important task in image restoration.In this paper,we introduce a general nonsmooth nonconvex model for recovering images degraded by blur and impulsive noise,which can easily include some prior information,such as box constraint or low rank,etc.To deal with the nonconvex problem,we employ the proximal linearized minimization algorithm.For the subproblem,we use the alternating direction method of multipliers to solve it.Furthermore,based on the assumption that the objective function satisfies the KurdykaLojasiewicz property,we prove the global convergence of the proposed algorithm.Numerical experiments demonstrate that our method outperforms both the l1TV and Nonconvex TV models in terms of subjective and objective quality measurements.
基金supported by the UCAS-ULille joint PhD training program and also partially supported within the frame of the EE4.0(Electrical Energy 4.0)project,which is co-financed by the European Union,with the financial support of the European Regional Development Fund(ERDF),the French state,and the French Region of Hautsde-France.
文摘Digital twin is considered the key technique for real-time monitoring and life-cycle management of electric equipment.To construct the digital twin model of electric equipment,a multi-parameter electromagnetic analysis is needed to generate a large amount of high fidelity data under various working condition.However,repeated solving such multi-parameter electromagnetic problems based on full order finite element method may lead to extreme scale calculations.To address this issue,a hybrid approach that combines tensor decomposition and proper orthogonal decomposition(POD)is introduced,which can effectively establish a reduced order model for multi-parameter electromagnetic field problems.The performance of the proposed approach is illustrated through three numerical examples,namely an electrical motor,a transformer and a voice coil actuator including parameter variations of operating conditions,geometric parameters,and material parameters.The numerical results show that the proposed hybrid approach has significant advantage compared to conventional reduced order model in situation where the solution changes dramatically within the parameter variation range and even more apparent when the parameter dimension is high.
