We consider the sparse identification of multivariate ARX systems, i.e., to recover the zero elements of the unknown parameter matrix. We propose a two-step algorithm, where in the first step the stochastic gradient (...We consider the sparse identification of multivariate ARX systems, i.e., to recover the zero elements of the unknown parameter matrix. We propose a two-step algorithm, where in the first step the stochastic gradient (SG) algorithm is applied to obtain initial estimates of the unknown parameter matrix and in the second step an optimization criterion is introduced for the sparse identification of multivariate ARX systems. Under mild conditions, we prove that by minimizing the criterion function, the zero elements of the unknown parameter matrix can be recovered with a finite number of observations. The performance of the algorithm is testified through a simulation example.展开更多
In power generation industries,boilers are required to be operated under a range of different conditions to accommodate demands for fuel randomness and energy fluctuation.Reliable prediction of the combustion operatio...In power generation industries,boilers are required to be operated under a range of different conditions to accommodate demands for fuel randomness and energy fluctuation.Reliable prediction of the combustion operation condition is crucial for an in-depth understanding of boiler performance and maintaining high combustion efficiency.However,it is difficult to establish an accurate prediction model based on traditional data-driven methods,which requires prior expert knowledge and a large number of labeled data.To overcome these limitations,a novel prediction method for the combustion operation condition based on flame imaging and a hybrid deep neural network is proposed.The proposed hybrid model is a combination of convolutional sparse autoencoder(CSAE)and least support vector machine(LSSVM),i.e.,CSAE-LSSVM,where the convolutional sparse autoencoder with deep architectures is utilized to extract the essential features of flame image,and then essential features are input into the least support vector machine for operation condition prediction.A comprehensive investigation of optimal hyper-parameter and dropout technique is carried out to improve the performance of the CSAE-LSSVM.The effectiveness of the proposed model is evaluated by 300 MW tangential coal-fired boiler flame images.The prediction accuracy of the proposed hybrid model reaches 98.06%,and its prediction time is 3.06 ms/image.It is observed that the proposed model could present a superior performance in comparison to other existing neural network models.展开更多
Joint sparse recovery(JSR)in compressed sensing(CS)is to simultaneously recover multiple jointly sparse vectors from their incomplete measurements that are conducted based on a common sensing matrix.In this study,the ...Joint sparse recovery(JSR)in compressed sensing(CS)is to simultaneously recover multiple jointly sparse vectors from their incomplete measurements that are conducted based on a common sensing matrix.In this study,the focus is placed on the rank defective case where the number of measurements is limited or the signals are significantly correlated with each other.First,an iterative atom refinement process is adopted to estimate part of the atoms of the support set.Subsequently,the above atoms along with the measurements are used to estimate the remaining atoms.The estimation criteria for atoms are based on the principle of minimum subspace distance.Extensive numerical experiments were performed in noiseless and noisy scenarios,and results reveal that iterative subspace matching pursuit(ISMP)outperforms other existing algorithms for JSR.展开更多
提出基于特征向量选择(feature vector selection,FVS)的稀疏最小二乘支持向量机(sparse least squares support vector machine,SLS-SVM)模型,解决最小二乘支持向量机(least squares support vector machine,LS-SVM)稀疏化问题。采用FV...提出基于特征向量选择(feature vector selection,FVS)的稀疏最小二乘支持向量机(sparse least squares support vector machine,SLS-SVM)模型,解决最小二乘支持向量机(least squares support vector machine,LS-SVM)稀疏化问题。采用FVS在特征空间构建特征向量子集,对训练样本进行稀疏线性重构;将稀疏化的特征向量作为支持向量,从而实现对LS-SVM稀疏化建模。将SLS-SVM模型进行弓网系统的仿真对比实验,结果表明SLS-SVM模型在取得高预报精度的同时,可实现支持向量的高度稀疏化,从而加快模型预报速度。展开更多
This paper presents a novel adaptive wide-band compressed spectrum sensing scheme for cognitive radio(CR)networks.Compared to the traditional CSS-based CR scenarios,the proposed approach reconstructs neither the recei...This paper presents a novel adaptive wide-band compressed spectrum sensing scheme for cognitive radio(CR)networks.Compared to the traditional CSS-based CR scenarios,the proposed approach reconstructs neither the received signal nor its spectrum during the compressed sensing procedure.On the contrary,a precise estimation of wide spectrum support is recovered with a fewer number of compressed measurements.Then,the spectrum occupancy is determined directly from the reconstructed support vector.To carry out this process,a data-driven methodology is utilized to obtain the mini-mum number of necessary samples required for support reconstruction,and a closed-form expression is obtained that optimally estimates the number of desired samples as a function of the sparsity level and number of channels.Following this phase,an adjustable sequential framework is developed where the first step predicts the optimal number of compressed measurements and the second step recovers the sparse support and makes sensing decision.Theoretical analysis and numerical simulations demonstrate the improvement achieved with the proposed algorithm to significantly reduce both sampling costs and average sensing time without any deterioration in detection performance.Furthermore,the remainder of the sensing time can be employed by secondary users for data transmission,thus leading to the enhancement of the total throughput of the CR network.展开更多
文摘We consider the sparse identification of multivariate ARX systems, i.e., to recover the zero elements of the unknown parameter matrix. We propose a two-step algorithm, where in the first step the stochastic gradient (SG) algorithm is applied to obtain initial estimates of the unknown parameter matrix and in the second step an optimization criterion is introduced for the sparse identification of multivariate ARX systems. Under mild conditions, we prove that by minimizing the criterion function, the zero elements of the unknown parameter matrix can be recovered with a finite number of observations. The performance of the algorithm is testified through a simulation example.
基金supported by the National Natural Science Foundation of China(Grant No.51976038)the Natural Science Foundation of Jiangsu Province,China for Young Scholars(Grant No.BK20190366)the China Scholarship Council(Grant No.202006090164).
文摘In power generation industries,boilers are required to be operated under a range of different conditions to accommodate demands for fuel randomness and energy fluctuation.Reliable prediction of the combustion operation condition is crucial for an in-depth understanding of boiler performance and maintaining high combustion efficiency.However,it is difficult to establish an accurate prediction model based on traditional data-driven methods,which requires prior expert knowledge and a large number of labeled data.To overcome these limitations,a novel prediction method for the combustion operation condition based on flame imaging and a hybrid deep neural network is proposed.The proposed hybrid model is a combination of convolutional sparse autoencoder(CSAE)and least support vector machine(LSSVM),i.e.,CSAE-LSSVM,where the convolutional sparse autoencoder with deep architectures is utilized to extract the essential features of flame image,and then essential features are input into the least support vector machine for operation condition prediction.A comprehensive investigation of optimal hyper-parameter and dropout technique is carried out to improve the performance of the CSAE-LSSVM.The effectiveness of the proposed model is evaluated by 300 MW tangential coal-fired boiler flame images.The prediction accuracy of the proposed hybrid model reaches 98.06%,and its prediction time is 3.06 ms/image.It is observed that the proposed model could present a superior performance in comparison to other existing neural network models.
基金supported by the National Natural Science Foundation of China(61771258)the Postgraduate Research and Practice Innovation Program of Jiangsu Province(KYCX 210749)。
文摘Joint sparse recovery(JSR)in compressed sensing(CS)is to simultaneously recover multiple jointly sparse vectors from their incomplete measurements that are conducted based on a common sensing matrix.In this study,the focus is placed on the rank defective case where the number of measurements is limited or the signals are significantly correlated with each other.First,an iterative atom refinement process is adopted to estimate part of the atoms of the support set.Subsequently,the above atoms along with the measurements are used to estimate the remaining atoms.The estimation criteria for atoms are based on the principle of minimum subspace distance.Extensive numerical experiments were performed in noiseless and noisy scenarios,and results reveal that iterative subspace matching pursuit(ISMP)outperforms other existing algorithms for JSR.
文摘提出基于特征向量选择(feature vector selection,FVS)的稀疏最小二乘支持向量机(sparse least squares support vector machine,SLS-SVM)模型,解决最小二乘支持向量机(least squares support vector machine,LS-SVM)稀疏化问题。采用FVS在特征空间构建特征向量子集,对训练样本进行稀疏线性重构;将稀疏化的特征向量作为支持向量,从而实现对LS-SVM稀疏化建模。将SLS-SVM模型进行弓网系统的仿真对比实验,结果表明SLS-SVM模型在取得高预报精度的同时,可实现支持向量的高度稀疏化,从而加快模型预报速度。
文摘This paper presents a novel adaptive wide-band compressed spectrum sensing scheme for cognitive radio(CR)networks.Compared to the traditional CSS-based CR scenarios,the proposed approach reconstructs neither the received signal nor its spectrum during the compressed sensing procedure.On the contrary,a precise estimation of wide spectrum support is recovered with a fewer number of compressed measurements.Then,the spectrum occupancy is determined directly from the reconstructed support vector.To carry out this process,a data-driven methodology is utilized to obtain the mini-mum number of necessary samples required for support reconstruction,and a closed-form expression is obtained that optimally estimates the number of desired samples as a function of the sparsity level and number of channels.Following this phase,an adjustable sequential framework is developed where the first step predicts the optimal number of compressed measurements and the second step recovers the sparse support and makes sensing decision.Theoretical analysis and numerical simulations demonstrate the improvement achieved with the proposed algorithm to significantly reduce both sampling costs and average sensing time without any deterioration in detection performance.Furthermore,the remainder of the sensing time can be employed by secondary users for data transmission,thus leading to the enhancement of the total throughput of the CR network.
