Control allocation considers the problem of controlling instruction distribution for control systems with multiple and redundant actuators. This paper focuses on the direct allocation method, making the time requireme...Control allocation considers the problem of controlling instruction distribution for control systems with multiple and redundant actuators. This paper focuses on the direct allocation method, making the time requirement of the algorithm analogous compared with modified pseudoinverse redistribution methods, linear programming methods solved by simplex method, and sub-gradient optimization method. To reduce off-line computations of constructing the attainable moment set of actuators, a new approach based on the null space of the control effectiveness matrix is proposed, which is superior when the number of actuators is less than 10 compared with traditional method. To decrease on-line computations, an improvement method of searching the facet that is aligned with the desired moment is presented, shortening the search time by checking only the facets that lie around the desired moment. To find such facets, the vertices of the attainable moment set are normalized and saved during off-line computations. Simulation results show that at least 32.22% of off-line computation time would be saved using null space-based construction when the number of actuators is less than 10. In on-line computations, the modified method performs superiorly compared with the three aforementioned methods. Furthermore, it may solve the problem of control allocation efficiently when a remarkable large number of redundant actuators are configured.展开更多
This paper proposes a mixed integer programming model for the allocation of rail mounted gantry cranes for four basic yard activities with different priorities. The model pays special attention to the typical features...This paper proposes a mixed integer programming model for the allocation of rail mounted gantry cranes for four basic yard activities with different priorities. The model pays special attention to the typical features of this kind of gantry cranes, such as a restricted traveling range and a limited number of adjustments during loading and discharging operations. In contrast to most of the literature dealing with these four yard activities individually, this paper models them into an integrated problem, whose computational complexity is proved to be NP-hard. We are therefore motivated to develop a Lagrangian relaxation-based heuristic to solve the problem. We compare the proposed heuristic with the branch-and-bound method that uses commercial software packages. Extensive computational results show that the proposed heuristic achieves competitive solution qualities for solving the tested problems.展开更多
We investigate a collaborative-relay beamforming design for simultaneous wireless information and power transfer. A non-robust beamforming design that assumes availability of perfect channel state information(CSI) in ...We investigate a collaborative-relay beamforming design for simultaneous wireless information and power transfer. A non-robust beamforming design that assumes availability of perfect channel state information(CSI) in the relay nodes is addressed. In practical scenarios, CSI errors are usually inevitable; therefore, a robust collaborativerelay beamforming design is proposed. By applying the bisection method and the semidefinite relaxation(SDR)technique, the non-convex optimization problems of both non-robust and robust beamforming designs can be solved.Moreover, the solution returned by the SDR technique may not always be rank-one; thus, an iterative sub-gradient method is presented to acquire the rank-one solution. Simulation results show that under an imperfect CSI case, the proposed robust beamforming design can obtain a better performance than the non-robust one.展开更多
The resource allocation scheme for the multiple description coding multicast （MDCM） in orthogonal frequency division multiplexing （OFDM-based） cognitive radio network （CRN） is studied in this paper, aiming at ma...The resource allocation scheme for the multiple description coding multicast （MDCM） in orthogonal frequency division multiplexing （OFDM-based） cognitive radio network （CRN） is studied in this paper, aiming at maximizing the total throughput of cognitive radio （CR） users, with constraints on sum transmit power, the maximal receiving rate of each CR user and the maximal total interference introduced to each primary user. With the analysis of the model, an algorithm, which consists of subcarrier assignment and power allocation using the sub-gradient updating method, is proposed. Meanwhile, to reduce the complexity, a suboptimal algorithm is also proposed, which divides the total transmit power into small slices and allocates them one by one. Moreover, the suboptimal algorithm is modified by adding an advanced water-filling process to improve the performance. The simulation results obtained in this paper show that the system throughput using the MDCM scheme is much higher than that using the conventional multicast （CM） scheme and the performance of the proposed suboptimal algorithms can approximate the MDCM scheme very well.展开更多
Soft margin support vector machine(SVM)with hinge loss function is an important classification algorithm,which has been widely used in image recognition,text classification and so on.However,solving soft margin SVM wi...Soft margin support vector machine(SVM)with hinge loss function is an important classification algorithm,which has been widely used in image recognition,text classification and so on.However,solving soft margin SVM with hinge loss function generally entails the sub-gradient projection algorithm,which is very time-consuming when processing big training data set.To achieve it,an efficient quantum algorithm is proposed.Specifically,this algorithm implements the key task of the sub-gradient projection algorithm to obtain the classical sub-gradients in each iteration,which is mainly based on quantum amplitude estimation and amplification algorithm and the controlled rotation operator.Compared with its classical counterpart,this algorithm has a quadratic speedup on the number of training data points.It is worth emphasizing that the optimal model parameters obtained by this algorithm are in the classical form rather than in the quantum state form.This enables the algorithm to classify new data at little cost when the optimal model parameters are determined.展开更多
基金National Natural Science Foundation of China （NSFC60704020） Changjiang Scholars and Innovative Research Team of China （PCSIRT0520） Research Fund for the Doctoral Program of Higher Education of China （20070213068）
文摘Control allocation considers the problem of controlling instruction distribution for control systems with multiple and redundant actuators. This paper focuses on the direct allocation method, making the time requirement of the algorithm analogous compared with modified pseudoinverse redistribution methods, linear programming methods solved by simplex method, and sub-gradient optimization method. To reduce off-line computations of constructing the attainable moment set of actuators, a new approach based on the null space of the control effectiveness matrix is proposed, which is superior when the number of actuators is less than 10 compared with traditional method. To decrease on-line computations, an improvement method of searching the facet that is aligned with the desired moment is presented, shortening the search time by checking only the facets that lie around the desired moment. To find such facets, the vertices of the attainable moment set are normalized and saved during off-line computations. Simulation results show that at least 32.22% of off-line computation time would be saved using null space-based construction when the number of actuators is less than 10. In on-line computations, the modified method performs superiorly compared with the three aforementioned methods. Furthermore, it may solve the problem of control allocation efficiently when a remarkable large number of redundant actuators are configured.
基金supported by the National Nature Science Foundation of China under grant numbers 71102011and 51105394Guangdong provincial department of science and technology(Number 2011B090400384)
文摘This paper proposes a mixed integer programming model for the allocation of rail mounted gantry cranes for four basic yard activities with different priorities. The model pays special attention to the typical features of this kind of gantry cranes, such as a restricted traveling range and a limited number of adjustments during loading and discharging operations. In contrast to most of the literature dealing with these four yard activities individually, this paper models them into an integrated problem, whose computational complexity is proved to be NP-hard. We are therefore motivated to develop a Lagrangian relaxation-based heuristic to solve the problem. We compare the proposed heuristic with the branch-and-bound method that uses commercial software packages. Extensive computational results show that the proposed heuristic achieves competitive solution qualities for solving the tested problems.
基金supported by the National Natural Science Foundation of China(No.61601295)the Zhejiang Provincial Natural Science Foundation of China(No.LY18F030015)
文摘We investigate a collaborative-relay beamforming design for simultaneous wireless information and power transfer. A non-robust beamforming design that assumes availability of perfect channel state information(CSI) in the relay nodes is addressed. In practical scenarios, CSI errors are usually inevitable; therefore, a robust collaborativerelay beamforming design is proposed. By applying the bisection method and the semidefinite relaxation(SDR)technique, the non-convex optimization problems of both non-robust and robust beamforming designs can be solved.Moreover, the solution returned by the SDR technique may not always be rank-one; thus, an iterative sub-gradient method is presented to acquire the rank-one solution. Simulation results show that under an imperfect CSI case, the proposed robust beamforming design can obtain a better performance than the non-robust one.
基金supported by the National Basic Research Program of China(2009CB320401)the National Natural Science Foundation of China(61101117)+2 种基金the Research Funds for Doctoral Program of Higher Education of China(20090005110003)the National Key Scientific and Technological Project of China(2010ZX03003-001, 2012ZX03004005002)the Fundamental Research Funds for the Central Universities(BUPT2012RC0112)
文摘The resource allocation scheme for the multiple description coding multicast （MDCM） in orthogonal frequency division multiplexing （OFDM-based） cognitive radio network （CRN） is studied in this paper, aiming at maximizing the total throughput of cognitive radio （CR） users, with constraints on sum transmit power, the maximal receiving rate of each CR user and the maximal total interference introduced to each primary user. With the analysis of the model, an algorithm, which consists of subcarrier assignment and power allocation using the sub-gradient updating method, is proposed. Meanwhile, to reduce the complexity, a suboptimal algorithm is also proposed, which divides the total transmit power into small slices and allocates them one by one. Moreover, the suboptimal algorithm is modified by adding an advanced water-filling process to improve the performance. The simulation results obtained in this paper show that the system throughput using the MDCM scheme is much higher than that using the conventional multicast （CM） scheme and the performance of the proposed suboptimal algorithms can approximate the MDCM scheme very well.
基金supported by the Beijing Natural Science Foundation(4222031)the National Natural Science Foundation of China(61976024,61972048)Beijing University of Posts and Telecommunications(BUPT)Innovation and Entrepreneurship Support Program(2021-YC-A206)
文摘Soft margin support vector machine(SVM)with hinge loss function is an important classification algorithm,which has been widely used in image recognition,text classification and so on.However,solving soft margin SVM with hinge loss function generally entails the sub-gradient projection algorithm,which is very time-consuming when processing big training data set.To achieve it,an efficient quantum algorithm is proposed.Specifically,this algorithm implements the key task of the sub-gradient projection algorithm to obtain the classical sub-gradients in each iteration,which is mainly based on quantum amplitude estimation and amplification algorithm and the controlled rotation operator.Compared with its classical counterpart,this algorithm has a quadratic speedup on the number of training data points.It is worth emphasizing that the optimal model parameters obtained by this algorithm are in the classical form rather than in the quantum state form.This enables the algorithm to classify new data at little cost when the optimal model parameters are determined.
