Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic an...Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic and lightweight SR framework designed for arbitrary scaling factors.DDNet integrates a residual learning structure with an Adaptively fusion Feature Block(AFB)and a scale-aware upsampling module,effectively reducing parameter overhead while preserving reconstruction quality.Additionally,we introduce DDNetGAN,an enhanced variant that leverages a relativistic Generative Adversarial Network(GAN)to further improve texture realism.To validate the proposed models,we conduct extensive training using the DIV2K and Flickr2K datasets and evaluate performance across standard benchmarks including Set5,Set14,Urban100,Manga109,and BSD100.Our experiments cover both symmetric and asymmetric upscaling factors and incorporate ablation studies to assess key components.Results show that DDNet and DDNetGAN achieve competitive performance compared with mainstream SR algorithms,demonstrating a strong balance between accuracy,efficiency,and flexibility.These findings highlight the potential of our approach for practical real-world super-resolution applications.展开更多
Compared with vertical and horizontal wells, the solution and computation of transient pressure responses of slanted wells are more complex. Vertical and horizontal wells are both simplified cases of slanted wells at ...Compared with vertical and horizontal wells, the solution and computation of transient pressure responses of slanted wells are more complex. Vertical and horizontal wells are both simplified cases of slanted wells at particular inclination, so the model for slanted wells is more general and more complex than other models for vertical and horizontal wells. Many authors have studied unsteady-state flow of fluids in slanted wells and various solutions have been proposed. However, until now, few of the published results pertain to the computational efficiency. Whether in the time domain or in the Laplace domain, the computation of integration of complex functions is necessary in obtaining pressure responses of slanted wells, while the computation of the integration is complex and time-consuming. To obtain a perfect type curve the computation time is unacceptable even with an aid of high-speed computers. The purpose of this paper is to present an efficient algorithm to compute transient pressure distributions caused by slanted wells in reservoirs. Based on rigorous derivation, the transient pressure solution for slanted wells of any inclination angle is presented. Assuming an infinite-conductivity wellbore, the location of the equivalent-pressure point is determined. More importantly, according to the characteristics of the integrand in a transient pressure solution for slanted wells, the whole integral interval is partitioned into several small integral intervals, and then the method of variable substitution and the variable step-size piecewise numerical integration are employed. The amount of computation is significantly reduced and the computational efficiency is greatly improved. The algorithm proposed in this paper thoroughly solved the difficulty in the efficient and high-speed computation of transient pressure distribution of slanted wells with any inclination angle.展开更多
This paper investigates the synchronization problem of clock oscillators in nonlinear dynamical network with arbitrary time-delays.First,a dynamic synchronization algorithm based on consensus control strategy,named fa...This paper investigates the synchronization problem of clock oscillators in nonlinear dynamical network with arbitrary time-delays.First,a dynamic synchronization algorithm based on consensus control strategy,named fast averaging syn-chronization algorithm(FASA),is presented to find a solution to the synchronization problem.This algorithm can compensate the clock skew and offset differences between clock nodes,achieving the synchronization of clock nodes in a shorter time as compared to previous synchronization methods.Second,because of the dynamical performance of FASA,it is characterized from the perspective of compartmental dynamical system with arbitrary time-delays.In this case,the algorithm guarantees the states of all clock nodes in dynamical network converge to Lyapunov stable equilibria.Finally,numerical simulations and experimental results demonstrate the correctness and effciency of the FASA,which means that the clock nodes can reach global consensus,and the synchronization error can reach nanosecond order of magnitude.展开更多
A memory compress algorithm for 12\|bit Arbitrary Waveform Generator (AWG) is presented and optimized. It can compress waveform memory for a sinusoid to 16×13bits with a Spurious Free Dynamic Range (SFDR) 90.7dBc...A memory compress algorithm for 12\|bit Arbitrary Waveform Generator (AWG) is presented and optimized. It can compress waveform memory for a sinusoid to 16×13bits with a Spurious Free Dynamic Range (SFDR) 90.7dBc (1/1890 of uncompressed memory at the same SFDR) and to 8×12bits with a SFDR 79dBc. Its hardware cost is six adders and two multipliers. Exploiting this memory compress technique makes it possible to build a high performance AWG on a chip.展开更多
An algorithm for partitioning arbitrary simple polygons into a number of convex parts was presented. The concave vertices were determined first, and then they were moved by using the method connecting the concave vert...An algorithm for partitioning arbitrary simple polygons into a number of convex parts was presented. The concave vertices were determined first, and then they were moved by using the method connecting the concave vertices with the vertices of falling into its region B,so that the primary polygon could be partitioned into two subpolygons. Finally, this method was applied recursively to the subpolygons until all the concave vertices were removed. This algorithm partitions the polygon into O(l) convex parts, its time complexity is max(O(n),O(l 2)) multiplications, where n is the number of vertices of the polygon and l is the number of the concave vertices.展开更多
To solve the hardware deployment problem caused by the vast demanding computational complexity of convolutional layers and limited hardware resources for the hardware network inference,a look-up table(LUT)-based convo...To solve the hardware deployment problem caused by the vast demanding computational complexity of convolutional layers and limited hardware resources for the hardware network inference,a look-up table(LUT)-based convolution architecture built on a field-programmable gate array using integer multipliers and addition trees is used.With the help of the Winograd algorithm,the optimization of convolution and multiplication is realized to reduce the computational complexity.The LUT-based operator is further optimized to construct a processing unit(PE).Simultaneously optimized storage streams improve memory access efficiency and solve bandwidth constraints.The data toggle rate is reduced to optimize power consumption.The experimental results show that the use of the Winograd algorithm to build basic processing units can significantly reduce the number of multipliers and achieve hardware deployment acceleration,while the time-division multiplexing of processing units improves resource utilization.Under this experimental condition,compared with the traditional convolution method,the architecture optimizes computing resources by 2.25 times and improves the peak throughput by 19.3 times.The LUT-based Winograd accelerator can effectively solve the deployment problem caused by limited hardware resources.展开更多
波达方向(direction of arrival,DOA)估计是阵列信号处理中的一个关键领域,而波达角(angle of arrival,AOA)则是其核心参数。将AOA估计作为主要研究目标,提高其在任意结构阵列中的准确性。通过分析现有文献中关于任意结构阵列的系统模型...波达方向(direction of arrival,DOA)估计是阵列信号处理中的一个关键领域,而波达角(angle of arrival,AOA)则是其核心参数。将AOA估计作为主要研究目标,提高其在任意结构阵列中的准确性。通过分析现有文献中关于任意结构阵列的系统模型,基于E-子空间理论,结合最大似然原理与Lloyd-like迭代算法,提出了一种新的子空间逼近迭代算法。此外,为了降低计算复杂度,结合MUSIC算法,提出了基于MUSIC峰值范围的迭代改进算法。最后,通过仿真验证了以上方法在毫米波信道估计中的有效性,相较于传统的MUSIC算法性能显著提升。展开更多
基金supported by Sichuan Science and Technology Program[2023YFSY0026,2023YFH0004].
文摘Recent Super-Resolution(SR)algorithms often suffer from excessive model complexity,high computational costs,and limited flexibility across varying image scales.To address these challenges,we propose DDNet,a dynamic and lightweight SR framework designed for arbitrary scaling factors.DDNet integrates a residual learning structure with an Adaptively fusion Feature Block(AFB)and a scale-aware upsampling module,effectively reducing parameter overhead while preserving reconstruction quality.Additionally,we introduce DDNetGAN,an enhanced variant that leverages a relativistic Generative Adversarial Network(GAN)to further improve texture realism.To validate the proposed models,we conduct extensive training using the DIV2K and Flickr2K datasets and evaluate performance across standard benchmarks including Set5,Set14,Urban100,Manga109,and BSD100.Our experiments cover both symmetric and asymmetric upscaling factors and incorporate ablation studies to assess key components.Results show that DDNet and DDNetGAN achieve competitive performance compared with mainstream SR algorithms,demonstrating a strong balance between accuracy,efficiency,and flexibility.These findings highlight the potential of our approach for practical real-world super-resolution applications.
基金financial support from the special fund of China’s central government for the development of local colleges and universities―the project of national first-level discipline in Oil and Gas Engineering, the National Science Fund for Distinguished Young Scholars of China (Grant No. 51125019)the National Program on Key fundamental Research Project (973 Program, Grant No. 2011CB201005)
文摘Compared with vertical and horizontal wells, the solution and computation of transient pressure responses of slanted wells are more complex. Vertical and horizontal wells are both simplified cases of slanted wells at particular inclination, so the model for slanted wells is more general and more complex than other models for vertical and horizontal wells. Many authors have studied unsteady-state flow of fluids in slanted wells and various solutions have been proposed. However, until now, few of the published results pertain to the computational efficiency. Whether in the time domain or in the Laplace domain, the computation of integration of complex functions is necessary in obtaining pressure responses of slanted wells, while the computation of the integration is complex and time-consuming. To obtain a perfect type curve the computation time is unacceptable even with an aid of high-speed computers. The purpose of this paper is to present an efficient algorithm to compute transient pressure distributions caused by slanted wells in reservoirs. Based on rigorous derivation, the transient pressure solution for slanted wells of any inclination angle is presented. Assuming an infinite-conductivity wellbore, the location of the equivalent-pressure point is determined. More importantly, according to the characteristics of the integrand in a transient pressure solution for slanted wells, the whole integral interval is partitioned into several small integral intervals, and then the method of variable substitution and the variable step-size piecewise numerical integration are employed. The amount of computation is significantly reduced and the computational efficiency is greatly improved. The algorithm proposed in this paper thoroughly solved the difficulty in the efficient and high-speed computation of transient pressure distribution of slanted wells with any inclination angle.
基金Supported by National Science Fund for Distinguished YoungScholars(60925011)
文摘This paper investigates the synchronization problem of clock oscillators in nonlinear dynamical network with arbitrary time-delays.First,a dynamic synchronization algorithm based on consensus control strategy,named fast averaging syn-chronization algorithm(FASA),is presented to find a solution to the synchronization problem.This algorithm can compensate the clock skew and offset differences between clock nodes,achieving the synchronization of clock nodes in a shorter time as compared to previous synchronization methods.Second,because of the dynamical performance of FASA,it is characterized from the perspective of compartmental dynamical system with arbitrary time-delays.In this case,the algorithm guarantees the states of all clock nodes in dynamical network converge to Lyapunov stable equilibria.Finally,numerical simulations and experimental results demonstrate the correctness and effciency of the FASA,which means that the clock nodes can reach global consensus,and the synchronization error can reach nanosecond order of magnitude.
文摘A memory compress algorithm for 12\|bit Arbitrary Waveform Generator (AWG) is presented and optimized. It can compress waveform memory for a sinusoid to 16×13bits with a Spurious Free Dynamic Range (SFDR) 90.7dBc (1/1890 of uncompressed memory at the same SFDR) and to 8×12bits with a SFDR 79dBc. Its hardware cost is six adders and two multipliers. Exploiting this memory compress technique makes it possible to build a high performance AWG on a chip.
文摘An algorithm for partitioning arbitrary simple polygons into a number of convex parts was presented. The concave vertices were determined first, and then they were moved by using the method connecting the concave vertices with the vertices of falling into its region B,so that the primary polygon could be partitioned into two subpolygons. Finally, this method was applied recursively to the subpolygons until all the concave vertices were removed. This algorithm partitions the polygon into O(l) convex parts, its time complexity is max(O(n),O(l 2)) multiplications, where n is the number of vertices of the polygon and l is the number of the concave vertices.
基金The Academic Colleges and Universities Innovation Program 2.0(No.BP0719013)。
文摘To solve the hardware deployment problem caused by the vast demanding computational complexity of convolutional layers and limited hardware resources for the hardware network inference,a look-up table(LUT)-based convolution architecture built on a field-programmable gate array using integer multipliers and addition trees is used.With the help of the Winograd algorithm,the optimization of convolution and multiplication is realized to reduce the computational complexity.The LUT-based operator is further optimized to construct a processing unit(PE).Simultaneously optimized storage streams improve memory access efficiency and solve bandwidth constraints.The data toggle rate is reduced to optimize power consumption.The experimental results show that the use of the Winograd algorithm to build basic processing units can significantly reduce the number of multipliers and achieve hardware deployment acceleration,while the time-division multiplexing of processing units improves resource utilization.Under this experimental condition,compared with the traditional convolution method,the architecture optimizes computing resources by 2.25 times and improves the peak throughput by 19.3 times.The LUT-based Winograd accelerator can effectively solve the deployment problem caused by limited hardware resources.
文摘波达方向(direction of arrival,DOA)估计是阵列信号处理中的一个关键领域,而波达角(angle of arrival,AOA)则是其核心参数。将AOA估计作为主要研究目标,提高其在任意结构阵列中的准确性。通过分析现有文献中关于任意结构阵列的系统模型,基于E-子空间理论,结合最大似然原理与Lloyd-like迭代算法,提出了一种新的子空间逼近迭代算法。此外,为了降低计算复杂度,结合MUSIC算法,提出了基于MUSIC峰值范围的迭代改进算法。最后,通过仿真验证了以上方法在毫米波信道估计中的有效性,相较于传统的MUSIC算法性能显著提升。
