Realizing balance control of single inverted pendulum without mathematic model by reinforcement learning has gained great success. However, further applying it to the multi-level inverted pendulum faces problems such ...
详细信息
ISBN:
(纸本)7121002159
Realizing balance control of single inverted pendulum without mathematic model by reinforcement learning has gained great success. However, further applying it to the multi-level inverted pendulum faces problems such as curse of dimensionality and difficulty in convergence. In this paper. we e propose a hierarchical reinforcement learning algorithm to control parallel double inverted pendulum. Firstly. control of single inverted pendulum is learned by Q-learning algorithm. then the learned control states of single inverted pendulum are used to direct control process of parallel double inverted pendulum so that the double one is controlled.
this Paper presents a new chip-structure to implement FFT the key features of this structure are parallel Computation and Hierarchically Memory Partition. this new structure is aimed at both increasing the data bandwi...
详细信息
ISBN:
(纸本)7121002159
this Paper presents a new chip-structure to implement FFT the key features of this structure are parallel Computation and Hierarchically Memory Partition. this new structure is aimed at both increasing the data bandwidth and decreasing die power dissipation, and it has been implemented on a 0.15-um manufacture process FPGA (Field Programmable Gate Array) chip. At a core supply voltage of 1.5V and a working clock of 50MHz, this chip can complete 1024-point 16-bit complex FFT in a little more than 30us while the chip core consumes about 500mW power.
the ability to extract a primary speech signal from an environment with multiple speakers is an important issue [1,2] in speech processing. this paper presents a method for incorporating multiple parallel beamformers ...
详细信息
ISBN:
(纸本)0780384075
the ability to extract a primary speech signal from an environment with multiple speakers is an important issue [1,2] in speech processing. this paper presents a method for incorporating multiple parallel beamformers with a Wiener filter. By iteratively improving the spectral magnitude estimates of each speech source, substantial improvement in overall signal separation can be obtained. the performance of the algorithm is illustrated using a simulated multiple speaker environment with resulting SNR and sSNR plots.
Implementing image-processing systems can require significant effort and resources due to information volume and algorithm complexity. Model Integrated Computing (MIC) based image processing systems show promise in su...
详细信息
ISBN:
(纸本)0769521258
Implementing image-processing systems can require significant effort and resources due to information volume and algorithm complexity. Model Integrated Computing (MIC) based image processing systems show promise in supporting solutions of these complex problems. While MIC has contributed to the advancement of performing complex image processing tasks on parallel-embedded systems, it has not addressed a challenging class of algorithmsthat adapt the image-processing algorithm based on the information or state of the image processing system. this proposed effort addresses creating an adaptive image-processing environment based on MIC that allows solutions of complex image processing problems to be built and executed rapidly. this effort will involve creating a new modeling representation for image processing adaptation mechanisms. the proposed MIC-based adaptive image-processing environment will generate a solution given the modeling constraints and execute it on a number of hardware architectures.
this paper presents a theoretical analysis in the sense of the Neyman-Pearson (N-P) test about die relationship between fusion rule and local decision rules in the parallel distributed detection fusion system with mul...
详细信息
ISBN:
(纸本)7121002159
this paper presents a theoretical analysis in the sense of the Neyman-Pearson (N-P) test about die relationship between fusion rule and local decision rules in the parallel distributed detection fusion system with multiple sensors. It combines wavelet filter based on soft-threshold withthe parallel distributed detection fusion system with multiple sensors ideally. and derives the determination of decision rules of two sub-optimal systems and globally optimal system completely. the detection performances of three systems above are computed numerically for the problem of detecting a known signal embeded in Rayleigh noise. the results obtained indicate that the presented method can enhance die radar detection performance remarkably.
Code partitioning is the problem of dividing sections of code among a set of processors for execution in parallel taking into account the communication overhead between the processors. Code partitioning of large amoun...
详细信息
ISBN:
(纸本)0769522297
Code partitioning is the problem of dividing sections of code among a set of processors for execution in parallel taking into account the communication overhead between the processors. Code partitioning of large amounts of code onto numerous processors requires variations to the classical partitioning algorithms, in part due to the memory and time requirements to partition a large set of data, but also due to the nature of the target machine and multiple constraints imposed by its architectural features. In this paper we present our experience in the design of enhancements to the classical multi-level k-way partitioning algorithm to deal with large graphs of over 1 million nodes, 5 constraints, and nodes of irregular size. Our algorithm was implemented to produce code for a massively parallel machine of up to 40,000 processors, and forms part of a hardware description language compiler. the algorithm and the compiler were tested on RTL designs for a next generation SPARC(R) processor. We present performance results and comparisons for partitioning multi-processor hardware designs.
In this paper, we present "rules of thumb" for the efficient and straight-forward parallelization of cellular neural networks (CNNs) processing image data on cluster architectures. the rules result from the ...
详细信息
In this paper, we present "rules of thumb" for the efficient and straight-forward parallelization of cellular neural networks (CNNs) processing image data on cluster architectures. the rules result from the application and optimization of the simple but effective structural data parallel approach, which is based on the SPMD model. Digital gray-scale images were used to evaluate the optimized parallel cellular neural network program. the process of parallelizing the algorithm employs HPF to generate an MPI-based program.
It is well known that 4 or 8-neighborhood property has an important role in many algorithms, for example, image processing, solving partial differential equation (PDE) etc. In this paper, we establish these properties...
详细信息
It is well known that 4 or 8-neighborhood property has an important role in many algorithms, for example, image processing, solving partial differential equation (PDE) etc. In this paper, we establish these properties on an OTIS-Mesh, an optoelectronic parallel computer. We show that these properties can be established in constant time withthe help of a new indexing scheme called processor data index (PDI), proposed in this paper.
this paper presents a new chip-structure to implement FFT. the key features of this structure are parallel Computation and Hierarchically Memory Partition. this new structure is aimed at botta increasing the data band...
详细信息
ISBN:
(纸本)0780384075
this paper presents a new chip-structure to implement FFT. the key features of this structure are parallel Computation and Hierarchically Memory Partition. this new structure is aimed at botta increasing the data bandwidth and decreasing the power dissipation, and it has been implemented on a 0.15-um manufacture process FPGA (Field Programmable Gate Array) chip. At a core supply voltage of 1.5V and a working clock of 50MHz, this chip can complete 1024-point 16-bit complex FFT in a little more than 30us while the chip core consumes about 500mW power.
暂无评论