Typical speech enhancement algorithms that operate in the Fourier domain only modify the magnitude component of the noisy speech. It is commonly understood that the phase component is perceptually unimportant, and thu...
详细信息
ISBN:
(纸本)9781424481835
Typical speech enhancement algorithms that operate in the Fourier domain only modify the magnitude component of the noisy speech. It is commonly understood that the phase component is perceptually unimportant, and thus, it is passed directly to the output. Nevertheless, it has been reported in recent experiments that the Short-Time Fourier Transform (STFT) phase spectrum contributes significantly to speech intelligibility. Motivated by this, we investigated the role of phase spectrum in speech enhancement using Wiener filtering and Martin's minimum statistics. In this paper we report on results obtained using optimization algorithms, for phase correction of each processed frame, that intend to match the waveform of the zero-phase Wiener filtered speech to the conventional filter output obtained with noisy phase characteristic. No a priori information on the original phase is assumed. We show that better results are achieved using phase correction for different noise types. Different criteria are used for optimization with results similar to the case when the actual clean speech phase is at hand. Almost as good results are also obtained when minimizing the Wiener filter impulse response dispersion. The achieved improvement is assessed through different measurements such as signal to noise ratio (SNR), Segmental signal to noise ratio, and Perceptual Estimation of Speech Quality (PESQ).
It is known that there is only information sharing in most particle swarm optimization. But competition among particles which is a good feature for searching progress does not exist. For all these, based on the idea o...
详细信息
ISBN:
(纸本)9783642040191
It is known that there is only information sharing in most particle swarm optimization. But competition among particles which is a good feature for searching progress does not exist. For all these, based on the idea of multiagent with emotion, bring in competition controlled by emotion to enhance performance of PSO after describing similarity between particles swarm and multi-agent system. And from the point of emotional view, problem that agents whether should compete or not is stated qualitatively. Furthermore, velocity threshold is deduced. Utilizing these, the method proposed could improve both local and global searching ability of particle swarm optimization. In addition, simulation results show that the improvement is effective. algorithm with it has good efficiency.
This paper presents the results obtained testing a tool designed to characterize and simulate the behavior of a Photovoltaic (PV) panel under real and/or simulated working conditions. The presented tool permits the co...
详细信息
ISBN:
(纸本)9781424428328;9781424428335
This paper presents the results obtained testing a tool designed to characterize and simulate the behavior of a Photovoltaic (PV) panel under real and/or simulated working conditions. The presented tool permits the continuous monitoring of the I-V (Current-Voltage) and P-V (Power-Voltage) characteristics of the panel and the comparison between actual and expected performance;in this way it is possible to prevent any possible decrease of the output power and to replace the monitored module before it goes out of order or its efficiency falls under a given threshold. The well known two-diode model is used to estimate the parameters of the electrical equivalent circuit of the PV panel and to simulate both the I-V and P-V characteristic curves in any given environmental condition of irradiation and/or temperature. The model and the estimation algorithm are implemented as a MATLAB script while the user interface is designed with LabVTEW. The presented tool has been validated against an experimentally characterized PV panel. The environmental parameters of the model such as irradiance and temperature have been measured directly, whereas the others parameters have been evaluated using a best-fit algorithm on the measured data.
Dynamic resectorization is a promising concept to accommodate the increasing and fluctuating demands of flight operations in the National Airspace System. At the core of dynamic resectorization is finding an optimal s...
详细信息
Dynamic resectorization is a promising concept to accommodate the increasing and fluctuating demands of flight operations in the National Airspace System. At the core of dynamic resectorization is finding an optimal sectorization. Finding such an optimal sectorization is challenging, because it mixes the graph partition problem and non-deterministic polynomial-time-hard optimization problem. This paper revisits Voronoi diagrams and genetic algorithms, and proposes a strategy that combines these algorithms with the iterative-deepening algorithm. Voronoi diagrams accomplish the graph partition, which then needs to be optimized. By defining a multi-objective cost, the combination of the genetic algorithm and iterative deepening algorithm solves the optimization problem. Experimental results show that this method can accomplish sector design by setting an appropriate cost. Without a need of clustering, this method can capture the dominant flow, which is one of the major concerns in sector design. The design can have balanced aircraft count and low coordination. If the capacity is defined and incorporated into the cost, the sectorization will lead to a design with increased capacity. The whole process can be finished within a feasible time period without the need for parallel schemes.
So far most of the K-means algorithms use the number of the labeled data as the K value, but sometimes it doesn't work well. In this paper, we propose a semi-supervised K-means algorithm based on the global optimi...
详细信息
ISBN:
(纸本)9780769536996
So far most of the K-means algorithms use the number of the labeled data as the K value, but sometimes it doesn't work well. In this paper, we propose a semi-supervised K-means algorithm based on the global optimization. It can select an appropriate number of clusters as the K value directly and plan a great amount of supervision data by using only a small amount of the labeled data. Combining the distribution characteristics of data sets and monitoring information in each cluster after clustering, we use the voting rule to guide the cluster labeling in the data sets. The experiments indicated that the global optimization algorithm for semi-supervised K-means is quite helpful to improve the K-means algorithm, it can effectively rind the best data sets for K values and clustering center and enhancing the performance of clustering.
The weapon target assignment problem is about finding the optimal allocation of weapons to threats in a way that minimizes the expected damage inflicted on the defender assets. In 1986 Lloyd and Witsenhousen demonstra...
详细信息
The weapon target assignment problem is about finding the optimal allocation of weapons to threats in a way that minimizes the expected damage inflicted on the defender assets. In 1986 Lloyd and Witsenhousen demonstrated that weapon target assignment was NP-complete with no exact algorithm. Researchers in this area are trying to provide an exact solution to special cases of the problem or heuristics that attempt to supply an approximate solution using a variety of tools and techniques from nonlinear network flows to artificial neural networks and genetic algorithms. A new approach for tackling the weapon target assignment problem is proposed in this paper. Such an approach is a novel goal-based system. The proposed novel approach combines state of the art goal-based optimization approach and the Hungarian method to preserve good performance under different air defense mission configurations. An air defense mission design and analysis package is developed to provide realistic air defense missions data to the algorithm. The proposed algorithm has the best performance when compared with other weapon target assignment doctrines.
This paper demonstrates the significant role that invariant manifolds play in the dynamics of low-thrust trajectories moving through unstable regions in the three-body problem. It shows that an optimization algorithm ...
详细信息
This paper demonstrates the significant role that invariant manifolds play in the dynamics of low-thrust trajectories moving through unstable regions in the three-body problem. It shows that an optimization algorithm incorporating no knowledge of invariant manifolds converges on low-thrust trajectories that use the invariant manifolds of unstable resonant orbits to traverse resonances. It is determined that the algorithm could both change the energy through thrusting to a level where the invariant manifolds could more easily be used, as well as use thrusting to move the trajectory along the invariant manifolds. Knowledge of this relationship has the potential to be very useful in developing initial guesses and new control laws for these optimization algorithms. In particular, this approach can speed up the convergence of the optimization process, retain the essential geometric and topological characteristics of the initial design, and provide a more accurate estimate of the Delta V and fuel usage based on the initial trajectory.
Visualization is often invaluable to understand the behavior of optimization algorithms, identify their bottlenecks or pathological behaviors, and suggest remedial techniques. Yet developing visualizations is often a ...
详细信息
Visualization is often invaluable to understand the behavior of optimization algorithms, identify their bottlenecks or pathological behaviors, and suggest remedial techniques. Yet developing visualizations is often a tedious activity requiring significant time and expertise. This paper presents a framework for the visualization of constraint-based local search (CBLS) algorithms. Given a high-level model and a declarative visualization specification, the CBLS visualizer systematically produces animations to visualize constraints and objectives, violations, and conflicts, as well as the temporal behavior of these measures. The visualization specification is declarative and typically composed of a triple (what,where,how) indicating what to display, where, and with which graphical objects. The visualizer architecture is compositional and extensible. It provides building blocks which can be assembled freely by the user and focuses almost exclusively on static aspects, the dynamic aspects being automated by the use of invariants. The paper highlights various functionalities of the visualizer and describes a blueprint for its implementation.
The hydrological component of the soil and water assessment tool (SWAT) model is adapted for two Ethiopian catchments based on primary knowledge of the coherence spectrum between rainfall and stream flow data. Spectru...
详细信息
The hydrological component of the soil and water assessment tool (SWAT) model is adapted for two Ethiopian catchments based on primary knowledge of the coherence spectrum between rainfall and stream flow data. Spectrum analysis using the available nearby climatic data is trade to limit the temporal and spatial scales (inverse rate coefficients) subject to the calibration of compartmentalized runoff models. The exclusion of unwarranted time scales in the calibration implies that the model efficiency (r(2) values) decrease only moderately between calibration and validation, and the optimization is focused on warranted problems. On the basis of the available data for the two Ethiopian catchments, the implication is that only periods longer than about 50 days can be reliably evaluated in the model. The model structure of SWAT for the surface runoff and groundwater flow response is modified to make the time scales consistent with the results of the spectrum analysis. An optimization algorithm is developed to constrain and combine the model parameters with the spectrum analysis results. Copyright (C) 2009 John Wiley & Sons, Ltd.
暂无评论