This paper presents a new approach to dynamic software tuning called the adaptive algorithm selection method (AASM). The AASM is built into the calling sequence of a library. When the library is called, the AASM is ac...
详细信息
This paper presents a new approach to dynamic software tuning called the adaptive algorithm selection method (AASM). The AASM is built into the calling sequence of a library. When the library is called, the AASM is activated. The AASM selects and executes the optimum algorithm from registered algorithms in a library, based on data and machine type. As a result, the software is automatically tuned and the execution time is shortened. The relation between the data and the best algorithm for a given machine is learned by a neural network from the results of performance tests of the registered algorithms. We experimented on a multi-strings search problem with the AASM on the following machines: the CRAY X-MP/216, FACOM M 1800/30, and SUN Sparc Station 2. From these experiments we demonstrated that the AASM is able to minimize the execution time.< >
Wind energy systems are being closely studied because of its benefits as an environmentally friendly and renewable source of energy. Because of its unpredictable nature, power management concepts are essential to extr...
详细信息
Wind energy systems are being closely studied because of its benefits as an environmentally friendly and renewable source of energy. Because of its unpredictable nature, power management concepts are essential to extract as much power as possible from the wind when it becomes available. In this paper an algorithm has been developed to keep the system at its highest possible efficiency at all times. The proposed algorithm uses a modified version of hill climb search (HCS) and intelligent memory to implement its power management scheme. Because it does not require that the turbinepsilas characteristics be pre-programmed to obtain the optimal operating points for maximum power transfer, it can be applied to a wide range of wind turbines. The algorithm determines the turbinepsilas internal characteristics through operation. Once the algorithm obtains knowledge of the turbinepsilas characteristics, it can infer the optimum rotor speeds for wind speeds that have not occurred before. The main focus of this paper is the algorithm structure and its effectiveness under fluctuating atmospheric conditions. PSIM simulation studies have been done to confirm the effectiveness of the proposed algorithm.
Pulse-to-pulse coherent Doppler sonar has been used to provide accurate velocity information, but the occurrence of "range-velocity" ambiguity seriously limits more general applications. Although multi-frequ...
详细信息
ISBN:
(纸本)9781479987375
Pulse-to-pulse coherent Doppler sonar has been used to provide accurate velocity information, but the occurrence of "range-velocity" ambiguity seriously limits more general applications. Although multi-frequency and dual-time-interval methods have been used to enlarge the range of ambiguity velocity, it remains limited. Therefore, a combined method of conventional and coherent Doppler sonar using a fixed range of ambiguity velocity has been proposed to provide accurate and precise velocity without velocity ambiguity. The combined method using a fixed range of ambiguity velocity worked well at high SNR, but at low SNR it was not able to provide the same accuracy as seen at high SNR. In order to provide accurate and precise velocities at a wide range of SNR, an adaptive algorithm of ambiguity velocity is proposed. Based on simulation results, the combined method using the adaptive algorithm can provide accurate and precise velocities at a wide range of SNR.
This paper presents an adaptive algorithm to compensate web tension disturbances caused by the eccentricity and non-circularity of the reel and rolls in web winding systems. The disturbances generated by the eccentric...
详细信息
This paper presents an adaptive algorithm to compensate web tension disturbances caused by the eccentricity and non-circularity of the reel and rolls in web winding systems. The disturbances generated by the eccentricity and non-circularity of the roll have a quasi-periodic form with a frequency that varies with the rotation speed of the roll. The algorithm is based on an estimation of the phase component and the quadrature phase component of the disturbances. Approximate harmonic analysis and parameter design are done. The ability of the algorithm to reject eccentricity and non-circularity tension disturbances with slowly varying frequencies is shown through simulations on the physical model as well as with experimental results.
This paper presents adaptive algorithm for updating the weights of a WFQ scheduler. The algorithm provides maximum revenue for the network operator, while the QoS requirements for the connections in different service ...
详细信息
This paper presents adaptive algorithm for updating the weights of a WFQ scheduler. The algorithm provides maximum revenue for the network operator, while the QoS requirements for the connections in different service classes are satisfied. The operation of the algorithm is verified in ns-2 simulator environment. The simulations show that the algorithm works well with TCP traffic and provides appropriate service for all service classes.
One of the essential problems of sensor acquisition systems is the correction of measurement affected by nonsystematic error. This paper continues the problem of error correction from, and is focused on a digital meth...
详细信息
One of the essential problems of sensor acquisition systems is the correction of measurement affected by nonsystematic error. This paper continues the problem of error correction from, and is focused on a digital method used in signal processing for correction of temperature measurement from inside of dam. The temperature measurements are done in a long time by same year. The measurements from inside of dam depend by ambient temperature of environment, so the temperature evolution respects the ambient temperature variation in good condition The purpose of this article is to interpret the temperature evolution as a signal and for correction is used an adaptive filtering method. The systematic error is a high frequency noise that affects the sensor measurement. The adaptive filter process two signals one is the utile signal (the sensor measurement) and the second is the reference signal (the environment temperature). This method consists in obtaining an estimate value for error and eliminates errors from sensor measurements. This method use the discreet time signal, the advantage is that this adaptive algorithm can be implemented with digital programmable circuit such as microcontroller, DSP. This algorithm can be implemented in data acquisition system, so the correction data will be done in real time.
The vestibular system is important for the maintenance of static and dynamic equilibrium of the human body. Dysfunctions within this system may lead to abnormal reflex eye movements known as nystagmus. It is possible ...
详细信息
The vestibular system is important for the maintenance of static and dynamic equilibrium of the human body. Dysfunctions within this system may lead to abnormal reflex eye movements known as nystagmus. It is possible to diagnose several pathologies of the human vestibular system through the analysis of this signal. In a previous work a computational algorithm that automatically evaluates the nystagmus' slow component velocity was developed. This algorithm gave good results when compared to the clinical analysis, but it did not foresee the possibility of expressive changes in the signal to noise ratio. Here, the author presents a new algorithm that monitors the level of noise associated to the signal and modifies its parameters in such a way as to improve the results. Comparisons between these two algorithms were done and the results indicate a clear superiority of this new technique over the previous one.
This paper presents new radar algorithm for detecting moving targets and measuring target velocities. The algorithm is based on calculating spectrum estimates by using the signals reflected from the adjacent range bin...
详细信息
This paper presents new radar algorithm for detecting moving targets and measuring target velocities. The algorithm is based on calculating spectrum estimates by using the signals reflected from the adjacent range bins. Statistical hypothesis checking is fulfilled on the difference between the adjacent spectra. The decision-making on distinction between different spectral components is used for estimating velocity of the target. Efficiency of the algorithm is investigated by using statistical modeling. The algorithm is checked by processing a real radar data containing a signal on the background of reflections from rain.
The sparse reconfigurable adaptive filter (SRAF) consists of an input delay bank, a photonic switch with reconfigurable connections and adaptive weights, and an output delay bank. This system architecture can be used ...
详细信息
The sparse reconfigurable adaptive filter (SRAF) consists of an input delay bank, a photonic switch with reconfigurable connections and adaptive weights, and an output delay bank. This system architecture can be used to implement a sparse tapped delay line (TDL) with a large number of adaptive weights and a wide range of time delays. In this paper, we investigate some delay properties of the SRAF and consider an adaptive algorithm based on a cross-correlation function of the pre-whitened input and a desired response signal. Computer simulations are presented to illustrate the behavior of the SRAF for a system identification application
A novel adaptive algorithm for MMSE multiuser detection in downlink CDMA is proposed in this paper. This algorithm uses the information in the pilot channel to compute the desired MMSE solution directly. Compared with...
详细信息
A novel adaptive algorithm for MMSE multiuser detection in downlink CDMA is proposed in this paper. This algorithm uses the information in the pilot channel to compute the desired MMSE solution directly. Compared with conventional adaptive algorithms and previously proposed blind algorithms, it does not require training sequences or channel estimation. Analysis shows that the MMSE detector obtained by the algorithm converges to the Wiener solution globally, and the computational complexity is O(N/sup 2/). Simulation examples illustrate that the algorithm can rapidly reach the steady state. The steady state performance can be enhanced by increasing the transmitted power in the pilot channel, but is worse than that of a conventional recursive least-square (RLS) algorithm in decision-directed mode.
暂无评论