This study presents an algorithm for multi-channel electroencephalographic (EEG) spike prescreening based on radial basis function (RBF) artificial neural network. Sixteen-channel EEG records from twenty-eight patient...
详细信息
This study presents an algorithm for multi-channel electroencephalographic (EEG) spike prescreening based on radial basis function (RBF) artificial neural network. Sixteen-channel EEG records from twenty-eight patients with epilepsy were examined online. The single-channel recognition network consisted of input-layer neurons receiving three EEG waveform parameters (peak angle, amplitude, and velocity), ten hidden-layer neurons, and one output neuron. An optimal network model was chosen according to receiver operating characteristics analysis. Identification of multi-channel geometric correlation was performed with an incidence matrix. Sensitivity and selectivity cross-over value was found to be 80% for the single-channel RBF network. Compared with error back-propagation, a substantial improvement in training efficiency by at least two to three orders of magnitude was obtained for the RBF network. Validation with visual analysis showed 87.3% sensitivity using the proposed online multi-channel classification algorithm. The computation time required for spike detection was significantly less than that needed for online display of the EEG record. The proposed algorithm was able to achieve classification accuracy similar to previous artificial neural network schemes based on error back-propagation, while the training needed for RBF network was substantially less. Therefore, the RBF neural network is potentially an effective tool in real-time prescreening for epileptic spikes.
作者：
Prieur, ChristopheTeel, Andrew R.LAAS-CNRS
7 Avenue du Colonel Roche 31077 Toulouse Cedex 4 France Center for Control
Dynamical Systems and Computation Department of Electrical and Computer Engineering University of California Santa Barbara CA 93106-9560 United States
We consider control systems for which we know two stabilizing output feedback controllers. One is globally asymptotically stabilizing, while the other one is only locally asymptotically stabilizing. We look for a comp...
详细信息
In this paper, a design of the 3-phase, 6-stator-pole and 4-rotor-pole, 15 kW switched reluctance motor for the propulsion system of the electric vehicle is presented. Its performance such as the torque per volume and...
详细信息
ISBN:
(纸本)8986510081
In this paper, a design of the 3-phase, 6-stator-pole and 4-rotor-pole, 15 kW switched reluctance motor for the propulsion system of the electric vehicle is presented. Its performance such as the torque per volume and its efficiency are compared to the machine which is designed based on the conventional design guideline.
In this paper a method to estimate the initial rotor position of the switched reluctance motor (SRM) at the stand still condition is proposed. The proposed method can be implemented by the simple microcontroller based...
详细信息
ISBN:
(纸本)8986510081
In this paper a method to estimate the initial rotor position of the switched reluctance motor (SRM) at the stand still condition is proposed. The proposed method can be implemented by the simple microcontroller based system, and it is not necessary to know the magnetization curves. The impulse gate drive signal is applied and the winding current is detected and its approximate value is used for calculating the rotor position. The compensation method to reduce the estimation error thorough one cycle is also mentioned. The proposed method is implemented with the 6-stator-pole and 4-rotor-pole, 1.5 kW 3-phase switched reluctance motor.
We present an algorithm based on hierarchical decomposition that finds close-to-optimal search paths for a cooperative team of agents searching for one or more targets on a graph. The method partitions the graph to cr...
详细信息
We present an algorithm based on hierarchical decomposition that finds close-to-optimal search paths for a cooperative team of agents searching for one or more targets on a graph. The method partitions the graph to create a high-level problem and several lower-level problems. Since the computations on each level are identical, the lower-level problems can be further decomposed. In this way, the problem becomes fractal in nature. We use best-case and worst-case instances of the decomposed problem to establish upper and lower bounds on the optimal search reward, and the bounds are determined with much less computation than what is required to solve the full problem. We show that as the number of decomposition levels increases, the computational complexity approaches O(n) at the expense of looser bounds on the optimal reward. A large-scale test case shows that this method is computationally fast, produces good results, and achieves true cooperation between agents.
作者：
Michael FryeChunjiang QianControl
Computation and Cybernetics Laboratory Department of Electrical and Computer Engineering University of Technology San Antonio TX USA
This paper examines the problem of controlling a system of airships that are operating as a temporary mobile cell phone network. The primary requirements are maintaining a formation that is subjected to environmental ...
详细信息
This paper examines the problem of controlling a system of airships that are operating as a temporary mobile cell phone network. The primary requirements are maintaining a formation that is subjected to environmental gust disturbances while holding either a fixed position or attempting to hold a reference velocity. The purpose of having both position and velocity hold modes is to have the performance of an aerostat while maintaining the ability to relocate as required by cell phone load usage. This paper examines the use of a predictive real-time control technique which solves an optimal control problem that takes into account minimum control energy, desired velocity, formation constraints, and control constraints. The control problem is reformulated as a linear programming problem for rapid convergence to a solution. This paper will illustrate the optimal control technique using a simulation on two airships.
作者：
Cai ChaohongGoebel RafalG. Sanfelice RicardoR. Teel AndrewDepartment of Electrical and Computer Engineering
University of California Santa Barbara CA USA Center for Control
Dynamical Systems and Computation Department of Electrical and Computer Engineering University of California Seattle WA USA Seattle
WA USA Center for Control
Dynamical Systems and Computation Department of Electrical and Computer Engineering University of California Santa Barbara CA 93106-9560 USA. E-mail: rsanfelice@ece.ucsb.edu Center for Control
Dynamical Systems and Computation Department of Electrical and Computer Engineering University of California Santa Barbara CA 93106-9560 USA. E-mail: teel@ece.ucsb.edu
Modeling issues for hybrid dynamical systems are discussed and fundamental stability analysis tools are summarized. These tools are useful for the development of hybrid control algorithms.
Modeling issues for hybrid dynamical systems are discussed and fundamental stability analysis tools are summarized. These tools are useful for the development of hybrid control algorithms.
The unique characteristic of a repetitive process is a series of sweeps or passes through a set of dynamics defined over a finite duration known as the pass length. At the end of each pass, the process is reset and th...
详细信息
This paper focuses on the asymptotic stability properties of omega limit sets for complex hybrid dynamical systems,which are commonly found in systems and *** spells out specific stability results that follow when a h...
详细信息
This paper focuses on the asymptotic stability properties of omega limit sets for complex hybrid dynamical systems,which are commonly found in systems and *** spells out specific stability results that follow when a hybrid dynamical system has certain structure,e.g.,when it admits a decomposition resembling a cascade of hybrid dynamical systems.
暂无评论