Papers by Keyword: Radial Basis Function Neural Network (RBFNN)

Abstract: In this work, we estimate Yunnan housing price from 1999 to 2009. Firstly, we analyze the correlation coefficients between housing price and characteristic variables, identify the characteristic variables. Then, we build the forecasting model using four techniques, support vector regression (SVR), radial basis function neural network (RBFNN), partial least square (PLS) and multiple regression analysis (MRA), based on whole variables and characteristic variables. The results show that PLS technique is the best one for housing price forecasting. Its mean absolute percentage error (MAPE) is only 2.45%. SVR and RBFNN are better techniques to obtain a satisfactory forecasting result with almost 5% MAPE. Furthermore, the performance of MRA and SVR can be obviously improved through variables selection.

Abstract: An accurate forecasting method for wind power generation of the wind energy conversion system (WECS) can help the power system’s operator to reduce the risk of unreliability of electricity supply. This paper proposed a radial basis function (RBF) neural network method to forecast the wind power generation of WECS. To demonstrate the effectiveness of the proposed method, the method is tested on the practical information of wind power generation of a WECS. The good agreements between the realistic values and forecasting values are obtained; the numerical results show that the proposed forecasting method is accurate and reliable.

Abstract: A radial basis function neural networks (RBFNs) mobile robot control system is automatically developed with the image processing and learned by the bacterial foraging particle swarm optimization (BFPSO) algorithm in this paper. The image-based architecture of robot model is self-generated to travel the routing path in the dynamical and complicated environments. The visible omni-directional image sensors capture the surrounding environment to represent the behavior model of the mobile robot system. Three parameterize RBFNs model with the centers and spreads of each radial basis function, and the connection weights to solve the mobile robot path traveling and routing problems. Several free parameters of radial basis functions can be automatically tuned by the direct of the specified fitness function. In additional, the proper number of radial basis functions of the constructed RBFNs can be chosen by the defined fitness function which takes this factor into account. The desired multiple objectives of the RBFNs control system are proposed to simultaneously approach the shorter path and avoid the unexpected obstacles. Evaluations of PSO and BFPSO show that the developed RBFNs robot systems skip the obstacles and efficiently achieve the desired targets as soon as possible.

Abstract: Dissolved Gas Analysis (DGA) is a popular method to detect and diagnose different types of faults occurring in power transformers. Improved three-ratio is an effective method for transformer fault diagnosis used in recent years. This paper applies appropriate Artificial Neural Networks (ANN) to resolve the online fault diagnosis problems for oil-filled power transformer based on improved three-ratio. Because of the characteristic of improved three-ratio boundary is too absolute, a method using fuzzy math theory to deal with the data of the neural network input is also proposed. A major kind of neural network, i.e. radial basis function neural network (RBFNN), is used to model the fault diagnosis structure. In addition, to improve the convergence speed, an improved gradient descent algorithm is used in training RBFNN. Through on-line monitoring the concentrations of the dissolved gases, the proposed diagnostic system can offer a way to interpret the incipient faults. The simulation diagnosis demonstrates the effectiveness and veracity of the proposed method.

Abstract: A new algorithm for training radial basis function neural network (RBFNN) is presented in this paper. This algorithm is based on the dynamic fuzzy clustering method (DFCM). The algorithm has a number of advantages compared to the traditional method based on k-means. For example, it does not need to know the number of the hidden nodes and to predicts more accurately. Due to these advantages, this method proves to be suitable for developing models for complex nonlinear systems.

Abstract: A nondestructive measurement approach is presented in this paper, which is capable of determining sugar content in cantaloupe from the dielectric property. The approach is based on measured equivalent capacitance and equivalent resistance of the cantaloupe, and on data analysis using quantum-behaved particle swarm optimization (QPSO) and Grey radial basis function (RBF) neural network. First, accumulated generating operation (AGO) in Grey forecasting is used to convert the initial observed data to obtain the accumulated data with strong regularity, which are employed to model and train the radial basis function neural network. Second, it adopted quantum-behaved particle swarm optimization algorithm to train the centers and widths of radial basis function. This model not only prevented the problem that the parameters of neural network are hard to be tuned, but also improved the network precision of prediction. Experimental results revealed that the predictive model as proposed has good predictive effect for the measurement of sugar in cantaloupes.

Abstract: In order to control the wire diameter stability for pitch carbon fiber melt-spinning effectively, this can affect the performance of carbon fiber. This paper presents an asphalt carbon fiber melt-spinning wire diameter stabilization method based on radial basis function neural network. Firstly, the relation model that pitch carbon fiber melt-spinning wire diameter, spinning temperature, spinning pressure and spinning roller speed was established through measured data based on radial basis function neural network. Then control the spinning temperature, pressure and spinning rollers speed coordination changes to ensure the stability of spinning wire diameter in spinning process. Finally, we apply this method to our laboratory measured data and compared with existing experience formula. The result shows that the method is feasible and effective

Abstract: Take serial robot with six DOF for example. On the basis of analyzing the characteristics of RBF neural network, inverse kinematics calculation of arc welding robot was achieved by RBF of six-input and single output. The forward and inverse kinematics could be seen as a nonlinear mapping between the joint space and the operation space of the robot. Take the algorithm based on RBF. Acquire RBF centers by the nearest neighbor clustering algorithm. The inverse kinematics of robot was solved. Through learning the training samples of the positive solutions to determine weight coefficient of neural network, the robots pose could be accurately solved. The example shows that the algorithm has the characteristics of simple calculation and effective solution, etc. The cumbersome derivation of traditional methods is avoided. It can be seen as kinematics trajectory tracking controller of serial mechanism system.

Abstract: The paper presents an approach to model the electro-hydraulic system of a certain explosive mine sweeping device using the Radial Basis Function (RBF) neural network. In order to obtain accurate and simple RBF neural network, a revised clustering method is used to train the hidden node centers of the neural network, in which the subtractive clustering(SC) algorithm was used to determine the initial centers and the fuzzy C – Means(FCM) clustering algorithm to further determined the centers data set. The spread factors and the weights of the neural network are calculated by the modified recursive least squares (MRLS) algorithm for relieving computational burden. The proposed algorithm is verified by its application to the modeling of an electro-hydraulic system, simulation and experiment results clearly indicate the obtained RBF network can model the electro-hydraulic system satisfactorily and comparison results also show that the proposed algorithm performs better than the other methods.

Abstract: In order to solve the problem that how to evaluate the complex system support concept, an evaluation method based on Radial Basis Function (RBF) neural network model was presented. Through researching the support system overall design characteristics and elements of support, on this basis, evaluation parameters of support concept were abstracted. Support concept evaluation model based on RBF was established and a mature and stable RBF neural network was trained to calculate the comprehensive evaluation value for support concept. Finally, the further demonstration and verification of the method are given through specific case application and compared with the result for evaluation results of data envelopment analysis (DEA) model.