RBF Network Structure Research Articles

RBFNN-based angles-only orbit determination method for non-cooperative space targets

Angles-only orbit determination during space situational awareness missions for non-cooperative targets such as space debris is subject to the well-known observability problem, that is, the orbital state is weakly observable or even unobservable because of missing range measurements. The key point to solve this problem is how to quickly capture the nonlinearity of orbit dynamics if only one space-based passive sensor is available to observe the target and translational maneuvers are not executed. To solve the short-arc angles-only problem, this study develops a simple and fast orbit determination method by exploiting the advantages of radial basis function neural network (RBFNN). Emphasis was placed on the design of the RBFNN, which was trained to construct a mapping network model from line-of-sight (LOS) measurements to the relative orbit. First, the perturbed relative orbit dynamics considering the Earth's non-spherical J2 term and air drag was established with the LOS measurement model. Second, a RBFNN-based angles-only orbit determination frame was developed, in which the generator of the training dataset, data preprocessing algorithm, and RBF network structure were designed. Finally, the proposed method was verified and tested using a set of numerical simulations of low-Earth orbit cases. The anti-noise and generalization performance of the designed network and the sensitivity of the solution accuracy to the measurement frequency and span were presented and discussed. Additionally, the advantages of the RBFNN-based method were confirmed by comparison with the results obtained using the BPNN (Back Propagation Neural Network)-based method. The simulation results show that the average orbit prediction error for LEO scenarios is approximately 0.08% using the index of initial relative distance, where only three sets of LOS angles are measured within a span of 24 s.

Application of RBF network structure and data mining in e-commerce network marketing

Application of RBF network structure and data mining in e-commerce network marketing

Research on Vehicle Network Security Situation Prediction Based on Improved CLPSO-RBF

In order to improve the accuracy and efficiency of the vehicle network security situation prediction, a prediction algorithm based on improved CLPSO-RBF is proposed. Firstly, for speeding up the optimization efficiency of CLPSO, a reasonable speed monitoring variable has been introduced. Secondly, we use the improved CLPSO algorithm to optimize the clustering radius of the RBF neural network, so that the optimal RBF network structure can be determined. Finally, we use the optimal RBF network to predict the security situation of the vehicle network. Simulation experiments have proved that the improved algorithm has higher accuracy and faster convergence rate in situation prediction, and has better prediction effects.

Prediction of concrete compressive strength based on principal component analysis and radial basis function neural network

Aiming at the concrete compressive strength data measured in the field laboratory, the data samples are standardized and the missing and abnormal values of the data samples are detected and processed. Then, the principal component factors of several influencing factors are extracted with the idea of “dimensionality reduction” by principal component analysis, and a new data sample set is established. Finally, the radial basis function neural network model is constructed to simulate the concrete compressive strength. The strength is predicted and the predicted value is obtained. The results show that three principal component influencing factors are extracted by principal component analysis, and the RBF network structure with 50 neuron nodes in the hidden layer is established. The error of the predicted value of concrete compressive strength is less than 4%, which meets the control requirements of engineering test accuracy. The predicted model can be used to predict the compressive strength of concrete.

An Improved Particle Swarm Optimization Algorithm Of Radial Basis Neural Network

The core issues of RBF network design are to design the minimum structure neural networks that can meet the accuracy requirements, in order to ensure the generalization ability of the network. For the purpose of simplifying the structure of RBF network, proposes a learning method of RBF network based on improved particle swarm. The method automatically constructs frugal structure of RBF network model by the combining algorithm of regularized least squares method and Doptimal experimental design; chooses three learning parameters of the combining algorithm that can affect network generalization performance by the improved particle swarm optimization algorithm. By nonlinear time series modeling, verifies the effectiveness of the method in this paper.

Voice Activity Detection Algorithm based on Improved Radial Basis Function Neural Network

Voice activity detection (VAD) is the key of voice recognition, voice synthesis and speech-sound enhancement.For the sake of improve the accuracy and robustness of speech endpoint detection system. Combining the advantages of adaptive genetic algorithm (AGA) and improved radial basis function network (RBF) defects in existing learning methods. This paper presents a comprehensive detection method-- Adaptive genetic algorithm radial basis function network. This method uses adaptive genetic algorithm to simultaneously optimize the center, the width and the structure of RBF network. The method using wavelet analysis to extract the characteristics of the speech signal, use them as an input amount to the radial basis function networks. Establish voice detection system model, this method enhance the accuracy of the detection system and has better robustness.

Method and Application of RBF Network Structure Optimization

The hidden unit number of RBF neural networks directly influences the performances of the whole net. A new strategy to prune the hidden units based on the singular value decomposition (SVD) of matrixes is proposed in the paper. At the basis of a structure involving enough more hidden units, the paper analyzes the outputs corresponding to some training samples with the SVD method and finds out the internal relations of them, then removes redundant ones according to the contribution rate of every hidden unit to the whole network, simplifies the structure of RBF neural network at last. The optimized network has strong generalization ability with simpler structure. At the end of this paper the new strategy is successfully used in the main steam system modeling of power plant and confirmed by simulation experiments.

Seismic Damage Prediction of Multistory Brick Buildings Based on RBF Neural Network Model

Seismic damage prediction of multistory brick buildings is a multi-factor nonlinear complex problems, this paper analyzed the deficiencies of the traditional methods for predicting the seismic damage, so a prediction model of multistory brick buildings based on RBF Neural Network model is established, RBF network structure elements and parameters are studied and obtained. With examples the research proved that the prediction results are similar to the actual seismic damage to multistory brick buildings by the RBP neural network model, the analytic method and process discussed in this paper can also be applied to seismic damage prediction of brick buildings of urban earthquake disaster prevention planning.

Data dimensionality reduction with application to simplifying rbf network structure and improving classification performance

For high dimensional data, if no preprocessing is carried out before inputting patterns to classifiers, the computation required may be too heavy. For example, the number of hidden units of a radial basis function (RBF) neural network can be too large. This is not suitable for some practical applications due to speed and memory constraints. In many cases, some attributes are not relevant to concepts in the data at all. In this paper, we propose a novel separability-correlation measure (SCM) to rank the importance of attributes. According to the attribute ranking results, different attribute subsets are used as inputs to a classifier, such as an RBF neural network. Those attributes that increase the validation error are deemed irrelevant and are deleted. The complexity of the classifier can thus be reduced and its classification performance improved. Computer simulations show that our method for attribute importance ranking leads to smaller attribute subsets with higher accuracies compared with the existing SUD and Relief-F methods. We also propose a modified method for efficient construction of an RBF classifier. In this method we allow for large overlaps between clusters corresponding to the same class label. Our approach significantly reduces the structural complexity of the RBF network and improves the classification performance.