Papers by Keyword: Radial Basis Function

Abstract: The availability of electrical energy is essential for human progress and economic development. Renewable energy solutions, including waste-to-energy (WtE) systems, present sustainable alternatives but require advanced control strategies for optimal performance. This research aims to enhance the control of drum level, temperature, and pressure in WtE steam boilers at Ethiopia's Reppie power plant. The existing Programmable Logic Controller (PLC) system is limited in its ability to predict future states and handle nonlinear system behaviors. To overcome these challenges, a Radial Basis Function Autoregressive with Exogenous input (RBF-ARX) model was developed and integrated with a Model Predictive Controller (MPC). The results demonstrate that the MPC approach significantly surpasses the performance of the Linear Quadratic Regulator (LQR) in terms of control efficiency. For temperature control, the MPC achieves a settling time of 0.3955 seconds and a rise time of 0.0195 seconds, compared to LQR's 5.99 seconds. Similarly, for pressure control, the MPC achieves a settling time of 0.6678 seconds, outperforming the LQR's 12.507 seconds. Drum level regulation further showcases the superiority of MPC, with a settling time of 0.5223 seconds versus the LQR's 8.302 seconds. This proposed RBF-ARX-based MPC framework not only optimizes control efficiency at Reppie but also demonstrates scalability and applicability to other WtE plants, enhancing operational performance under varying conditions. MATLAB/Simulink was used for the modeling and simulation, confirming the robustness of this approach for global adoption in WtE systems.

Abstract: Despite recent advances in the finite element method, mesh distortion due to large deformations may still occur in some problems such as footings subjected to deep penetration or objects penetrating into a soil layer. In order to overcome mesh distortion, robust remeshing techniques are required. In this paper the performance of four remeshing methods is studied by analysing a free falling penetrometer penetrating into an undrained layer of soil. These techniques are implemented within the framework of the Arbitrary Lagrangian-Eulerian method and include the refinement based on an elastic relocation, a technique based on the Radial Basis Functions, the Spring Analogy method, and the Elastic Hardening method. Since one of the challenging problems in a large deformation analysis is dealing with complex boundary shapes, a scheme based on the B-Splines used in isogeometric analysis is also presented here.

Abstract: Optimum design in civil structures like domes and vaults is a very old and ongoing research field. These structures are preferably designed to transport loads via membrane action. In this paper, we have considered a reinforced concrete dome and vault, where the bending moment and strain energy were used as objective function to be minimized using genetic algorithm, and model reduction method by proper orthogonal decomposition based on the results of finite element analysis of gradually changed design parameters. The proposed approach results are of a high accuracy compared to finite element based optimization.

Abstract: The interest in the localisation of wireless sensor networks has grown in recent years. A variety of machine-learning methods have been proposed in recent years to improve the optimisation of the complex behaviour of wireless networks. Network administrators have found that traditional classification algorithms may be limited with imbalanced datasets. In fact, the problem of imbalanced data learning has received particular interest. The purpose of this study was to examine design modifications to neural networks in order to address the problem of cost optimisation decisions and financial predictions. The goal was to compare four learning-based techniques using cost-sensitive neural network ensemble for multiclass imbalance data learning. The problem is formulated as a combinatorial cost optimisation in terms of minimising the cost using meta-learning classification rules for Naïve Bayes, J48, Multilayer Perceptions, and Radial Basis Function models. With these models, optimisation faults and cost evaluations for network training are considered.

Abstract: — Neural networks are frequently used as a classifier for tasks in many classifications. However there are disadvantages in terms of amount of training data required, and length of training time. This paper, develop an intelligent diagnosis system for zinc oxide (ZnO) surge arrester fault classification. First the features were extracted from 600 ZnO surge arrester thermal images and leakage currents. Then these features were presented to several neural network architectures to investigate the most suitable network model for classifying the ZnO surge arrester fault condition effectively. Three classification models were used namely feed forward back propagation (FFBP), radial basis function (RBF) and learning vector quantization (LVQ) algorithm. The performance of the networks was compared based on resulted of misclassify and correct rate. The method was evaluated using 24 testing datasets. Comparison results showed that LVQ was the best training algorithm for the ZnO surge arrester fault classification compared to the others system. Also the LVQ is faster than FFBP and RBF.

Abstract: A mathematical prediction model has been developed in order to detect particles with a diameter of 10 micrometers or less (PM-10) that are responsible for adverse health effects because of their ability to cause serious respiratory conditions in areas of high pollution such as Chiang Mai City moat area. The prediction model is based on 3 types of Artificial Neural Networks (ANNs), including Multi-layer perceptron (MLP-NN), Radial basis function (RBF-NN), and hybrid of RBF and Genetic algorithm (RBF-NN-GA). The model uses 8 input variables to predict PM-10, consisting of 4 air pollution substances ( CO, O₃, NO₂ and SO₂) and 4 meteorological variables related PM-10 (wind speed, temperature, atmospheric pressure and relative humidity). These 3 types of ANN have proved efficient instrument in predicting the PM-10. However, the performance of RBF-NN was superior in comparison with MLP-NN and RBF-NN-GA respectively.

Abstract: Downscaling Global Circulation Model (GCM) output is important in order to understand the present climate as well as future climate changes at local scale. In this study, Radial basis function (RBF) neural network was used to downscale the mean monthly rainfall in an arid coastal region located in Baluchistan province of Pakistan. The RBF model was used to downscale monthly rainfall from National Center for environmental prediction (NCEP) reanalysis dataset at four observation stations in the area. The potential predictors were selected using principal component analysis of NCEP variables at grid points located around the study area. Power transformation method was used to remove the bias in the prediction. The results showed that the RBF model was able to establish a good relation between NCEP predictors and local rainfall. The power transformation method was also found to perform well to correct errors in prediction. It can be concluded that RBF and power transformation methods are reliable and effective methods for downscaling rainfall in an arid coastal region.

Abstract: In order to improve the modeling efficiency of RBF neural network, an Artificial Fish Swarm Algorithm (AFSA) training algorithm with an adaptive mechanism is proposed. In the training algorithm, the search step size and visible domain of AFSA algorithm can be adjusted dynamically according to the convergence characteristics of artificial fish swarm, and then the improved AFSA algorithm is used to optimize the parameters of RBF neural network. The example shows that, the proposed model is a better approximation performance for the nonlinear function.

Abstract: Meshless collocation method based on generalized multiquadrics radial basis function is used to study the free vibration of simply supported laminated composite plates. The generalized multiquadric radial basis function g=[r²+c²]^q has the exponent q and shape parameter c that play an important role in the accuracy of the approximation. Genetic algorithm is utilized to optimize the shape parameter and exponent of generalized multiquadrics radial basis function. The natural frequencies of simply supported laminated composite plates are calculated using the generalized multiquadrics with optimal shape parameter, exponent and compared with the analytical solutions.

Abstract: Radial Basis Function (RBF) interpolation and trilinear interpolation techniques are compared in the soot particle tracking inside the cylinder of a direct injection engine. The interpolation techniques are used separately in an efficient routine written in Matlab codes which is developed to track the movement or pathline of soot particles in the engine operation cycle ranged from inlet valve closing (IVC) to exhaust valve opening (EVO). Soot particles are treated as a massless body and in spherical shape which will move under the influence of bulk gases flow inside the cylinder. Movement of soot particles are examined through the selection factors of particle's initial coordinate (r,Ɵ,z) and soot concentration level at different instant crack angle. Results obtained from both interpolation techniques are compared and good agreement is achieved with some minor relative difference. However, RBF interpolation has wider applications potential where it can be applied to variety type of mesh geometry as compared to trilinear interpolation which is best used in mesh with hexahedral shape.