RBFNN Model Research Articles

Medium term load demand forecast of Kano zone using neural network algorithms

Electricity load forecasting refers to projection of future load requirements of an area or region or country through appropriate use of historical load data. One of several challenges faced by the Nigerian power distribution sectors is the overloaded power distribution network which leads to poor voltage distribution and frequent power outages. Accurate load demand forecasting is a key in addressing this challenge. This paper presents a comparison of generalized regression neural network (GRNN), feed-forward neural network (FFNN) and radial basis function neural network for medium term load demand estimation. Experimental data from Kano electricity distribution company (KEDCO) were used in validating the models. The simulation results indicated that the neural network models yielded promising results having achieved a mean absolute percentage error (MAPE) of less than 10% in all the considered scenarios. The generalization capability of FFNN is slightly better than that of RBFNN and GRNN model. The models could serve as a valuable and promising tool for the forecasting of the load demand.

Coordinate Transformation between Global and Local Datums Based on Artificial Neural Network with K-Fold Cross-Validation: A Case Study, Ghana

The popularity of Artificial Neural Network (ANN) methodology has been growing in a wide variety of areas in geodesy and geospatial sciences. Its ability to perform coordinate transformation between different datums has been well documented in literature. In the application of the ANN methods for the coordinate transformation, only the train-test (hold-out cross-validation) approach has usually been used to evaluate their performance. Here, the data set is divided into two disjoint subsets thus, training (model building) and testing (model validation) respectively. However, one major drawback in the hold-out cross-validation procedure is inappropriate data partitioning. Improper split of the data could lead to a high variance and bias in the results generated. Besides, in a sparse dataset situation, the hold-out cross-validation is not suitable. For these reasons, the K-fold cross-validation approach has been recommended. Consequently, this study, for the first time, explored the potential of using K-fold cross-validation method in the performance assessment of radial basis function neural network and Bursa-Wolf model under data-insufficient situation in Ghana geodetic reference network. The statistical analysis of the results revealed that incorrect data partition could lead to a false reportage on the predictive performance of the transformation model. The findings revealed that the RBFNN and Bursa-Wolf model produced a transformation accuracy of 0.229 m and 0.469 m, respectively. It was also realised that a maximum horizontal error of 0.881 m and 2.131 m was given by the RBFNN and Bursa-Wolf. The obtained results per the cadastral surveying and plan production requirement set by the Ghana Survey and Mapping Division are applicable. This study will contribute to the usage of K-fold cross-validation approach in developing countries having the same sparse dataset situation like Ghana as well as in the geodetic sciences where ANN users seldom apply the statistical resampling technique.

Assessment of Machine Learning Techniques for Monthly Flow Prediction

Monthly flow predictions provide an essential basis for efficient decision-making regarding water resource allocation. In this paper, the performance of different popular data-driven models for monthly flow prediction is assessed to detect the appropriate model. The considered methods include feedforward neural networks (FFNNs), time delay neural networks (TDNNs), radial basis neural networks (RBFNNs), recurrent neural network (RNN), a grasshopper optimization algorithm (GOA)-based support vector machine (SVM) and K-nearest neighbors (KNN) model. For this purpose, the performance of each model is evaluated in terms of several residual metrics using a monthly flow time series for two real case studies with different flow regimes. The results show that the KNN outperforms the different neural network configurations for the first case study, whereas RBFNN model has better performance for the second case study in terms of the correlation coefficient. According to the accuracy of the results, in the first case study with more input features, the KNN model is recommended for short-term predictions and for the second case with a smaller number of input features, but more training observations, the RBFNN model is suitable.

Accurate Identification of the Sex and Species of Silkworm Pupae Using Near Infrared Spectroscopy

The present study proposes a novel method to discriminate the sex and species of silkworm pupae using NIR spectroscopy (800–2778 nm). The spectra from 840 silkworm pupae were collected then divided into a calibration set (700) and a test set (140) using the Kennard–Stone (KS) algorithm. The recognition models were built using the radial basis function and neural network (RBF–NN) and support vector machine (SVM) approaches. The species and sex identification results using the RBF–NN and SVM models based on full spectral data achieved a low accuracy of 5% and 33.57%, respectively. To improve the accuracy and decrease the processing time, both principal component analysis (PCA) and linear discriminant analysis (LDA) were used to reduce the data dimensions. The performance of the optimized SVM model (92.14%) was much better than the RBF–NN model (19.29%) based on PCA. Overall, the best discrimination results were obtained using the RBF–NN and SVM models based on LDA, providing an accuracy of 100%. These promising results have shown that the LDA–SVM and LDA–RBF–NN models can accurately recognize the sex and species of silkworm pupae using NIR spectroscopy.

Forecasting of Horizontal Gas Well Production Decline in Unconventional Reservoirs using Productivity, Soft Computing and Swarm Intelligence Models

The most widely used production decline forecasting tools are numerical reservoir simulation, material balance estimates and advanced methods of production decline analysis. Besides, these existing production decline evaluation techniques for unconventional reserves estimations have underlying limitations and assumptions incorporated into their formulation which can result into under- and overestimation. This further raises the debate about which decline curve analysis (DCA) is better than the other for unconventional reservoirs predictions. Currently, data-driven artificial neural network (ANN) has emerged as a new paradigm capable of mapping complex functional relationships. In this present study, ANN technology was used to calibrate tight gas carbonate field historical declining trends which exhibit high early peak production rate and quick decline. Hereafter, making reliable predictions posed as a challenge for the complex target field and hard computing protocols. Therefore, this research applied and tested the capability of backpropagation artificial neural network (BPANN), radial basis function neural network (RBNN) and generalized regression neural network (GRNN) as DCA techniques for predicting the historical production decline trends of an ultra-low porosity and permeability tight gas reservoir. The optimum trained ANN DCA models developed were validated with another surrounding well’s dataset, producing acceptable results in agreement with the actual field data. The GRNN DCA model’s testing performance was poor, and its smoothing parameter was optimized with particle swarm optimization (PSO) algorithm which gave satisfactory results comparable to standalone BPANN and RBFNN DCA models. Furthermore, the optimum ANN DCA models’ generalization strength across the entire field dataset revealed that the developed models’ predictions were robust as compared to the data-driven Arps hyperbolic and power law exponential DCA models. This was evident from the statistical performance criteria employed which indicated that BPANN, RBFNN and PSO-GRNN DCA models are plausible better fit models for matching the target field’s historical decline performances. Also, a novel non-Darcy flow horizontal well productivity evaluation model for the target field was developed based on stress sensitivity coefficient (SSC) of permeability, tortuosity factor, Klinkenberg effect, near-wellbore turbulence effect and threshold pressure gradient (TPG) for validating the ANN DCA models predictions. The productivity model was validated with published horizontal well model with closely matched results. For inflow performances, the horizontal well model with turbulence minimizes negative effects on non-Darcy flow rates than without turbulence. Additionally, pressure drawdown influences the tight gas well productivity such that the lower the pressure drawdown, the smaller the tight gas well productivity. The operating points of the tight gas well were determined through inflow performance relation and tubing performance relation at different SSC and TPG for $$ 3{\raise0.5ex\hbox{$\scriptstyle 1$} \kern-0.1em/\kern-0.15em \lower0.25ex\hbox{$\scriptstyle 2$}} $$ in tubing size. In another case, synthetic unconventional simulation data with long production history were used for future forecast of 5, 24 and 70 months of production rate and cumulative production which gave pretty close results for all the DCA models, unlike the real field datasets, where the empirical rate-time models under and overestimate. In all, these ANN DCA models and the horizontal well productivity model derived will serve as new computational tools for complementing existing DCA techniques for better understanding of unconventional reservoirs’ production decline performance.

Safety Monitoring Model of a Super-High Concrete Dam by Using RBF Neural Network Coupled with Kernel Principal Component Analysis

Effective deformation monitoring is vital for the structural safety of super-high concrete dams. The radial displacement of the dam body is an important index of dam deformation, which is mainly influenced by reservoir water level, temperature effect, and time effect. In general, the safety monitoring models of dams are built on the basis of statistical models. The temperature effect of dam safety monitoring models is interpreted using approximate functions or the temperature values of a few points of measurement. However, this technique confers difficulty in representing the nonlinear features of the temperature effect on super-high concrete dams. In this study, a safety monitoring model of super-high concrete dams is established through the radial basis neural network (RBF-NN) and kernel principal component analysis (KPCA). The RBF-NN with strong nonlinear fitting capacity is utilized as the framework of the model, and KPCA with different kernels is adopted to extract the temperature variables of the dam temperature dataset. The model is applied to a super-high arch dam in China, and results show that the Hybrid-KPCA -RBF-NN model has high fitting and prediction precision and thus has practical application value.

Adaptive Prediction Algorithm Based on ARMA and RBFNN Models for Radar Performance

To implement health management for large complex systems, it is of great importance to monitor health online and track the trend of working performance. Fault prediction is a relatively weak but critical part in Prognostics and Health Management. When selecting characteristic parameter time series to monitor and predict the health of the system, applicability of a single model is limited. To satisfy the requirement of real-time performance and accuracy, an algorithm based on Autoregressive Moving Average Model and Radial Basis Function Neural Network Model was promoted, updating the model to predict automatically when the current model is no longer applicable. Multi-threaded parallel modeling strategy can be used in engineering application. The simulation results show that compared with algorithm employing the single model, the adaptive algorithm can effectively improve the real-time performance and accuracy.

Accurate modeling of vapor–liquid equilibria of binary mixtures of refrigerants using intelligent models

Developing simple, accurate and general models for prediction of different properties of hydrofluorocarbons (HFCs) and hydrocarbons (HCs) with hydrofluoro-olefins (HFOs) mixtures is of crucial importance in the design of new refrigeration system. In this communication, four computer based models namely Radial Basis Function Neural Network, Multilayer Perceptron Neural Network, Least Square Support Vector Machine optimized by Coupled Simulated Annealing and Adaptive Neuro Fuzzy Inference System trained by Hybrid method were used for prediction of vapor–liquid equilibrium (VLE) for binary mixtures of different HFC and HC compounds with HFO refrigerants. Results reveal that the developed models are accurate and effective for prediction of experimental VLE data for different systems. However, the RBF-NN model provides better predictions compared to other models. Moreover, the predictions of the developed models were better than the Peng-Robinson (PR) and Soave–Redlich–Kwong (SRK) equations of state (EoSs).

Thermodynamics and ANN models for predication of the equilibrium CO2 solubility in aqueous 3-dimethylamino-1-propanol solution

In this work, the CO2 equilibrium solubility of 3-dimethylamino-1-propanol (3DMA1P) is presented over the concentration range of 1–3mol/L, the temperature range of 298–333K, and the CO2 partial pressure of 3–101kPa. Several thermodynamic models (i.e. Kent-Eisenberg model, Austgen model, Hu-Chakma model, Liu et al. model) were used to correlate and predict the CO2 equilibrium solubility of 3DMA1P solution. It was found that the Liu et al. model could be considered to be an appropriate model to predict the CO2 equilibrium solubility in 3DMA1P solutions with an absolute average deviation (ADD) of 9.4%. In addition, the ANNs models (BPNN and RBFNN model) were developed and used to correlate the CO2 equilibrium solubility of 3DMA1P. It was found that the ANNs models could predict the experimental values very well with excellent ADDs of 3.0% for the BPNN model and 4.4% for the RBFNN model, respectively. In addition, a comparison of thermodynamic models and ANN models was presented in terms of prediction of CO2 solubility in 3DMA1P solution.

The analysis of solubility, absorption kinetics of CO2 absorption into aqueous 1‐diethylamino‐2‐propanol solution

In this present work, the CO2 absorption performance of aqueous 1‐diethylamino‐2‐propanol (1DEA2P) solution was studied with respect to CO2 equilibrium solubility, absorption kinetics, and absorption heat. The equilibrium solubility of CO2 in 2M 1DEA2P solution was measured over the temperature range from 298 to 333 K and CO2 partial pressure range from 8 to101 kPa. The absorption kinetics data were developed and analyzed using the base‐catalyzed hydration mechanism and artificial neural network models (radial basis function neural network [RBFNN] and back‐propagation neural network [BPNN] models) with an acceptable absolute average deviation of 10% for base‐catalyzed hydration mechanism, 2.6% for RBFNN model and 1.77% for BPNN model, respectively. The CO2 absorption heat of 1DEA2P was estimated to be −43.6 kJ/mol. In addition, the ions (1DEA2P, 1DEA2PH+, , CO32−) speciation plots of the 1DEA2P‐CO2‐H2O system were developed to further understand the reaction process of 1DEA2P with CO2. Based on a comparison with conventional amines (e.g., MEA, DEA, MDEA) and alternative amines (i.e., 1DMA2P and 4‐(diethylamino)−2‐butanol [DEAB]), 1DEA2P exhibited good performance with respect to CO2 equilibrium solubility, reaction kinetics, and CO2 absorption heat. Meanwhile, the overall evaluation of 1DEA2P for application in CCS in terms of absorption and desorption is presented, giving helpful information for the screening of these novel amines. © 2017 American Institute of Chemical Engineers AIChE J, 63: 2694–2704, 2017

Utilizing RBF-NN and ANFIS Methods for Multi-Lead ahead Prediction Model of Evaporation from Reservoir

Evaporation as a major meteorological component of the hydrologic cycle plays a key role in water resources studies and climate change. The estimation of evaporation is a complex and unsteady process, so it is difficult to derive an accurate physical-based formula to represent all parameters that effect on estimate evaporation. Artificial intelligence-based methods may provide reliable prediction models for several applications in engineering. In this research have been introduced twelve networks with the RBF-NN and ANFIS methods. These models have applied to prediction daily evaporation at Layang reservoir, located in the southeast part of Malaysia. The used meteorological data set to develop the models for prediction daily evaporation rate from water surface for Layang reservoir includes daily air temperature, solar radiation, pan evaporation, and relative humidity that measured at a case study for fourteen years. The obtained result denote to the superiority of the RBF-NN models on the ANFIS models. A comparison of the model performance between RBF-NN and ANFIS models indicated that RBF-NN method presents the best estimates of daily evaporation rate with the minimum MSE 0.0471 , MAE 0.0032, RE and maximum R2 0.963.

A Hybrid ARIMA and RBF Neural Network Model for Tourist Quantity Forecasting : A Case Study for Chiangmai Province

Applications of a single model may not be able to capture different data patterns well enough, especially in the tourist forecast problem which is often complex in nature. An autoregressive integrated moving average (ARIMA) is a famous linear model while an artificial neural network (ANN) is a promising alternative to a traditional linear method. The ARIMA model may not be adequate for nonlinear problems while ANN can well reveal the correlation of nonlinear patterns. However, overfitting due to a learning process is the main disadvantage of ANN as well as being trapped in a local optimum for parameters optimization. To improve the forecast performance of both ARIMA and ANN for high accuracy, the two hybridization models, i.e. hybrid ARIMA-RBFNN model and hybrid RBFNN-ARIMA model are employed to examine the Chiangmai’s tourist time series data. Statistics test and parameter designed experiments were used to optimize these models and the sum-square of error (SSE) was used to indicate their performances. In this case study, the hybrid RBFNN-ARIMA model has proved that the RBFNN can priori capture the non-stationary non-linear component while the fully linearly stationary residuals were accurately predicted by ARIMA. The experimental results demonstrated that the hybrid RBFNN-ARIMA model outperformed 42% by averaging over the hybrid ARIMA-RBFNN model, an improvement of hybrid ARIMA-RBFNN model, RBFNN model, and ARIMA model.

Prediction of Spherical Phase Change Material Temperature Using RBFNN Model

PCM system particularly plays an important candidate at off peak to improve thermal storage system. In this paper, predication of output temperature in integrated solar heat using RBFNN was proposed. First, a brief PCM numerical model including geometrical and operational parameters and discusses the effect this parameter on the output temperature. The working fluid temperature is uniform distributed and equal to collector output temperature. The learning data is theoretical generated by using standard solar collector integrated with 95 PCM modules. The RBFNN model has 3 input nodes representing spherical size, duct length and number of ball and two output node represented by first and last layer PCM temperature. Simulation result shows the predicted PCM temperature at first and last layer closely match with the analytical data from CFD.

A PSO-GRNN Model for Railway Freight Volume Prediction: Empirical Study from China

Purpose: The purpose of this paper is to propose a mathematical model for the prediction of railway freight volume, and therefore provide railway freight resource allocation with an accurate direction. With an accurate railway freight volume prediction, railway freight enterprises can integrate the limited resources and organize transport more reasonably. Design/methodology/approach: In this paper, a PSO-GRNN model is proposed to predict the railway freight volume. In this model, GRNN is applied to carry out the nonlinear regression analysis and output the prediction value, PSO algorithm is applied to optimize the GRNN model by searching the best smoothing parameter. In order to improve the performance of PSO algorithm, time linear decreasing inertia weight algorithm and time varying acceleration coefficient algorithm are applied in the paper. Originality/value: A railway freight volume prediction index system containing seventeen indexes from five aspects is established in this paper. And PSO-GRNN model constructed in this paper are applied to predict the railway freight volume from 2007 to 2011. Finally, an empirical study is given to verify the feasibility and accuracy of the PSO-GRNN model by comparing with RBFNN model and BPNN model. The result shows that PSO-GRNN model has a good performance in reducing the prediction error, and can be applied in actual production easily.

Genetic algorithm based multi-objective least square support vector machine for simultaneous determination of multiple components by near infrared spectroscopy

The near infrared (NIR) spectrum contains a global signature of composition, and enables to predict different proper ties of the material. In the present paper, a genetic algorithm and an adaptive modeling technique were applied to build a multiobjective least square support vector machine (MLS-SVM), which was intended to simultaneously determine the concentrations of multiple components by NIR spectroscopy. Both the benchmark corn dataset and self-made Forsythia suspense dataset were used to test the proposed approach. Results show that a genetic algorithm combined with adaptive modeling allows to efficiently search the LS-SVM hyperparameter space. For the corn data, the performance of multi-objective LS-SVM was significantly better than models built with PLS1 and PLS2 algorithms. As for the Forsythia suspense data, the performance of multi-objective LS-SVM was equivalent to PLS1 and PLS2 models. In both datasets, the over-fitting phenomena were observed on RBFNN models. The single objective LS-SVM and MLS-SVM didn't show much difference, but the one-time modeling convenience al lows the potential application of MLS-SVM to multicomponent NIR analysis.

ANALYSIS OF MICROSTRIP LINE FEED SLOT LOADED PATCH ANTENNA USING ARTIFICIAL NEURAL NETWORK

In this article, the parametric analysis of the slot-loaded microstrip line feed patch antenna is investigated using artiflcial neural network model. The bandwidths of the proposed antenna obtained at TM01, TM02, and TM03 frequency modes are 10.2GHz, 13.6GHz, and 17.2GHz, respectively. The performance of the proposed antenna is analyzed using artiflcial neural network model. The changes obtained in bandwidth due to the position of slot length and slot width are reported. The antenna parameters such as return loss, VSWR, gain and e-ciency are also calculated. The simulated results obtained with the help of IE3D simulation software are trained and tested using ANN. Theoretical results are compared with simulated and experimental ones, and they are in close agreement. Microstrip antennas have sparked interest among researchers because of their attractive features like low proflle, light weight, and conformal to mounting structures, but they have two most serious limitations, narrow bandwidth and low gain (1). Available methods for the analysis of MSA have high level of complexity. Generally, there are two methods for analysis of microstrip antenna such as numerical method and analytical method. The numerical methods is complicated compared to analytical methods and require much more time, whereas, analytical methods are easy and specifled to only some deflnite shapes of microstrip antenna. An artiflcial neural network (ANN) is used here for reducing some of these problems. ANNs are computational tools that learn from experience (training), which provide fast and accurate models for microwave modeling, simulation, and optimization. Using artiflcial neural network microstrip patch antenna was reported by Vegni and Toscano (2) in which they proposed the analysis of MSA using artiflcial neural network. Further, Mishra and Patnaik (3,4) proposed a CAD model for the design of square patch antenna and artiflcial model for efiective dielectric constant of microstrip line. Later, CAD model using spectral domain formulation (5) was also reported for the designing of rectangular patch antenna. Therefore, Guney and Sarikaya (6) proposed a comparative study of MAMDANI and Sugeno fuzzy interface system models for the resonant frequency calculation of rectangular microstrip patch antenna. Several other papers have also used ANN model to analyze and synthesize the microwave circuits (7{15). Thakare and Singhal (16) proposed the analysis of broadband slot-loaded inset feed microstrip patch antenna using ANN model. Most of these papers are based on ANN model presented only simulated and experimental results. In this paper, we provide a theoretical investigation and compare its results with simulated and experimental ones. In this paper, the microstrip line feed slot-loaded patch antenna is proposed. Its theoretical, simulated and ANN results have also been verifled experimentally. The proposed antenna is investigated for triple frequency-band operation, so that single antenna can be utilized for more than one frequency bands. The theoretical analysis of the proposed antenna is discussed using circuit theory concept based on modal expansion cavity model whereas ANN used RBFNN model.

RGBW LED 이용한 RBFNN 기반 감성조명 시스템 설계

In this paper, we introduce the LED emotional lighting system realized with the aid of both intelligent algorithm and RGB LED combined with White LED. Generally, the illumination is known as a design factor to form the living place that affects human"s emotion and action in the light- space as well as the purpose to light up the specific space. The LED emotional lighting system that can express emotional atmosphere as well as control the quantity of light is designed by using both RGB LED to form the emotional mood and W LED to get sufficient amount of light. RBFNNs is used as the intelligent algorithm and the network model designed with the aid of LED control parameters (viz. color coordinates (x and y) related to color temperature, and lux as inputs, RGBW current as output) plays an important role to build up the LED emotional lighting system for obtaining appropriate color space. Unlike conventional RBFNNs, Fuzzy C-Means(FCM) clustering method is used to obtain the fitness values of the receptive function, and the connection weights of the consequence part of networks are expressed by polynomial functions. Also, the parameters of RBFNN model are optimized by using PSO(Particle Swarm Optimization). The proposed LED emotional lighting can save the energy by using the LED light source and improve the ability to work as well as to learn by making an adequate mood under diverse surrounding conditions.

Genetic Optimization of Artificial Neural Networks to Forecast Virioplankton Abundance from Cytometric Data

Since viruses are able to influence the trophic status and community structure they should be accessed and accounted in ecosystem functioning and management models. So, this work met a set of biological, chemical and physical time series in order to explore the correlations with marine virioplankton community across different trophic gradients. The case studied is the Arraial do Cabo upwelling system, northeast of Rio de Janeiro State in Southeast coast of Brazil. The main goal is to evolve three type of artificial neural network (ANN) by genetic algorithm (GA) optimization to predict virioplankton abundance and dynamic. The input variables range from the abundance of phytoplankton, bacterioplankton and its ratios acquired by one in situ and another ex situ flow cytometers. These data were collected with weekly frequency from August 2006 to June 2007. Our results show viruses being highly correlated to their host, and that GA provided an efficient method of optimizing ANN architectures to predict the virioplankton abundance. The RBF-NN model presented the best performance to an accuracy of 97% for any period in the year. A discussion and ecological interpretations about the system behavior is also provided.

Intelligent sensor system for discrimination of material type using neural networks

Neural network based classification of material type even with the variation in the sensor parameter is investigated in this paper. The sensor is developed by means of a lightweight plunger probe and an optical mouse sensor. An experimental prototype was developed which involves bouncing or hopping of the plunger based impact probe freely on the plain surface of an object under test. The experiment is conducted to obtain the bouncing signals for plain surface of an objects kept at different distances from the probe. During the bouncing of the probe, time varying signals are generated from optical mouse that are recorded in data files on PC. Some dominant unique features are then extracted using signal processing tools to optimize neural network based classifier. The time and features of bouncing signal are related to the material type, and each material has a unique set of such properties. It is found that the sensor system is intelligent due to its ability to classify the material type even with the variation in the sensor parameter (distance between the sensor probe and plain objects). The classifiers are developed using two neural networks configurations, namely a well-known Multi-layer Perceptron Neural Networks (MLP NN), and Radial Basis Function Neural Networks (RBF NN). MLP NN and RBF NN models are designed to maximize accuracy under the constraints of minimum network dimension. The optimal parameters of MLP NN and RBF NN models based on various performance measures that include percentage classification accuracy (PCLA) on the testing data, and area under Receiver Operating Characteristics (ROC), and are determined. For the sensor data set, the PCLA of both the classifiers are found reasonable consistently in respect of rigorous testing using different data partitions. The areas under the ROC curves are close to unity. Performances of the two classifiers have been compared. It has been found that the RBF NN is more robust to noise, and epochs required for training are very less as compared to that for MLP NN.

Radial basis function neural network in fault detection of automotive engines

Fault detection and isolation have become one of the most important aspects of automobile design. A fault detection (FD) scheme is developed for automotive engines in this paper. The method uses an independent Radial Basis Function (RBF) Neural Network model to model engine dynamics, and the modelling errors are used to form the basis for residual generation. A dependent RBFNN model is a model which uses output data of a plant as a target output then use it to train the neural network, while, The independent RBFNN model is a higher accuracy than the dependent model and the errors can be detected by this model, this is because this model does not dependent on the output of the plant and it will use its output as a target, so if any faults in the plant will be not effect in the model and this faults will be detected easily and clearly. The method is developed and the performance assessed using the engine benchmark, the Mean Value Engine Model (MVEM) with Matlab/Simulink. Five faults have been simulated on the MVEM, including three sensor faults, one component fault and one actuator fault. The three sensor faults considered are 10-20% change superimposed on the outputs of manifold pressure, temperature and crankshaft speed sensors; one component fault considered is air leakage in intake manifold and Exhaust Gas Recycle (EGR); the actuator fault considered is the malfunction of fuel injector. The simulation results showed that all the simulated faults can be clearly detected in the dynamic condition throughout the operating range. Keywords: Automotive engine, independent RBFNN model, RBF neural network, fault detection International Journal of Engineering, Science and Technology, Vol. 2, No. 10, 2010, pp. 1-8