Search results

The purpose of this paper is to improve the application of neural networks on vehicle engine systems for fault detecting and analysing engine oils.

Three types of neural networks are employed to find exact neural network predictor of vehicle engine oil performance and quality. Nevertheless, two oil types are analysed for predicting performance in the engine. These oils are used and unused oils. In experimental work, two accelerometers are located at the bottom of the car engine to measure related vibrations for analysing oil quality of both cases.

The results of both computer simulation and experimental work show that the radial basis neural network predictor gives good performance at adapting different cases.

The results of the proposed neural network analyser follow the desired results of the vehicle engine's vibration variation. However, this kind of neural network scheme can be used to analyse oil quality of the car in experimental applications.

As theoretical and practical studies are evaluated together, it is hoped that oil analysers and interested researchers will obtain significant results in this application area.

This paper is an original contribution on vehicle oil quality analysis using a proposed artificial neural network and it should be helpful for industrial applications of vehicle oil quality analysis and fault detection.

This paper aimed a fractional-order sliding mode-based lateral lane-change control method that was proposed to improve the path-tracking accuracy of vehicle lateral motion.

In this paper the vehicle presighting and kinematic models were established, and a new sliding mode control isokinetic convergence law was devised based on the fractional order calculus to make the front wheel turning angle approach the desired value quickly. On this basis, a fractional gradient descent algorithm was proposed to adjust the radial basis function (RBF) neuron parameter update rules to improve the compensation speed of the neural network.

The simulation results revealed that, compared to the traditional sliding mode control strategy, the designed controller eliminated the jitter of the sliding mode control, sped up the response of the controller, reduced the overshoot of the system parameters and facilitated accurate and fast tracking of the desired path when the vehicle changed lanes at low speeds.

This paper combines the idea of fractional order calculus with gradient descent algorithm, proposed a fractional-order gradient descent method applied to RBF neural network and fast adjustment the position and width of neurons.

The purpose of this paper is to propose an adaptive neural-sliding mode-based observer for the estimation and reconstruction of unknown faults and disturbances for time-varying nonlinear systems such as aircraft, to ensure preciseness in the diagnosis of fault magnitude as well as the shape without enhancement of system complexity and cost. Fault-tolerant control (FTC) strategy based on adaptive neural-sliding mode is also proposed in the existence of faults for ensuring the stability of the faulty system.

In this paper, three strategies are presented: adaptive radial basis functions neural network (ARBFNN), conventional radial basis functions neural network (CRBFNN) and integral-chain differentiator. For the purpose of enhancement of fault diagnosis and isolation, a new sliding mode-based concept is introduced for the weight updating parameters of radial basis functions neural network (RBFNN).The main objective of updating the weight parameters adaptively is to enhance the effectiveness of fault diagnosis and isolation without increasing the computational complexities of the system. Results depict the effectiveness of the proposed ARBFNN approach in fault detection (FD) and approximation compared to CRBFNN, integral-chain differentiator and schemes existing in literature. In the second step, the FTC strategy is presented separately for each observer in the presence of unknown faults and failures for ensuring the stability of the system, which is validated on Boeing 747 100/200 aircraft.

The proposed adaptive neural-sliding mode approach is investigated, which depicts more effectiveness in numerous situations such as faults, disturbances and uncertainties compared to algorithms used in literature. In this paper, both the fault approximation and isolation and the fault tolerance approaches are studied.

For the enhancement of safety level as well as for avoiding any kind of damage, timely FD and fault tolerance have always had a significant role; therefore, the algorithms proposed in this research ensure the tolerance of faults and failures, which plays a vital role in practical life for avoiding any kind of damage.

In this study, a new neural-sliding mode concept is adopted for the adaptive faults approximation and reconstruction, and then the FTC algorithms are studied for each observer separately, whereas in previous studies, only the fault detection and isolation (FDI) or the fault tolerance problems were studied. Results demonstrate the effectiveness of the proposed strategy compared to the approaches given in the literature.

The purpose of this paper is to implement the hardware structure for radial basis function (RBF) neural network based on stochastic logic computation.

The hardware implementation of artificial neural networks (ANNs) has a complicated structure and is normally space consuming due to huge size of digital multiplication, addition/subtraction, non‐linear activation function, etc. Also the unavailability of ANN hardware at an attractive price limits its use for real time applications. In stochastic logic theory, the real numbers are converted to random streams of bits instead of a binary number. The performance of the proposed structure is analyzed using very high speed integrated circuit hardware description language.

Stochastic theory‐based arithmetic and logic approach provides a way to carry out complex computation with very simple hardware and very flexible design of the system. The Gaussian RBF for hidden layer neuron is employed using stochastic counter that reduces the hardware resources significantly. The number of hidden layer neurons in RBF neural network structure is adaptively varied to make it an intelligent system.

The paper outlines the stochastic neural computation on digital hardware for implementing radial basis neural network. The structure has considered the optimized usage of hardware resources.

Omnidirectional mobile platforms are still plagued by the problem of heading deviation. In four-Mecanum-wheel systems, this problem arises from the phenomena of dynamic imbalance and slip of the Mecanum wheels while driving. The purpose of this paper is to analyze the mechanism of omnidirectional motion using Mecanum wheels, with the aim of enhancing the heading precision. A proportional-integral-derivative (PID) setting control algorithm based on a radial basis function (RBF) neural network model is introduced.

In this study, the mechanism of omnidirectional motion using Mecanum wheels is analyzed, with the aim of enhancing the heading precision. A PID setting control algorithm based on an RBF neural network model is introduced. The algorithm is based on a kinematics model for an omnidirectional mobile platform and corrects the driving heading in real time. In this algorithm, the neural network RBF NN2 is used for identifying the state of the system, calculating the Jacobian information of the system and transmitting information to the neural network RBF NN1.

The network RBF NN1 calculates the deviations ?Kp, ?Ki and ?Kd to regulate the three coefficients Kp, Ki and Kd of the heading angle PID controller. This corrects the driving heading in real time, resolving the problems of low heading precision and unstable driving. The experimental data indicate that, for a externally imposed deviation in the heading angle of between 34º and ∼38°, the correction time for an omnidirectional mobile platform applying the algorithm during longitudinal driving is reduced by 1.4 s compared with the traditional PID control algorithm, while the overshoot angle is reduced by 7.4°; for lateral driving, the correction time is reduced by 1.4 s and the overshoot angle is reduced by 4.2°.

In this study, the mechanism of omnidirectional motion using Mecanum wheels is analyzed, with the aim of enhancing the heading precision. A PID setting control algorithm based on an RBF neural network model is introduced. The algorithm is based on a kinematics model for an omnidirectional mobile platform and corrects the driving heading in real time. In this algorithm, the neural network RBF NN2 is used for identifying the state of the system, calculating the Jacobian information of the system and transmitting information to the neural network RBF NN1. The method is innovative.

The increasing complexity of industrial systems is at the heart of the development of many fault diagnosis methods. The artificial neural networks (ANNs), which are part of these methods, are widely used in fault diagnosis due to their flexibility and diversification which makes them one of the most appropriate fault diagnosis methods. The purpose of this paper is to detect and locate in real time any parameter deviations that can affect the operation of the blowout preventer (BOP) system using ANNs.

The starting data are extracted from the tables of the HAZOP (HAZard and OPerability) method where the deviations of the parameters of normal BOP operating (pressure, flow, level and temperature) are associated with an initial rule base for establishing cause and effect of relationships between the causes of deviations and their consequences; these data are used as a database for the neural network. Three ANNs were used, the multi-layer perceptron network (MLPN), radial basis functions network (RBFN) and generalized regression neural networks (GRNN). These models were trained and tested, then, their comparative performances were presented. The respective performances of these models are highlighted following their application to the BOP system.

The performances of the models are evaluated using determination coefficient (R2), root mean square error (RMSE) and mean absolute error (MAE) statistics and time execution. The results of this study show that the RMSE, MAE and R2 values of the GRNN model are better than those corresponding to the RBFN and MLPN models. The GRNN model can be applied with better performance, to establish a diagnostic model that can detect and to identify the different causes of deviations in the parameters of the BOP system.

The performance of the trained network is found to be satisfactory for the real-time fault diagnosis. Therefore, future studies on modeling the BOP system with soft computing techniques can be concentrated on the ANNs. Consequently, with the use of these techniques, the performance of the BOP system can be ensured performing only a limited number of monitoring operations, thus saving engineering effort, time and funds.

Global optimization in electrical engineering using stochastic methods requires usually a large amount of CPU time to locate the optimum, if the objective function is calculated either with the finite element method (FEM) or the boundary element method (BEM). One approach to reduce the number of FEM or BEM calls using neural networks and another one using multiquadric functions have been introduced recently. This paper compares the efficiency of both methods, which are applied to a couple of test problems and the results are discussed.

Sensor arrays and pattern recognition-based electronic nose (E-nose) is a typical detection and recognition instrument for indoor air quality (IAQ). The E-nose is able to monitor several pollutants in the air by mimicking the human olfactory system. Formaldehyde concentration prediction is one of the major functionalities of the E-nose, and three typical machine learning (ML) algorithms are most frequently used, including back propagation (BP) neural network, radial basis function (RBF) neural network and support vector regression (SVR).

This paper comparatively evaluates and analyzes those three ML algorithms under controllable environment, which is built on a marketable sensor arrays E-nose platform. Variable temperature (T), relative humidity (RH) and pollutant concentrations (C) conditions were measured during experiments to support the investigation.

Regression models have been built using the above-mentioned three typical algorithms, and in-depth analysis demonstrates that the model of the BP neural network results in a better prediction performance than others.

Finally, the empirical results prove that ML algorithms, combined with low-cost sensors, can make high-precision contaminant concentration detection indoor.

Diagnostics of electrical machines is a very important task. The purpose of this paper is the presentation of coupling three numerical techniques, a finite element analysis, a signal analysis and an artificial neural network, in diagnostics of electrical machines. The study focused on detection of a time-varying inter-turn short-circuit in a stator winding of induction motor.

A finite element method is widely used for the calculation of phase current waveforms of induction machines. In the presented results, a time-varying inter-turn short-circuit of stator winding has been taken into account in the elaborated field-circuit model of machine. One of the time-varying short-circuit symptoms is a time-varying resistance of shorted circuit and consequently the waveform of phase current. A general regression neural network (GRNN) has been elaborated to find a number of shorted turns on the basis of fast Fourier transform (FFT) of phase current. The input vector of GRNN has been built on the basis of the FFT of phase current waveform. The output vector has been built upon the values of resistance of shorted circuit for respective values of shorted turns. The performance of the GRNN was compared with that of the multilayer perceptron neural network.

The GRNN can contribute to better detection of the time-varying inter-turn short-circuit in stator winding than the multilayer perceptron neural network.

It is argued that the proposed method based on FFT of phase current and GRNN is capable to detect a time-varying inter-turn short-circuit. The GRNN can be used in a health monitoring system as an inference module.

The objective of this research is to investigate various neural network (NN) observer techniques for sensors fault identification and diagnosis of nonlinear system in consideration of numerous faults, failures, uncertainties and disturbances. For the importunity of increasing the faults diagnosis and reconstruction preciseness, a new technique is used for modifying the weight parameters of NNs without enhancement of computational complexities.

Various techniques such as adaptive radial basis functions (ARBF), conventional radial basis functions, adaptive multi-layer perceptron, conventional multi-layer perceptron and extended state observer are presented. For increasing the fault detection preciseness, a new technique is used for updating the weight parameters of radial basis functions and multi-layer perceptron (MLP) without enhancement of computational complexities. Lyapunov stability theory and sliding-mode surface concepts are used for the weight-updating parameters. Based on the combination of these two concepts, the weight parameters of NNs are updated adaptively. The key purpose of utilization of adaptive weight is to enhance the detection of faults with high accuracy. Because of the online adaptation, the ARBF can detect various kinds of faults and failures such as simultaneous, incipient, intermittent and abrupt faults effectively. Results depict that the suggested algorithm (ARBF) demonstrates more confrontation to unknown disturbances, faults and system dynamics compared with other investigated techniques and techniques used in the literature. The proposed algorithms are investigated by the utilization of quadrotor unmanned aerial vehicle dynamics, which authenticate the efficiency of the suggested algorithm.

The proposed Lyapunov function theory and sliding-mode surface-based strategy are studied, which shows more efficiency to unknown faults, failures, uncertainties and disturbances compared with conventional approaches as well as techniques used in the literature.

For improvement of the system safety and for avoiding failure and damage, the rapid fault detection and isolation has a great significance; the proposed approaches in this research work guarantee the detection and reconstruction of unknown faults, which has a great significance for practical life.

In this research, two strategies such Lyapunov function theory and sliding-mode surface concept are used in combination for tuning the weight parameters of NNs adaptively. The main purpose of these strategies is the fault diagnosis and reconstruction with high accuracy in terms of shape as well as the magnitude of unknown faults. Results depict that the proposed strategy is more effective compared with techniques used in the literature.