Neuro-Fuzzy Networks

In any transport system, especially at industrial railway junctions, it is fundamentally important to build an effective timetable (traffic schedule) to regulate traffic flows. The task is complicated by the high dimensionality of the railway network of the node, the large number of variable parameters associated with scheduling the use of a traction resource (locomotives) during operation for sorting wagons and transporting payloads (ore, fuel, finished products and empty wagons). The problem is that most plotting problems are NP-hard, i.e. the algorithms for solving them, used to automate the process, may require an unacceptably long execution time by traditional methods of solving this problem (sequential, using reference information; method of thread laying). The article deals with the issues of building a mathematical model for dispatching an industrial railway junction to minimize the time of using locomotives in order to increase the efficiency of its operation. The mathematical model uses the technique of neuro-fuzzy computing to determine the parameters for identifying fuzzy systems and calculating the priorities of operations in the framework of creating a flexible schedule for the decision support system of the dispatching service. The results of modeling and recommendations on the use of the developed methodology are presented.

In this paper, a soft computing approach is presented for modeling electrical power generating plants in order to characterize the essential dynamic behavior of the plant subsystems. The structure of the soft computing method consists of fuzzy logic, neural networks and genetic algorithms. The measured data from a complete set of field experiments is the basis for training the models including the extraction of linguistic rules and membership functions as well as adjusting the other parameters of the fuzzy model. The genetic algorithm is applied to the modeling approach in order to optimize the procedure of the training. Comparison between the responses of the proposed models with the responses of the plants validates the accuracy and performance of the modeling approach. A similar comparison between the responses of these models with the models obtained based on the thermodynamical and physical relations of the plant shows the effectiveness and feasibility of the developed model in terms of more accurate and less deviation between the responses of the models and the corresponding subsystems.

In this paper, an application of dynamic neuro-fuzzy systems is presented for modeling the subsystems of the heat recovery steam generator (HRSG). The dynamic neuro-fuzzy models were developed based on the formal NARX models topology. The clustering techniques were employed to define the structure of the fuzzy models by dividing the entire operating regions into smaller subspaces. The optimal cluster centers and corresponding membership functions are captured by FCM, where the parameters of consequent were adjusted by recursive LSE method. A comparison between the responses of the proposed models and the responses of the plants ware preformed, which validates the accuracy and performance of the modeling approach.

In this paper, a soft computing approach is presented for modeling electrical power generating plants in order to characterize the essential dynamic behavior of the plant subsystems. The structure of the soft computing method consists of fuzzy logic, neural networks and genetic algorithms. The measured data from a complete set of field experiments is the basis for training the models including the extraction of linguistic rules and membership functions as well as adjusting the other parameters of the fuzzy model. The genetic algorithm is applied to the modeling approach in order to optimize the procedure of the training. Comparison between the responses of the proposed models with the responses of the plants validates the accuracy and performance of the modeling approach. A similar comparison between the responses of these models with the models obtained based on the thermodynamical and physical relations of the plant shows the effectiveness and feasibility of the developed model in terms of more accurate and less deviation between the responses of the models and the corresponding subsystems.

Induction motors have gained wide acceptance with respect to industrial and domestic drives due to their minimal maintenance requirements and low costs. However, speed and position control of an AC induction motors exhibits complex characteristics, various control methods have been adopted for the speed and position control of the motors which have exhibited certain limitations like poor accuracy and high harmonic distortion. A hybrid network which combines the Neural Network and the Fuzzy Logic System known as the neuro-fuzzy networks is developed in this work to control the speed and position of a three phase inductor motor. The stator voltage model was used to generate the data used to train the neuro-fuzzy network for the control of an inductor motor's position while a speed-frequency model was used to generate the data used to train the network for speed control of a typical industrial three phase inductor motor. Developed network for position control was implemented directly in matlab using neuro fuzzy tool box while the developed network for the control of speed implementation was done in the existing state vector pulse width modulation (SVPWM) technique for speed control of inductor motor. The integrated scheme known as neuro-fuzzy state vector pulse width modulation (NFSVPWM) was implemented in Simulink. Results show that neuro-fuzzy network for position control which achieved complete 360 degree detection and also having 50% improvement on the accuracy of the existing methods. The NFSVPWM improved the accuracy of SVPWM by 1.87% while achieving a zero total harmonic distortion (THD). These results can be deployed to industries by implementing them digitally using microcontrollers as target devices.

ABSTRACT

This article discusses issues related to problems encountered on electrical energy measurement due to harmonics distortions in the network. Such distortions may be able to cause considerable errors and consequently an undue amount of energy collection by the supplier. In view of this problem, the objective of this article is to make the process of measuring more efficient electricity through concepts related to Artificial Neural Networks and Neuro-Fuzzy Networks.

RESUMO

Este artigo aborda assuntos relacionados aos problemas encontrados na medição de energia elétrica em decorrência de distorções harmônicas da rede. Tais distorções podem ser capazes de provocar erros consideráveis, acarretando, consequentemente, em cobrança de uma quantidade indevida de energia pela empresa fornecedora. Tendo em vista esse problema, o objetivo deste artigo consiste em tornar o processo de medição de energia elétrica mais eficiente por meio de conceitos referentes a Redes Neurais Artificias e Redes Neuro-Fuzzy. Palavras-chave: Medição de energia elétrica. Distorções Harmônicas. Redes Neurais Artificiais. Redes Neuro-Fuzzy.

Increasing demand for electricity and growing need for more and safer power generation has motivated investigation into dynamic analysis of power plants to design more sophisticated and reliable control systems. In this paper, simple first order models are developed for the subsystems of a subcritical once through boiler, based on thermodynamics principles and energy-mass balance, together with parameter estimation routines. These routines are applied on the experimental data obtained from a complete set of field experiments. However, since most of processes in boiler are categorized as multi input and multi output systems, mathematical boiler models which are derived from physical structure and parameters estimation routines lead to a time consuming procedure, and employing such models in control algorithms becomes so complex. Therefore, to improve the dynamics modeling, a concise multilayer neuro fuzzy model of the boiler is developed. Next, these two models are compared based on the performance of the real system. This comparison validates the accuracy of both original and neuro fuzzy models, while the latter can be successfully employed in modern model-based control systems. Finally, a new Fuzzy P 2 ID controller is developed to use for superheaters temperature control. Simulation results show very good performance of this controller in terms of more accurate and less fluctuation in the temperature of corresponding subsystems, compared to the existing classic controllers.

References 205-4487 Abstract References 205-4571 Abstract References 205-4634 Abstract References 205-4637 Abstract References 205-4677 Abstract References Abstract References 205-4850 Abstract References 205-4874 Abstract References 205-4878 CONVERSION OF LADDER DIAGRAM TO PETRI NET USING MODULE SYNTHESIS TECHNIQUE Abstract References 205-4901 Abstract References 205-4934 Abstract References Issue: 2 Free 205-4517 Abstract References 205-4527 Abstract References 205-4589 Abstract References 205-4689 Abstract References 205-4747 Abstract References 205-4763 Abstract References 205-4904 Abstract References 205-4910 Abstract References 205-4918 Abstract References 205-5006 Abstract References 205-5100 Abstract References 205-5111 Abstract References 205-4809 Abstract References Abstract References 205-4823 Abstract References 205-4841 Abstract References 205-4870 Abstract References 205-4889 Abstract References 205-4961 Abstract References 205-4996 Abstract References 205-5009 Abstract References 205-5020 Abstract References 205-5035 Abstract References Issue: 4 Free 205-4795 Abstract References 205-4808 Abstract References 205-4842 Abstract References 205-4945 Abstract References 205-4976 SECOND-ORDER CONE OPTIMIZATION APPROACH TO GROUNDWATER QUANTITY MANAGEMENT Abstract References 205-4986 Abstract References Abstract References 205-5150 Abstract References 205-5167 Abstract References

In this paper, a soft computing approach is presented for modeling electrical power generating plants in order to characterize the essential dynamic behavior of the plant subsystems. The structure of the soft computing method consists of fuzzy logic, neural networks and genetic algorithms. The measured data from a complete set of field experiments is the basis for training the models including the extraction of linguistic rules and membership functions as well as adjusting the other parameters of the fuzzy model. The genetic algorithm is applied to the modeling approach in order to optimize the procedure of the training. Comparison between the responses of the proposed models with the responses of the plants validates the accuracy and performance of the modeling approach. A similar comparison between the responses of these models with the models obtained based on the thermodynamical and physical relations of the plant shows the effectiveness and feasibility of the developed model in terms of more accurate and less deviation between the responses of the models and the corresponding subsystems. r

In this paper, an application of model-based coordinated control concept for improving the performance and maneuverability of once-through power plants is presented. In this structure, neuro-fuzzy based Hammerstein models are employed as the core of feedforward (FF) controller to generate the reference trajectories for the plant’s subsystems. The set points for fuel, electrical power and steam pressure are provided by FF controller with respect to load demands. In order to diminish the effect of disturbances and to compensate deviations between variables and corresponding feedforward setpoints, feedback controllers are considered. The error signal defined by the deviation of desired boiler output pressure from its actual value is used to compensate the demand signals for turbine and boiler. The performance of the proposed control system is compared with that of a conventional turbine-follower and also turbine-follower coordinated control modes during load changes. The results indicate the effectiveness and load tracking capability of the proposed control system at different operating conditions.

Abstract: In unit steam-boiler, load control system is to provide continuous active power to satisfy the load demand of power system. The process is characterized by nonlinear and uncertainty. Neurofuzzy modelling based on local mechanism model and dynamic experiments is discussed. A relative fuzzy PI controller is constituted to control the load system in boiler-following mode. The performance of the control system has been verified by simulation and then tested on real-time process.

Keywords: load, fuzzy logic, neural networks, modelling, steam-boiler

Model-reference adaptive control with neurofuzzy methodology is derived in this paper. Associate memory network(AMN) is investigated in detail to be the possible implementation as the direct self-tuning nonlinear controller. The essence of the
neurofuzzy controller has been discussed and the local stability of the system is reached.
The performance of the model-reference adaptive neurofuzzy controller is illustrated by examples involving both linear and nonlinear systems.

Keywords: model-reference adaptive control, neural networks, fuzzy control, nonlinear

A neuro-fuzzy network predictive approach is
introduced to design a control system for nonlinear industrial
process. While the nonlinear process is modeled by neuro-fuzzy
technique containing local CARMA model, traditional
generalized minimum variance predictive control method can be
extended to nonlinear case in a neuro-fuzzy fashion. Boiler steam temperature process is chosen as realistic system for the
demonstration of the techniques discussed and the neuro-fuzzy
controller was found to provide a satisfactory performance over
the complex system.

Model reference adaptive control (MRAC) is a popular approach to control linear systems, as it is relatively simple to implement. However, the performance of the linear MRAC deteriorates
rapidly when the system becomes nonlinear. In this paper, a
nonlinear MRAC based on neurofuzzy networks is derived. Neurofuzzy networks are chosen not only because they can approximate nonlinear functions with arbitrary accuracy, but also they are compact in their supports, and the weights of the network can be readily updated on-line. The implementation of the neurofuzzy network-based MRAC is discussed, and the local stability of the system controlled by the proposed controller is established. The performance of the neurofuzzy network-basedMRACis illustrated by examples involving both linear and nonlinear systems.