Hybrid Neuro-fuzzy Controller Research Articles

Mobile robot controller using novel hybrid system

Hybrid neuro-fuzzy controller is one of the techniques that is used as a tool to control a mobile robot in unstructured environment. In this paper a novel neuro-fuzzy technique is proposed in order to tackle the problem of mobile robot autonomous navigation in unstructured environment. Obstacle avoidance is an important task in the field of robotics, since the goal of autonomous robot is to reach the destination without collision. The objective is to make the robot move along a collision free trajectory until it reaches its target. The proposed approach uses the artificial neural network instead of the fuzzified engine then the output from it is processed using adaptive inference engine and defuzzification engine. In this approach, the real processing time is reduce that is increase the mobile robot response. The proposed neuro-fuzzy controller is evaluated subjectively and objectively with other approaches and also the processing time is taken in consideration.

COMPARISON OF CONTROLLER PERFORMANCE FOR THE ANALYSIS OF SUPERFICIAL CANCER THERAPY

The primary objective of the research work presented is to design a thermal dose controller for hyperthermia treatment using several soft computing techniques, perform simulation studies on the mathematical model of the carcinogenic tissue and compare the controller performance. The non-linear relationship between temperature and dosage, after effects of treatment and constraints on normal tissue protection makes design of thermal dose controller a challenging task. Intelligent control of thermal dose using soft computing techniques like Fuzzy Logic Controller (FLC) and Fuzzy Adaptive Learning Control Network (FALCON) has been simulated on the tissue model and the performance of the controllers are compared in terms of accuracy of output and patient comfort. In the proposed system, optimum thermal dose to which the carcinogenic tissue can be exposed is determined. State space representation of the thermal model of the tissue is used for simulation studies. The performance metrics are the transient response characteristics like rise time, settling time and mean square error. These characteristics are compared to assess the controller performance. The FLC and FALCON performance for both linear and exponential parametrization has been compared. It has been observed that implementation of Fuzzy Controller and FALCON for thermal dose is successful in minimizing the treatment time. Model validation has been done by observing the mean square error which is well within the permissible limits. The concept of Automatic thermal dose controller has been enhanced using hybrid Neuro-fuzzy Controller (FALCON). The developed controller is applicable for a wide range of thermal therapies like thermal ablation and thermal radiotherapy.

Bitcoin price forecasting with neuro-fuzzy techniques

Cryptocurrencies, with Bitcoin being the most notable example, have attracted considerable attention in recent years, and they have experienced large fluctuations in their price. While a few studies employ conventional statistical and econometric approaches to reveal the driving factors of Bitcoin's prices, research on the development of forecasting models to be used as decision support tools in investment strategies is scarce. This study proposes a computational intelligence technique that uses a hybrid Neuro-Fuzzy controller, namely PATSOS, to forecast the direction in the change of the daily price of Bitcoin. The proposed methodology outperforms two other computational intelligence models, the first being developed with a simpler neuro-fuzzy approach, and the second being developed with artificial neural networks. Furthermore, the investment returns achieved by a trading simulation, based on the signals of the proposed model, are 71.21% higher than the ones achieved through a naive buy-and-hold strategy. The performance of the PATSOS system is robust to the use of other cryptocurrencies.

PREDICTION OF PERFORMANCE FOR SAVONIUS-DARRIEUS WIND ROTOR BY HYBRID NEURO-FUZZY CONTROLLER

Vertical Axis Wind Turbine (VAWT) is a viable proposition for small-scale uses like, rural electrification, pumping, desalinating, household applications etc in many developing countries including India. In this paper, a hybrid neuro-fuzzy controller has been developed using gradient-based training algorithm to evaluate the performance of a combined three-bladed Savonius-Darrieus rotor. The objective of the study is to design a controller that causes more uniform loading on the generator by minimizing fluctuations in output parameters with change of input and also that improves rotor performance. A two-input-single-output controller has been designed. The tip speed ratio and overlap have been taken as input parameters, and output parameters are power coefficients and torque coefficients. At the first step, the input data are fuzzified by assigning fuzzy levels to the input data sets, and then trained outputs are obtained by back propagation learning algorithm. The controller results are in good agreement with the experimental results both qualitatively and quantitatively. For power coefficient (Cp), the agreement is within 4.5%, and for torque coefficient (Ct) it is within 2%. Moreover, the performance of the hybrid neuro-fuzzy controller has also been compared with Fuzzy Logic Controller (FLC) & ANN controller. The present controller predicts smoother values of performance parameters compared with other controllers.

Torque and Flux Ripple Minimization of Induction Motor Using Hybrid Neuro Fuzzy Controller

This paper proposed to design, implementation and simulation of Direct Torque Control of Induction Motor drive system is to minimize stator current distortion, electromagnetic torque and flux ripples. In this paper, Hybrid Neuro-Fuzzy Controller is proposed to replace the conventional PI torque and flux controller to achieve desired torque and flux with zero steady state error and also with good dynamic fast response. Neuro-Fuzzy based torque and flux controllers are designed to optimize the stator voltages in d-q reference frame that applied to Direct Torque Control. Simulation and the performance of the proposed Hybrid NFC are analyzed. Simulation results showed that a significant improvement in dynamic torque and speed response in both steady and transient states and also a considerable reduction in torque and flux minimization compared to the other AI techniques.

PREDICTION OF PERFORMANCE FOR SAVONIUSDARRIEUS WIND ROTOR BY HYBRID NEURO-FUZZY CONTROLLER

Vertical Axis Wind Turbine (VAWT) is a viable proposition for small-scale uses like, rural electrification, pumping, desalinating, household applications etc in many developing countries including India. In this paper, a hybrid neuro-fuzzy controller has been developed using gradient-based training algorithm to evaluate the performance of a combined three-bladed Savonius-Darrieus rotor. The objective of the study is to design a controller that causes more uniform loading on the generator by minimizing fluctuations in output parameters with change of input and also that improves rotor performance. A two-input-single-output controller has been designed. The tip speed ratio and overlap have been taken as input parameters, and output parameters are power coefficients and torque coefficients. At the first step, the input data are fuzzified by assigning fuzzy levels to the input data sets, and then trained outputs are obtained by back propagation learning algorithm. The controller results are in good agreement with the experimental results both qualitatively and quantitatively. For power coefficient (Cp), the agreement is within 4.5%, and for torque coefficient (Ct) it is within 2%. Moreover, the performance of the hybrid neuro-fuzzy controller has also been compared with Fuzzy Logic Controller (FLC) & ANN controller. The present controller predicts smoother values of performance parameters compared with other controllers.

Development of a hybrid fuel cell system operated simplified neuro-fuzzy sliding-mode control based IM drive deploying linearization approach: An effort to enhance the performance

This article focuses on the design and implementation of a hybrid fuel cell supply system to deliver the power demanded by the feedback linearization (FBL) based induction motor (IM) drive. The proposed modified simple hybrid neuro-fuzzy sliding-mode control (NFSMC) with intuitive FBL substantially reduces torque chattering and improves speed response, giving optimal drive performance under system uncertainties and disturbances. This novel technique also has the benefit of reduced computational burdens by improving computational efficiency over conventional NFSMC and thus suitable for real-time industrial applications. The parameters of the modified neuro-fuzzy control are tuned by an adaptive mechanism based on sliding-mode control. A fuel cell stack with a dc/dc boost converter is considered here as a separate external source during interruption of main supply for maintaining the supply to the motor drive control through the inverter, thereby reducing the burden and average rating of the inverter. A rechargeable battery used as an energy storage supplements the fuel cell during different operating conditions of the drive system. The effectiveness of the proposed method using the fuel cell-based linearized IM drive is investigated in simulation, and the efficacy of the proposed controller is validated by experiment using DSP2812. It is evident from the results that the system provides optimal dynamic performance in terms of remarkably reduced torque ripple, good load disturbance rejection, and reduced speed settling time responses by around 92%, 50%, and 41%, respectively, compared to the PI-controller. The system is also robust in terms of parameter variation and external load.

Using computational intelligence to forecast carbon prices

European Union has introduced the European Trading System (ETS) as a tool for developing and implementing international treaties related to climate changes and to identify the most cost-effective methods for reducing greenhouse gas emissions, in particular carbon dioxide (CO2), which is the most substantial. Companies producing carbon emissions must effectively manage associated costs by buying or selling carbon emission futures. Viewed from this perspective, this paper provides a model for managing the risk by buying and selling carbon emission futures by implementing techniques that leverage computational intelligence. Three computational intelligence techniques are proposed to provide accurate and timely forecasts for changes in the price of carbon: a novel hybrid neuro-fuzzy controller that forms a closed-loop feedback mechanism called PATSOS; an artificial neural network (ANN) based system; an adaptive neuro-fuzzy inference system (ANFIS). Results are based on 1074 daily carbon price observations collected to comprise a useful time-series dataset and for evaluation of the proposed techniques. The extra-sample performance of the proposed techniques is calculated. Analysis results are compared with those produced by other models. Comparison studies reveal that PATSOS is the most accurate and promising methodology for predicting the price of carbon. It is stated that this paper registers a first attempt to apply a hybrid neuro-fuzzy controller to forecasting carbon prices.

Hybrid Neuro Fuzzy Controller for Automatic Generation Control of Multi Area Deregulated Power System

This paper is intended in investigating the Automatic Generation Control (AGC) problem of a deregulated power system using Adaptive Neuro Fuzzy controller. Here, three area control structure of Hydro-Thermal generation has been considered for different contracted scenarios under diverse operating conditions with non-linearities such as Generation Rate Constraint (GRC) and Backlash. In each control area, the effects of the feasible contracts are treated as a set of new input signals in a modified traditional dynamical model. The key benefit of this strategy is its high insensitivity to large load changes and disturbances in the presence of plant parameter discrepancy and system nonlinearities. This newly developed scheme leads to a flexible controller with a simple structure that is easy to realize and consequently it can be constructive for the real world power system. The results of the proposed controller are evaluated with the Hybrid Particle Swarm Optimisation (HCPSO), Real Coded Genetic Algorithm (RCGA) and Artificial Neural Network (ANN) controllers to illustrate its robustness.

Four-Area Load Frequency Control of an Interconnected Power System Using Neuro-Fuzzy Hybrid Intelligent Proportional and Integral Control Approach

AbstractThis article presents a novel control approach, hybrid neuro-fuzzy (HNF), for the load frequency control (LFC) of a four-area interconnected power system. The advantage of this controller is that it can handle nonlinearities, and at the same time, it is faster than other existing controllers. The effectiveness of the proposed controller in increasing the damping of local and inter-area modes of oscillation is demonstrated in a four-area interconnected power system. Areas 1 and 2 consist of a thermal reheat power plant, whereas Areas 3 and 4 consist of a hydropower plant. Performance evaluation is carried out by using fuzzy, artificial neural network (ANN), adaptive neuro-fuzzy inference system, and conventional proportional and integral (PI) control approaches. Four different models with different controllers are developed and simulated, and performance evaluations are carried out with said controllers. The result shows that the intelligent HNF controller has improved dynamic response and is at the same time faster than ANN, fuzzy, and conventional PI controllers.

05/00377 Neuro-PID tracking control of a discharge air temperature system: Zaheer-uddin, M. and Tudoroiu, N. Energy Conversion and Management, 2004, 45, (15–16), 2405–2415

A hybrid neuro-fuzzy control strategy and its corresponding rule generating approach is proposed. According to this approach, the fuzzy control rules can be generated automatically via fuzzy inputs, and then the appropriate control action can be deduced efficiently by a simplified fuzzy inference engine. By combining the use of an incremental PI algorithm and a positional PD algorithm, a PID fuzzy control strategy can be implemented simply from two input variables. It results in the number of control rules being significantly reduced without decreasing the control performance. The control parameters can be self-tuned by introducing a single neuron together with a modified back-propagation learning algorithm. Simulation results show that the proposed fuzzy controller is able to control unknown processes and provide good performance. Compared to traditional self-organising and neural-network-based fuzzy controllers, this method has simpler control algorithms and less computational burden.

Intelligent Control of Continuous Process Using Neuro-Fuzzy System

The main goal of the paper presented is to outline the possibility of hybrid neuro-fuzzy control systems application for continuous process intelligent control focusing on heating technology. The principle used pointed out the importance of a structured form of the operator's knowledge; further the realization of fuzzy decisions from the supervisory control level as well as the tuning of modified neuro- fuzzy system NEFCON with parametrical learning. The practical use of such approach in the domain of building complexes heating system control is described.Software realization of the neuro-fuzzy control algorithms for the process optimization on heuristical base used as well as the communication between supervisory SCADA system and the process operator represent an open applied software package making possible its integration to complex monitoring and control system of so called intelligent building. Results attained show possibilities to gain energy savings, thermal comfort improvement and pollution reduction as well.

A hybrid neuro-fuzzy PID controller

A hybrid neuro-fuzzy control strategy and its corresponding rule generating approach is proposed. According to this approach, the fuzzy control rules can be generated automatically via fuzzy inputs, and then the appropriate control action can be deduced efficiently by a simplified fuzzy inference engine. By combining the use of an incremental PI algorithm and a positional PD algorithm, a PID fuzzy control strategy can be implemented simply from two input variables. It results in the number of control rules being significantly reduced without decreasing the control performance. The control parameters can be self-tuned by introducing a single neuron together with a modified back-propagation learning algorithm. Simulation results show that the proposed fuzzy controller is able to control unknown processes and provide good performance. Compared to traditional self-organising and neural-network-based fuzzy controllers, this method has simpler control algorithms and less computational burden.

Toward learning individual characteristics in a hybrid neuro-fuzzy wheelchair controller

We have designed and built a fuzzy wheelchair controller, which can input the vague and shakey inputs of severe hand tremor and output a smooth control signal. The system has a new contribution in that it attempts to model the fuzziness and uncertainty of the control action of the user. We report the results of the project so far, describe continuing experiments, and propose future work to learn individual characteristics of wheelchair users. The significance of this work is that it explores the application of fuzzy logic to vague or imprecise human input. While this is breaking new ground in the application of fuzzy logic, our prototype demonstrates that this new type of application is not only possible, but also efficient. © 1998 John Wiley & Sons, Inc.

A method for design of a hybrid neuro-fuzzy control system based on behavior modeling

It is known that control signals from a fuzzy logic controller are determined by a response behavior of a controlled object rather than its analytical models. That implies that the fuzzy controller could yield a similar control result for a set of plants with a similar dynamic behavior. This idea lends to modeling of a plant with unknown structure by defining several types of dynamic behaviors. On the basis of dynamic behavior classification, a new method is presented for the design of a neuro-fuzzy control system in two steps: 1) we model a plant with unknown structure by choosing a set of simplified systems with equivalent behavior as "templates" to optimize their fuzzy controllers off-line; and 2) we use an algorithm for system identification to perceive dynamic behavior and a neural network to adapt fuzzy logic controllers by matching the "templates" online. The main advantage of this method is that convergence problem can be avoided during adaptation process. Finally, the proposed method is used to design neuro-fuzzy controllers for a two-link manipulator.