• Proceedings of the IEEK Conference
• /
• 2002.07c
• /
• pp.1626-1629
• /
• 2002

This paper describes novel methods to construct fuzzy inference rules with gradient descent. The present methods have a constructive mechanism of the rule unit that is applicable in two parameters: the central value and the width of the membership function in the antecedent part. The first approach is to create the rule unit at the nearest position from the input space, for the central value of the membership function in the antecedent part. The second is to create the rule unit which has the minimum width, for the width of the membership function in the antecedent part. Experimental results are presented in order to show that the proposed methods are effective in difference on the inference error and the number of learning iterations.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.36 no.1
• /
• pp.54-59
• /
• 2000

This paper presents a control system that uses a fuzzy algorithm in controlling the depth of a model midwater trawl net, and experimental results carried out in the circulating water channel by using a model trawl winch system.The fuzzy controller calculates the length of the warp to be changed, based on the depth error between the desired depth and actual depth of the model trawl net and the ratio of change in the depth error. The error and the error change are calculated every sampling time. Then the control input, i.e. desirable length of the warp, is determined by inference from the linguistic control rules which an experienced captain or navigator uses in controlling the depth of the trawl winch controller and the length of the warp is changed. Two kinds of fuzzy control rules were tested, one was obtained from the actual operations used by a skilled skipper or navigator, and the other was a modified from the former by considering the hydrodynamic characteristics of the model trawl system.Two kinds of fuzzy control were tested, one was obtained fro the actual operations used by a skilled skipper or navigator, and the other was a modified from the former by considering the hydrodynamic characteristics of the model trawl system.The results of these model experiments indicate that the proposed fuzzy controllers rapidly follow the desired depth without steady-state error although the desired depth was given in one step, and show robustness properties against changes in the parameters such as the change of the towing sped. Especially, a modified rule shows smaller depth fluctuations and faster setting times than those obtained by a field oriented rule.

• Journal of Broadcast Engineering
• /
• v.5 no.1
• /
• pp.123-129
• /
• 2000

This paper addresses a new double detection algorithm which is based on the fuzzy control in the adaptive echo canceller of communication system. In this method, the two input of the fuzzy inference for detecting double talk condition are used. The one is the cross-correlation coefficient between the error signal and the primary signal which is the summed signal of the real echo signal and the near-end signal. The other is the cross-correlation coefficient between the estimation error signal and the primary signal. The fuzzy controller made a fuzzification for two inputs by the membership functions of trapezoid and them became the composition using inference rules. The composed result is defuzzificated by the center gravity method. The output is compared with two threshold values to detect double talk and echo path variation effectively. It is confirmed by computer simulation that this fuzzy double talk detector is able to track echo path variation accurately.

• Journal of Institute of Control, Robotics and Systems
• /
• v.19 no.9
• /
• pp.839-845
• /
• 2013

A CSTR (Continuous Stirred Tank Reactor) is a highly nonlinear process with varying parameters during operation. Therefore, tuning of the controller and determining the transition policy of controller parameters are required to guarantee the best performance of the CSTR for overall operating regions. In this paper, a methodology employing the 2DOF (Two-Degree-of-Freedom) PID controller, the anti-windup technique and a fuzzy gain scheduler is presented for the temperature control of the CSTR. First, both a local model and an EA (Evolutionary Algorithm) are used to tune the optimal controller parameters at each operating region by minimizing the IAE (Integral of Absolute Error). Then, a set of controller parameters are expressed as functions of the gain scheduling variable. Those functions are implemented using a set of "if-then" fuzzy rules, which is of Sugeno's form. Simulation works for reference tracking, disturbance rejecting and noise rejecting performances show the feasibility of using the proposed method.

• Journal of Institute of Control, Robotics and Systems
• /
• v.16 no.12
• /
• pp.1150-1158
• /
• 2010

In this study, the Polynomial-based Radial Basis Function Neural Networks is proposed as one of the recognition part of overall face recognition system that consists of two parts such as the preprocessing part and recognition part. The design methodology and procedure of the proposed pRBFNNs are presented to obtain the solution to high-dimensional pattern recognition problem. First, in preprocessing part, we use a CCD camera to obtain a picture frame in real-time. By using histogram equalization method, we can partially enhance the distorted image influenced by natural as well as artificial illumination. We use an AdaBoost algorithm proposed by Viola and Jones, which is exploited for the detection of facial image area between face and non-facial image area. As the feature extraction algorithm, PCA method is used. In this study, the PCA method, which is a feature extraction algorithm, is used to carry out the dimension reduction of facial image area formed by high-dimensional information. Secondly, we use pRBFNNs to identify the ID by recognizing unique pattern of each person. The proposed pRBFNNs architecture consists of three functional modules such as the condition part, the conclusion part, and the inference part as fuzzy rules formed in 'If-then' format. In the condition part of fuzzy rules, input space is partitioned with Fuzzy C-Means clustering. In the conclusion part of rules, the connection weight of pRBFNNs is represented as three kinds of polynomials such as constant, linear, and quadratic. Coefficients of connection weight identified with back-propagation using gradient descent method. The output of pRBFNNs model is obtained by fuzzy inference method in the inference part of fuzzy rules. The essential design parameters (including learning rate, momentum coefficient and fuzzification coefficient) of the networks are optimized by means of the Particle Swarm Optimization. The proposed pRBFNNs are applied to real-time face recognition system and then demonstrated from the viewpoint of output performance and recognition rate.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1998.10a
• /
• pp.191-196
• /
• 1998

A new Fuzzy Logic Controller(FLC) called a Gradient-Descent Based Self-Organizing Controller is presented. The Self-Organizing Controller(SOC) has two inputs such as error and change of error, and updates control rules with monitoring a performance measure. There are many works in the SOC which concentrate on the self-organizing ability in control rule base, but have a few research on the performance measure which is akin to sliding mode control. With this procedure, we can get a robust performance measure on the SOC. To verify the perfomance of proposed controller, we have performed for the cart-pole system which is one of the well-known benchmark problem in the control literature.

• International Journal of Control, Automation, and Systems
• /
• v.3 no.2
• /
• pp.183-194
• /
• 2005

In this study, we introduce an advanced architecture of genetically optimized Hybrid Fuzzy Neural Networks (gHFNN) and develop a comprehensive design methodology supporting their construction. A series of numeric experiments is included to illustrate the performance of the networks. The construction of gHFNN exploits fundamental technologies of Computational Intelligence (CI), namely fuzzy sets, neural networks, and genetic algorithms (GAs). The architecture of the gHFNNs results from a synergistic usage of the genetic optimization-driven hybrid system generated by combining Fuzzy Neural Networks (FNN) with Polynomial Neural Networks (PNN). In this tandem, a FNN supports the formation of the premise part of the rule-based structure of the gHFNN. The consequence part of the gHFNN is designed using PNNs. We distinguish between two types of the linear fuzzy inference rule-based FNN structures showing how this taxonomy depends upon the type of a fuzzy partition of input variables. As to the consequence part of the gHFNN, the development of the PNN dwells on two general optimization mechanisms: the structural optimization is realized via GAs whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the gHFNN, the models are experimented with a representative numerical example. A comparative analysis demonstrates that the proposed gHFNN come with higher accuracy as well as superb predictive capabilities when comparing with other neurofuzzy models.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.1
• /
• pp.102-109
• /
• 2014

In this paper, an algorithm to design fuzzy PID controllers is proposed. The proposed controllers are composed of fuzzy rules of which consequences are linear PID controllers and are designed with help of TSK fuzzy controllers. TSK fuzzy controllers are designed from TSK fuzzy model using pole assignment and have outstanding ability making the output response of nonlinear systems similar to the desired one. However, because of its structure complexity the TSK fuzzy controller is difficult to be used in industry. The proposed controllers have PID controller structure which can be easily realized, and are designed by using the data obtained from control simulations with TSK fuzzy controllers. To verify the proposed algorithm, two example simulations are performed.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1315-1318
• /
• 2004

Based on fuzzy 2-mean classification and template matching method, we propose a new algorithm to detect the edges of an image. In the algorithm, fuzzy 2-mean classification can classify all pixels in the mask into two clusters whatever the mask in the dark or light region; and template matching not only determines the edge's direction, but also thins the detected edge by a set of inference rules and, by the way, reduces the impulse noises.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.7 no.5
• /
• pp.14-20
• /
• 1997

This paper proposes a new design method for the SMC-type FLC and shows that a SMC-type LFC is an extension of the SMC with BL. The conventional SMC-type FLC uses error and change-of-error as inputs of the FLC and generates the absolute value of a switching magnitude. Then, the fuzzy rule table is constructed on a two-dimensional space of the phase plane and has commonly the skew symmetric property. In this paper, we introduce a new variable, signed distance, from the skew symmetric property of the rule table. And thd variable becomes only a fuzzy variable that is used to generate the control input of a SMC-type FLC. that is, we design a new SMC-type FLC that uses a signed distance and a control input as the variables representing the contents of the rule-antecedent and the rule-con-sequent, respectively. Then the number of total rules is reduced and the control performance is almost the same as that of the conventional SMC-type FLC. Additionally, we derive the control law of the ordinary SMC with BL from a new SMC-type FLC. Namely, we show that a FLC is an extension of the SMC with BL.