• Journal of Sensor Science and Technology
• /
• v.21 no.5
• /
• pp.329-338
• /
• 2012

This paper presents an effective design method for a gas identification system. The design method adopted the sequential combination between the hybrid genetic algorithms and the TSK fuzzy logic system. First, the sensor grouping method by hybrid genetic algorithms led the effective dimensional reduction as well as effective pattern analysis from a large volume of pattern dimensions. Second, the fuzzy identification sub-models allowed handling the uncertainty of the sensor data extensively. By these advantages, the proposed identification model demonstrated high accuracy rates for identifying the five different types of gases; it was confirmed throughout the experimental trials.

• Proceedings of the KIEE Conference
• /
• 2008.07a
• /
• pp.1849-1850
• /
• 2008

Type-2 퍼지 집합은 Type-1 퍼지 집합의 확장으로 Type-1 퍼지 집합으로는 다루기 힘든 언어적인 불확실성을 다루기 위해 고안되었다. 대표적인 퍼지 논리 시스템(Fuzzy Logic System; FLS)으론 Mamdani FLS 모델과 TSK FLS모델이 있다. 본 논문에서는 Interval Type-2 TSK FLS를 구성한다. FLS 구성을 위한 전반부는 가우시안 형태의 Type-2 멤버쉽 함수를 사용하며, 전.후반부 파라미터들은 오류역전파 알고리즘을 통한 학습으로 결정한다. 본 논문에서는 Type-1 TSK FLS와 Interval Type-2 TSK FLS를 설계하고 가스로 공정 데이터에 적용하여 성능을 비교 분석한다. 또한 노이즈를 추가한 데이터들을 통하여 노이즈에 대한 성능도 비교 분석한다.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.62 no.11
• /
• pp.1590-1597
• /
• 2013

Electric power is an important part in economic development. Moreover, an accurate load forecast can make a financing planning, power supply strategy and market research planned effectively. This paper used the fuzzy logic system to predict the regional electric power load. To design the fuzzy prediction system, the correlation-based clustering algorithm and TSK fuzzy model were used. Also, to improve the prediction system's capability, the moving average technique and relative increasing rate were used in the preprocessing procedure. Finally, using four regional electric power load in Taiwan, this paper verified the performance of the proposed system and demonstrated its effectiveness and usefulness.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2001.05a
• /
• pp.4-7
• /
• 2001

A stable TSK-type FLC can be designed by the method of Parallel Distributed Compensation (PDC), but in this case, solving the LMI problem is not a trivial task. To overcome such a difficulty, a Time-Delay based FLC (TDFLC) is proposed. TSK-type TDFLC consists of Time-Delay Control (TDC) and Sliding Mode Control (SMC) schemes, which result in a robust controller basaed upon an integral sliding surface.

• Proceedings of the KIEE Conference
• /
• 2008.04a
• /
• pp.57-58
• /
• 2008

Type-2 퍼지 집합은 언어의 불확실성을 다루기 위하여 고안된 Type-1 퍼지집합의 확장이다. TSK 퍼지 로직 시스템(TSK Fuzzy Logic Systems; TSK FLS)은 Mamdani FLS과 함께 가장 널리 사용되는 퍼지 로직 시스템 모델이다. 본 논문에서는 Type-2 퍼지 집합을 이용하여 전반부 멤버쉽 함수를 구성하고 후반부 다항식 함수를 상수와 1차식, 2차식으로 확장한 다항식 Type-2 TSK FLS 설계한다. 다항식 Type-2 TSK FLS의 파라미터를 동정하기 위해 Back-propagation 방법을 사용한다. 제안된 다항식 Type-2 TSK FLS을 노이즈 섞인 비선형 시스템의 모델링에 적용하여 그 성능을 비교 분석한다.

• Proceedings of the KIEE Conference
• /
• 2008.10b
• /
• pp.251-252
• /
• 2008

Type-2 퍼지 집합은 Type-1 퍼지 집합에서는 다루기 어려운 언어적인 불확실성을 더욱 효과적으로 다룰 수 있다. TSK 퍼지 로직 시스템(TSK Fuzzy Logic Systems; TSK FLS)은 후반부를 1차 및 2차 함수식으로 나타내며 Mamdani 모델과 함께 가장 널리 사용되는 모델이다. 본 연구의 Interval Type-2 TSK FLS은 전반부에서 Type-2 퍼지 집합을 이용하고 후반부는 계수가 Type-1 퍼지집합인 1차식을 사용한다. 또한 전반부는 가우시안 형태의 Type-2 멤버쉽 함수를 사용하며, 오류역전파 학습알고리즘을 사용하여 파라미터들을 최적화 한다. 또한 학습에 앞서 PSO(Particle Swarm Optimization) 알고리즘을 사용하여 최적 학습률을 찾아 모델의 학습능력을 보다 효율적으로 한다. 본 논문에서는 Type-1과 Type-2 FLS의 성능을 가스로 공정 데이터를 적용하여 두 모델의 성능을 비교하고 노이즈를 추가한 데이터를 이용하여 노이즈에 대한 성능도 비교 분석한다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.5 no.2
• /
• pp.95-101
• /
• 2005

In this paper, we proposed a learning algorithm for the neuro-fuzzy modeling using a learning rule to adapt clustering. The proposed algorithm includes the data partition, assigning the rule into the process of partition, and optimizing the parameters using predetermined threshold value in self-constructing algorithm. In order to improve the clustering, the learning method of neuro-fuzzy model is extended and the learning scheme has been modified such that the learning of overall model is extended based on the error-derivative learning. The effect of the proposed method is presented using simulation compare with previous ones.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.6 no.1
• /
• pp.52-57
• /
• 2006

In this paper, we have proposed a new adaptive fuzzy control algorithm based on Takagi-Sugeno fuzzy model. The regulation problem for the uncertain SISO nonlinear system is solved by the proposed algorithm. Using the advanced stability theory, the stability of the state, the control gain and the parameter approximation error is proved. Unlike the existing feedback linearization based methods, the proposed algorithm can guarantee the global stability in the presence of the singularity in the inverse dynamics of the plant. The performance of the proposed algorithm is demonstrated through the problem of balancing and swing-up of an inverted pendulum on a cart.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.4 no.2
• /
• pp.187-192
• /
• 2004

In this paper, we proposed the method to design fuzzy controller using the experience of the operating expert and experimental numeric data for the robust control about the noise and disturbance instead of the traditional PID controller for the main steam temperature control of the thermal power plant. The temperature of main steam temperature process has to be controlled uniformly for the stable electric power output. The process has the problem of the hunting for the cases of various disturbances. In that case, the manual action of the operator happened to be introduced in some cases. We adopted the TSK (Takagi-Sugeno-Kang) model as the fuzzy controller and designed the fuzzy rules using the informations extracted directly from the real plant and various operating condition to solve the above problems and to apply practically. We implemented the real fuzzy controller as the Function Block module in the DCS(Distributed Control System) and evaluated the feasibility through the experimental results of the simulation.

• Journal of Industrial Technology
• /
• v.31 no.B
• /
• pp.45-50
• /
• 2011

Recently, machine olfactory systems that have been proposed as an artificial substitute of the human olfactory system are being studied by many researchers because they can scent dangerous gases and identify the type of gases in contamination areas instead of the human. In this paper, we present an effective design method for the gas identification system. The design method adopted the sequential combination between genetic algorithms and TSK fuzzy logic system. First, the proposed method allowed the designed gas identification system effectively performing the pattern analysis because it was able to avoid the curse of dimensionality caused by use of a large number of sensors. Secondly, the method led the gas identification system to good performance because it was able to deal with drift characteristics of the sensor data by using description ability of the fuzzy system for nonlinear data. In simulation, we demonstrated the effectiveness of the designed gas identification system by using the simulation results of five types of gases.