Jia‐Shing Sheu^a, I‐Chen Chen^b, Yi‐Syong Liao^a

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 395-404, 2019, DOI:10.31209/2019.100000101

Abstract This study proposed a household energy state monitoring system (HESMS) and a household energy load monitoring system (HELMS) for monitoring smart appliances. The HESMS applies reinforcement learning to receive changes in the external environment and the state of an electrical appliance, determines if the electrical appliance should be turned on, and controls the signals sent to the HELMS according to these decisions. The HELMS implements an ON/OFF control mechanism for household appliances according to the control signals and the power consumption state. The proposed systems are based on the wireless communication network and can monitor household appliances’ energy usage, control… More >

Peizhong Liu¹, Xiaofang Liu¹, Yanming Luo², Yongzhao Du¹, Yulin Fan¹, Hsuan‐Ming Feng³

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 385-394, 2019, DOI:10.31209/2019.100000100

Abstract Aiming at the drawback of artificial bee colony algorithm (ABC) with slow convergence speed and weak exploitation capacity, an enhanced exploitation artificial bee colony algorithm is proposed, EeABC for short. Firstly, a generalized opposition-based learning strategy (GOBL) is employed when initial population is produced for obtaining an evenly distributed population. Subsequently, inspired by the differential evolution (DE), two new search equations are proposed, where the one is guided by the best individuals in the next generation to strengthen exploitation and the other is to avoid premature convergence. Meanwhile, the distinction between the employed bee and the onlooker bee is not… More >

Ching‐Han Chen^a, Ching‐Yi Chen^b, Nai‐Yuan Liu^a

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 375-384, 2019, DOI:10.31209/2019.100000099

Abstract This study introduces a dynamic gesture recognition system applicable in IPTV remote control. In this system, we developed a hardware accelerator for realtime moving object detection. It is able to detect the position of hand block in each frame at high speed. After acquiring the information of hand block, the system can capture the robust dynamic gesture feature with the moving trail of hand block in the continuous images, and input to FNN classifier for starting recognition process. The experimental results show that our method has a good recognition performance, and more applicable to real gesture-controlled human-computer interactive environment. More >

Nan Pan^1*, Dilin Pan², Yi Liu², Gang Li³

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 367-373, 2019, DOI:10.31209/2019.100000098

Abstract A set of laser detection system for shearing tools is developed, By holding breakage of the cable, firstly, using single-point laser displacement sensors to pick up surface features signal of line trace, then wavelet decomposition is used to reduce the noise, and the signal after noise reduction is obtained. After that, the threshold based sequence comparison method is used to achieve matches of similar coincidence for trace features, and then using a gradient descent method to getting the minimum cost of cost function value through continuous iterative, and finally realizing the fast traceability of corresponding shearing tool. More >

Nan Pan¹, Yi Liu², Dilin Pan², Junbing Qian¹, Gang Li³

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 359-366, 2019, DOI:10.31209/2019.100000097

Abstract It will face a lot of problems when using existing image-processing and 3D scanning methods to do the similarity analysis of the line traces, therefore, an effective comparison algorithm is put forward for the purpose of making effective trace analysis and infer the criminal tools. The proposed algorithm applies wavelet decomposition to the line trace 1-D detection signals to partially reduce background noises. After that, the sequence comparison strategy based on wavelet domain DTW is employed to do trace feature similarity matching. Finally, using linear regression machine learning algorithm based on gradient descent method to do constant iteration. The experiment… More >

Jihyuck Jo^a, Han‐Gyu Kim^b, In‐Cheol Park^a, Bang Chul Jung^c, Hoyoung Yoo^c

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 351-358, 2019, DOI:10.31209/2019.100000096

Abstract A modified Viterbi scoring procedure is presented in this paper based on Dijkstra’s shortest-path algorithm. In HMM-based speech recognition systems, the Viterbi scoring plays a significant role in finding the best matching model, but its computational complexity is linearly proportional to the number of reference models and their states. Therefore, the complexity is serious in implementing a high-speed speech recognition system. In the proposed method, the Viterbi scoring is translated into the searching of a minimum path, and the shortest-path algorithm is exploited to decrease the computational complexity while preventing the recognition accuracy from deteriorating. In addition, a two-phase comparison… More >

Byoung-Doo Oh^a,b, Hye-Jeong Song^a,b, Jong-Dae Kim^a,b, Chan-Young Park^a,b, Yu-Seop Kim^a,b

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 343-350, 2019, DOI:10.31209/2019.100000095

Abstract Accurate prediction of fine dust (PM10) concentration is currently recognized as an important problem in East Asia. In this paper, we try to predict the concentration of PM10 using Deep Neural Network (DNN). Meteorological factors, yellow dust (sand), fog, and PM10 are used as input data. We test two cases. The first case predicts the concentration of PM10 on the next day using the day’s weather forecast data. The second case predicts the concentration of PM10 on the next day using the previous day’s data. Based on this, we compare the various performance results from the DNN model. In the… More >

Chung Wen Hung, Wei Lung Mao, Han Yi Huang

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 329-341, 2019, DOI:10.31209/2019.100000093

Abstract Nonlinear system modeling and identification is the one of the most important areas in engineering problem. The paper presents the recurrent fuzzy neural network (RFNN) trained by modified particle swarm optimization (MPSO) methods for identifying the dynamic systems and chaotic observation prediction. The proposed MPSO algorithms mainly modify the calculation formulas of inertia weights. Two MPSOs, namely linear decreasing particle swarm optimization (LDPSO) and adaptive particle swarm optimization (APSO) are developed to enhance the convergence behavior in learning process. The RFNN uses MPSO based method to tune the parameters of the membership functions, and it uses gradient descent (GD) based… More >

Ching‐Yi Chen, Yi‐Jen Lin

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 319-327, 2019, DOI:10.31209/2019.100000093

Abstract In this study, a systematic data-driven adaptive neuro-fuzzy inference system (ANFIS) modelling methodology is proposed. The new methodology employs an unsupervised competitive learning scheme to build an initial ANFIS structure from input-output data, and a high-performance PSO-LSE method is developed to improve the structure and to identify the consequent parameters of ANFIS model. This proposed modelling approach is evaluated using several nonlinear systems and is shown to outperform other modelling approaches. The experimental results demonstrate that our proposed approach is able to find the most suitable architecture with better results compared with other methods from the literature. More >

Wen-Hsiang Hsieh, Jerzy W Rozenblit, Minvydas Ragulskis

Intelligent Automation & Soft Computing, Vol.25, No.2, pp. 313-314, 2019, DOI:10.31209/2019.100000092

Abstract This article has no abstract. More >