Search

Wind is one of the most essential sources of clean, renewable energy, and is, therefore, a critical element in responsible power consumption and production. The accurate prediction of wind speed plays a key role in decision-making and management in wind power generation. This study proposes a model using a deep belief network with genetic algorithms (DBNGA) for wind speed forecasting. The genetic algorithms are used to determine parameters for deep belief networks. Wind speed and weather-related data are collected from Taiwan’s central weather bureau for this purpose. This paper uses both time series data and multivariate regression data to forecast wind speed. The seasonal autoregressive integrated moving average (SARIMA) method and the least squares support vector regression for time series with genetic algorithms (LSSVRTSGA) are used to forecast wind speed in a time series, and the least squares support vector regression with genetic algorithms (LSSVRGA) and DBNGA models are used to predict wind speed in a multivariate format. The empirical results show that forecasting wind speed by the DBNGA models outperforms the other forecasting models in terms of forecasting accuracy. Thus, the DBNGA model is a feasible and effective approach for wind speed forecasting.

This study applies an intuitionistic type-2 fuzzy inference system (FIS) to e-learning system. This study utilizes the advantages of intuitionistic fuzzy set and type-2 fuzzy inference system, which have been proven to be most effective in uncertain environment, to recommend e-learning courses. The learning readiness is difficultly found in e-leaning company. In the study, triangular fuzzy numbers (TFS) are used to represent uncertain data. Intuitionistic type-2 is employed to model vaguely defined relations between TFS. The numerical example is used to verify the intuitionistic type-2 can effectively infer the learning readiness, and recommend more objective e-learning courses than traditional fuzzy inference systems.

This paper develops a new the weakest t-norm (T ω ) fuzzy Decision Making Trial and Evaluation Laboratory (DEMATEL) to evaluate green supply chain management (GSCM). Traditional DEMATEL is a popular method for making decision. In order to apply a realistic environment the fuzzy DEMATEL had to be developed for analyzing expert judgments with preferences, which are often unclear and ambiguous. The fuzzy DEMATEL usually adopts α-cut arithmetic operations to calculate results. However, general fuzzy DEMATEL model has not really handled fuzzy cause and effect relationships due to the accumulating phenomenon of fuzziness of the a-cut arithmetic. In the numerical example, using novel T ω fuzzy DEMATEL find out real cause and effect relationships with fuzzy bounds among criteria of GSCM. The T ω fuzzy DEMATEL can provide more credible information/results and analysis of across quadrants to evaluate cause and effected relationships, because the decision-maker usually wants to accurately grasp uncertain influential factors in the GSCM.

Recent several researches have developed similarity measures method for handling pattern recognition problems based on intuitionistic fuzzy sets. However, in some cases, those proposed method can not give sufficient information to discriminate from some patterns. Therefore, the purpose of this paper is to present a new similarity measures model based on intuitionistic fuzzy sets. This proposed method is provided to enhance measure accuracy degree of similarity between basic pattern and each sample, so as to give additional information for decision makers. A comparison between previous existing methods and our proposed method has been discussed. In addition, an application for radar target identification has been illustrated through using the proposed method.

This study develops the kernel intuitionistic fuzzy c-means clustering (KIFCM), and applies KIFCM in E-learning customer analysis. KIFCM combines intuitionistic fuzzy sets (IFSs) with kernel fuzzy c-means clustering (KFCM). The KIFCM has advantages of IFSs and KFCM which can effectively handle uncertain data and simultaneously map data to kernel space. The proposed KFCM has better performance than k-mean (KM) and fuzzy c-means (FCM) in numerical example. Furthermore, the study adopts the advanced clustering technology in E-learning customer clustering analysis, and analyses customer data based on clustering results by correlation analysis. The customer analysis result can provide for sales department, and assist to obtain customer's learning tendency in E-learning platform.

This study will apply specific operations data of a real logistics firm to analyze Inventory-Routing problems. We first develop a mathematical model for the problem, and then apply a Savings method for the initial feasible solution, and finally implement Genetic Algorithm to obtain the optimal solution. The numerical examples are given to acquire the optimal total inventory routing cost, the optimal delivery routes, the economic order quantities, the optimal service levels, and the reorder points for the retail stores in these proposed routes. The results show that this model can be successfully applied in Inventory-Routing problem in a very efficient way. The performance of this proposed approach is also compared with Savings method and the current routing planning implemented by this specific logistics company. Consequently, the outcome of this study is very promising in many aspects and should be helpful to this logistics company for its future planning of the distribution system.

Long-term business cycle forecasting is a very important issue in economic evaluation. This study presents a novel intuitionistic fuzzy least-squares support vector regression (IFLS-SVR) model for accurately forecasting long-term index of business cycle. Traditional support vector regression (SVR) and least-squares support vector regressions (LS-SVR) have been successfully applied in forecasting problems especially complex/nonlinear system. In this paper, the prediction model adopts two least-squares support vector regressions with intuitionistic fuzzy sets to approach fuzzy upper and lower bounds respectively, and then approach the crisp predict values. Genetic algorithms (GAs) are also employed in order to select two parameters of IFLS-SVR models. In this study, IFLS-SVR, LS-SVR and SVR are employed for the Taiwan business index forecasting. Empirical results indicate that the IFLS-SVR has better performance in terms of forecasting accuracy. Therefore, the IFLS-SVR model can efficiently provide credible long-term prediction for the Taiwan business index forecasting

The fuzzy weighted average (FWA), which is a function of fuzzy numbers and is useful as an aggregation method in management and engineering science based on fuzzy sets theory by Zadeh. It provides a discrete approximate solution by a-cuts level representation of fuzzy sets and interval analysis. Since the FWA method has an exponential complexity, thus several researches have focused on reducing this complexity. This paper also presents an enhanced fuzzy weighted average approach to achieve the objective of reducing the complexity. This proposed approach is through an improved initial solution for original FWA algorithm, and a two-phase concept by extending and applying both the algorithms of Chang et al. (2006) and Guu (2002). Its complexity is O(n) the same as Guu (2002) which is the best level achieved to date. This paper a practical example for unmanned aerial vehicle (UAV) selected under military requirements, which have illustrated and demonstrated the usefulness of this study.

Thin-Film Transistor Liquid-Crystal Display (TFT-LCD) manufacturing in Taiwan is booming; and the revenues from the TFT-LCD industry have grown significantly in recent years. One of the main problems in the TFT-LCD manufacturing process is to diagnose faulty products. This study employed support vector machines (SVM) with principal components analysis (PCA) to diagnose root causes in sputtering operations of the TFT-LCD industry. The PCA technique was used to transfer original manufacturing parameters into essential factors; and the SVM model was applied in classifying faulty products. Besides, the genetic and tabu (GA/TS) search algorithms were utilized to select SVM parameters. In terms of classification accuracy and efficiency, simulation results indicated that the SVM with PCA procedure is a feasible and promising way to diagnose faulty products in TFT-LCD manufacturing processes.

Fuzzy weighted average, as functions of fuzzy numbers, happens with the problem of multiple occurrences of fuzzy parameters (weights). Additional fuzziness may be introduced in the cut arithmetic. This paper based on an improved algorithm, called IFWA, applies orthogonal arrays method that is a fractional factorial experiment, to analysis weights adjustment. We will show a numerical example and demonstrate the usefulness of the orthogonal arrays method.

This paper develops a vendor managed inventory (VMI) dynamics based on the fuzzy arithmetic. The VMI has been used to reduce the "Bullwhip Effect" in supply chain. This research observes and analyzes the phenomena of VMI by system dynamics based on fuzzy arithmetic of the weakest t-norm (TW). The traditional crisp system dynamics has been widely used. However, it is observable that some variables/parameters may belong to the uncertain factors. Therefore, traditional crisp system dynamics may be necessary to be extended to treat the vague variables or parameters. In this paper the research examines the fuzzy arithmetic of TW in VMI and compares the performance of the traditional crisp system dynamics and system dynamics with fuzzy TW arithmetic. The system dynamics of fuzzy arithmetic can provide the extended information regarding the system behavior uncertainties for the decision-makers.

In this paper a novel way to construct pairwise alignment of protein sequence is proposed. Currently in protein sequence alignment the vital problem is having too many uncertain factors and causes significant data loss while using crisp data. For this important premise, fuzzy concept is introduced and fuzziness is implemented in the matrix for 250 point accepted mutations (PAMs) in sequence aligning, and integrated with the genetic algorithm (GA). The purpose for this implementation is to reduce the effects of uncertain factor, avoid making use of crisp values or weights resulting in significant data loss, and increase solution accuracy and method suitability. Experimental results of fuzzy matrix for 250 PAM can find more continuous and identical protein sequence after sequence alignment by GA.