Hirosato Seki

SIRMs connected fuzzy inference method adopting emphasis and suppression

The single input rule modules connected fuzzy inference method (SIRMs method) can decrease the number of fuzzy rules drastically in comparison with the conventional fuzzy inference methods. However, the inference results obtained by the SIRMs method is generally simple compared with those of the conventional fuzzy inference methods. For example, the SIRMs method may be not equivalent to the product-sum-gravity method and fuzzy singleton-type inference method, if the fuzzy sets of the antecedent parts are limited to normal fuzzy sets. In this paper, we propose a fuzzy singleton-type SIRMs method, which weights the rules of the SIRMs method, in order to solve the above problem. This paper also clarifies the property of the fuzzy singleton-type SIRMs method, from the view point of equivalence and monotonicity. Moreover, the fuzzy singleton-type SIRMs method is shown to be superior to the conventional SIRMs method by applying to a medical diagnosis system.

Nonlinear Identification Using Single Input Connected Fuzzy Inference Model

Abstract The single input connected fuzzy inference model (SIC model) by Hayashi et al. can decrease the number of fuzzy rules drastically in comparison with the conventional fuzzy inference models. In this paper, we first show the SIC model and its learning algorithm, and clarify the applicability of the SIC model by applying it to identification of nonlinear functions.

On the monotonicity of fuzzy inference models

Monotonicity property is very important in real systems. The monotonicity may need to be satisfied in a variety of application domains, e.g., control, medical diagnosis, educational evaluation, etc. A search in the literature reveals that the importance of the monotonicity in fuzzy inference system has been highlighted. Therefore, this paper surveys the works relating the monotonicity for various fuzzy inference systems. It firstly focuses on the monotonicity of the Mamdani inference model. Themonotonicity ofMamdani model is shown by using a defuzzification method in cases of three t-norms. Secondly, the monotonicity conditions and applications of the T–S inference model are stated. Finally, the monotonicity of the single input type fuzzy inference models is surveyed.

Fuzzy singleton-type SIC fuzzy inference model

This paper firstly proposes a fuzzy singleton-type Single Input Connected fuzzy inference model (fuzzy singleton-type SIC model) which attaches weights to the rules of the conventional SIC model. Second, it shows the property of the proposed model from the point of view of the equivalence. Thirdly, the learning algorithm of the fuzzy singleton-type SIC model is derived by using steepest descent method. Finally, the proposed model is applied to medical diagnosis, and compared with the conventional fuzzy inference model. From the above results, the applicability of the proposed model is clarified.

Type-2 fuzzy functional SIRMs connected inference model

This paper proposes a type-2 fuzzy functional SIRMs connected inference model (type-2 FF-SIRMs model) which extends fuzzy sets of the antecedent parts of the SIRMs model to type-2 fuzzy sets. It next discusses some properties of the type-2 FF-SIRMs model. We show that the inference results of type-2 FF-SIRMs model can be easily obtained from the area and center of gravity in consequent parts even where general type-2 fuzzy sets are used to the antecedent parts. Moreover, we state monotonicity of type-2 FF-SIRMs model, and show its applicability.

Some Consideration of SIRMs Connected Fuzzy Inference Model with Functional Weights

This paper discusses the SIRMs (Single-Input Rule Modules) connected fuzzy inference model with functional weights (SIRMs model with FW). The SIRMs model with FWconsists of a number of groups of simple fuzzy if-then rules with only a single attribute in the antecedent part. The final outputs of conventional SIRMs model are obtained by summarizing product of the functional weight and inference result from a rule module. In the SIRMs model of the paper, we use square functional weights, and compare with the conventional model.

Medical diagnosis and monotonicity clarification using SIRMs connected fuzzy inference model with functional weights

This paper discusses the SIRMs (Single-Input Rule Modules) connected fuzzy inference model with functional weights (SIRMs model with FW). The SIRMs model with FW consists of a number of groups of simple fuzzy if-then rules with only a single attribute in the antecedent part. The final outputs of conventional SIRMs model are obtained by summarizing product of the functional weight and inference result from a rule module. In the SIRMs model of the paper, we firstly clarify its monotonicity. Secondly, we apply the SIRMs model with FW to medical diagnosis.

SIRMs connected fuzzy inference model with compatibility functions

The single input rule modules connected fuzzy inference model (SIRMs model) can decrease the number of fuzzy rules drastically in comparison with the conventional fuzzy inference models. However, the inference results obtained by the SIRMs model is generally simple comapred with the conventional fuzzy inference models. For example, the SIRMs model can not transform to the product-sum-gravity model, if the fuzzy sets of the antecedent parts are limited to normal fuzzy sets. In this paper, we propose a SIRMs model with compatibility functions, which weights the rules of the SIRMs model. Moreover, this paper shows that the inference results of the proposed model can be easily obtained even as the proposed model uses involved compatibility functions.

Type-2 SIC Fuzzy Inference Models

The Single Input Connected fuzzy inference model (SIC model) by Hayashi et al. can reduce the number of fuzzy rules drastically compared with conventional fuzzy inference models. However, since the number of rules of the SIC model is limited compared to the conventional inference models, inference results gained by the SIC model are simple in general. From this reason, this paper proposes two type-2 SIC fuzzy inference models. We firstly propose a general type-2 SIC model as most common extended model. Secondly, we propose a type-2 SIC model with fuzzy functions. Moreover, additive type-2 SIC model with fuzzy functions are shown to be transformed to the SIC model with functional weights, and its inference results can be easily obtained.

Foreword: Special issue on recent advance in Intelligent Manufacturing Systems

Manufacturing systems are facing continuing transformational changes. The introduction of new technologies and the advances in equipment intelligence capability are having profound effects on production and logistics systems in practice. New solutions for manufacturing systems are evolving to respond to various needs. In particular, production and services are transformed from vertically integrated firms for product-based business into virtual collaborations of horizontal modular partners, while huge quantities of data are increasingly accumulated due to business integration. Intelligence and decision technologies for manufacturing systems are being developed to analyze data and to generate intelligent algorithms that enable automated manufacturing and logistics systems to control work flow, material flow, and information flow of the global supply chain networks. By seamless integration of intelligence and decision technologies, manufacturing systems have been completely changing the way we manage our factories, logistics, outsourcing, and supply chain networks. This special issue aims to address some critical issues involved in Intelligent Manufacturing Systems. This special issue entitled ‘‘Recent Advance in Intelligent Manufacturing Systems’’ contains expanded versions of selected papers from the proceedings of the 40th International Conference on Computers and Industrial Engineering (CIE40). The conference was sponsored by Computers & Industrial Engineering: An International Journal, and held in Awaji Island, Japan, July 25–28, 2010. It was hosted and organized by Kobe Gakuin University and Osaka Institute of Technology. The conference proceedings are available at the following IEEE Xplore website: http://ieeexplore.ieee.org/xpl/tocresult.jsp?sortType%3Dasc_p_ Sequence%26filter%3DAND%28p_IS_Number%3A5668158%29%26 pageNumber%3D3%26rowsPerPage%3D100&pageNumber=1. One of the themes of the conference CIE40 was ‘‘Industrial Engineering and Intelligent Manufacturing Systems’’, and the papers included in this issue are based on this theme. Among the 299 papers presented at CIE40, the authors of 31 papers were invited to submit enhanced versions of them for possible publication in the special issue. Of these, 31 papers were submitted. They were evaluated by the rigorous double-blind review