Generation of Fuzzy Rules with Subtractive Clustering

Abstract

Pembelajaran sistem pangkalan peraturan kabur menggunakan algoritma genetik mempunyai masa depan yang cerah bagi menyelesaikan beberapa masalah. Lojik kabur menawarkan cara sederhana bagi menyimpulkan maklumat input yang kasar, kabur, cacat atau tidak jelas. Model lojik kabur adalah berasaskan kaedah–kaedah empirik bergantung kepada pengalaman operator berbanding dengan pengetahuan teknikal daripada sistem. Dalam metod lojik kabur, sebarang input yang munasabah dapat diproses dan sebilangan output dapat dijana meskipun penakrifan pangkalan peraturan secara cepat dapat menjadi rumit sekiranya terlalu banyak input dan output yang dipilih untuk sebuah penggunaan. Bergantung kepada sistem, semakin rumit input dan output yang ingin diselesaikan oleh sistem, maka akan semakin banyak jumlah bilangan peraturan dan kerumitan tetapi juga akan menambah mutu kawalan dari sistem. Banyak kaedah telah dicadangkan bagi menjana peraturan kabur. Idea asas daripada penyelidikan ini adalah untuk mempelajari serta menjana peraturan paling optimum yang diperlukan bagi mengawal input tanpa mengurangi mutu kawalan. Kertas kerja ini yang mencadangkan penjanaan peraturan kabur menggunakan penggugusan subtraktif pada lojik kabur Takasi–Sugeno–Kang (TSK) bagi kegunaan kawalan lampu isyarat lalu lintas. Kata kunci: Lojik kabur TSK, sistem pangkalan peraturan kabur, teknik penggugusan subtraktif Learning fuzzy rule–based systems with genetic algorithms can lead to very useful descriptions of several problems. Fuzzy logic (FL) provides a simple way to arrive at a definite conclusion based upon vague, ambiguous, imprecise, noisy or missing input information. The FL model is empirically based, relying on an operator’s experience rather than their technical understanding of the system. In the FL method, any reasonable number of inputs can be processed and numerous outputs will be generated, although defining the rule–base quickly becomes complex if too many inputs and outputs are chosen for a single implementation since rules defining their interrelations must also be defined. This will increase the number of fuzzy rules and complexity but may also increase the quality of the control. Many methods were proposed to generate fuzzy rules–base. The basic idea is to study and generate the optimum rules needed to control the input without compromising the quality of control. The paper proposed the generation of fuzzy rule base by subtractive clustering technique in Takagi–Sugeno–Kang (TSK) fuzzy method for traffic signal control system. Key words: TSK fuzzy logic, fuzzy rule base system, subtractive clustering technique

Adaptive neuro-fuzzy inference system for temperature and humidity profile retrieval from microwave radiometer observations

“…Fuzzy logic (FL) provides a simple way to arrive at a definite conclusion based upon vague, ambiguous, imprecise, noisy and missing input information (Priyono et al, 2005). Most of the FL models are empirically based, relying on an operator's experience rather than a technical understanding of the system.…”

Data-Based Fuzzy TOPSIS for Alternative Ranking

“…*) Penulis korespondensi: victor.utomo@ gmail.com Given a set of data, set the normal value based on and use the following model (Priyono, 2005).…”

Designing Fuzzy Inference System to Diagnosis down syndrome by Face Processing

“…Mamdani fuzzy system used is a 1-input, 1-output system applying the Euclidean distance between the color of each pixel to the average skin color sub-space as an input, and the likelihood of being skin pixel as an output. Subtractive clustering [11] is applied on input space (contain 132000 skin and non skin pixels) to decide about the number of MFs and rules. Utilizing the obtained four clusters information and experimental knowledge, input and output MFs are designed.…”