Integrating Fuzzy Logic with a Modified Ichimoku Indicator for Bitcoin Price Forecasting and Trading Decision Support

This paper presents a hybrid decision-support framework that combines the Ichimoku Kinko Hyo technical analysis system with fuzzy logic for forecasting Bitcoin prices and generating trading signals. The Ichimoku indicator integrates multiple components to evaluate trend direction, momentum, support/resistance levels, and equilibrium, but its signals are typically qualitative and reliant on subjective judgment. Fuzzy logic is applied to address this by modeling uncertainty through linguistic variables, membership functions, and IF-THEN rules derived from expert knowledge. A modified Kijun-sen line (extended to ۱۰۳ periods) is introduced to enhance multi-timeframe stability. Key Ichimoku-derived variables are fuzzified, and a Mamdani fuzzy inference system produces buy, sell, or hold recommendations. The model is implemented in MATLAB and tested on real daily Bitcoin (BTC/USD) data from Yahoo Finance and CoinMarketCap, spanning November ۲۰۲۵ to February ۲۰۲۶. Results indicate consistent and interpretable signals aligned with price movements, including the correction from approximately $۱۰۱,۳۰۰ on November ۶, ۲۰۲۵, to levels near $۶۷,۹۶۵ on February ۲۷, ۲۰۲۶. The approach offers improved decision consistency for traders in volatile cryptocurrency markets.

کلیدواژه ها:

Ichimoku system ، fuzzy logic ، Bitcoin forecasting ، trading signals ، fuzzy rules