Abstract: An identification model of taste signals is developed based on fuzzy neural networks.The data compression and feature extraction of the sampled taste signals obtained using taste sensors are implemented employing wavelet transformation.Fuzzy neural networks are used to identify the taste signals.The training of network weights and the optimization of membership functions are conducted employing genetic algorithms.The data processing and fuzzy identification of mixed acid and sweet taste signals are realized.Simulated experimental results show that it is feasible and effective to introduce fuzzy neural networks into the fuzzy identification of taste signals.