A novel generalized nonlinear predictive controller based on T-S fuzzy model and small-world algorithm is proposed. The dynamic of nonlinear system object is described by T-S fuzzy model based on chaos genetic algorithm, so a multi-step fuzzy predictor is derived. Furthermore, a novel small-world optimization algorithm with real-coding is proposed in order to avoid the deficiency of small-world optimization algorithm with binary and decimal coding or decoding, function tests and applying in moving horizon of nonlinear predictive control reflect the strong ability to optimize. Finally, generalized nonlinear predictive controller based on T-S fuzzy model of real data and a novel small-world optimization algorithm with real-coding meets the requirements of speed and real-time of control system well.