System Identification and Application Based on Parameters Self-Adaptive SMO

This paper studies the identification algorithm of parameters self adaptive SMO based on linear kernel function, and analyses its performance and advantages. For ARX model and long-term prediction model, the method is used to identify the model of main steam pressure of thermal system and dual-lane gas turbine engine of aero system. The simulation results show that the algorithm can effectively identify model parameters and has a higher accuracy, reducing the requirements of training data including quantity and quality, so that its engineering applications and implementation are easier.