In recent years, due to the improvement in industrial and living standards, the demand for electricity has repeatedly reached new heights every year. Because of the low price of coal, coal-fired power generation accounts for a large proportion of the total power generated. In addition to emitting large amounts of carbon dioxide and exacerbating climate change, coal combustion emits a large number of harmful substances, such as PM2.5, that poses a direct threat to human health. The Paris Agreement came into force at the end of 2016. With the worsening global air pollution problem, all governments began taking relevant measures. Among these measures, the low-emission and high-efficiency coal gasification technology is the most practical method. Coal gasification is caused by the pyrolysis (thermal cracking) of coal, air, and water to produce syngas. However, due to the uncertainty of coal quality, controlling the optimal proportion of coal, water, and air response to the output during pyrolysis is very important. First, in this study, Matlab or Simulink were used to establish a dynamic simulation platform for the gasifier. Second, these software were used to find the optimal proportion of coal, water, and air through the Taguchi method. Finally, the multiple adaptive neural fuzzy inference system (MANFIS)/particle swarm optimization (PSO)/proportional-integral-derivative controller was designed using MANFISs and PSO. As the water-gas reaction and the Boudouard reaction are the major reactions of the gasification process, the uncertainty of coal quality was compensated in this study by adjusting the amount of water to improve the output of synthetic gas and reduce slag. (C) 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.