Aiming at the characteristics of significant nonlinearity, large time delays, and strong coupling in the coking production process, this study designs a net coking time judgment system based on the Elman neural network, an expert net coking database, and an intelligent control system were designed. The expert system guided the coke oven heating control system based on coking coal characteristics, planned production, and net coking time. Following the guidance from the expert system, the heating control system adopted an adaptive neural network PID control strategy to optimize the coking process, thereby improving coke quality, saving energy, and reducing pollution. Through long-term operation, the system had demonstrated significant economic and social benefits, verifying its feasibility, advanced nature, and substantial potential for widespread application.