The use of industrial big data high-level optimization to diagnose and analyze DCC devices has been applied in the petrochemical production process, promoting the revolutionary breakthrough of production process control. Combined with the company's idea of enhancing its core competitiveness through information technology, it then elaborates the existing data modeling methods, builds a software model, and tries to optimize its process big data and models through neural network algorithms to improve the potential of process control, so as to achieve the purpose of reducing costs and increasing efficiency. Finally, the characteristics of highlevel optimization of industrial big data and the problems and challenges that may be faced in the process of realizing this are discussed, in order to achieve a breakthrough in the company's overall business value.