An Online Sintering Batching System Based on Machine Learning and Intelligent Algorithm

Aiming at the problem that accuracy and economy of traditional off-line batching method are not high, online system (BSMLIA) based on machine learning intelligent algorithms was put forward from three aspects: real-time, technical requirements economic benefits. The accurate solution on-line fast calculation sintering raw material ratio under influence multiple factors solved. Specifically, a BSMLIA architecture with levels data communication layer (DCL), parameter prediction optimization (PPBOL), diagnostic decision (DDL) first designed to realize monitoring abnormal diagnosis sinter performance. Then, adjustment module (SBAOM) elaborated. mixture performance model developed by MLR LightGBM algorithm, can be composition quality index current production process parameters calculate appropriate In addition, pre-batching were established achieve lowest cost for given Finally, actual used verify SBAOM. results proved only quickly plan meets requirements, but also reduce RMB 29.54/ton.