توسعه يك مدل عددي و تعيين بهينه دبي آب دوش‌ها در فرايند ميز خنك‌كاري ورق فولادي

System modeling at the second level of automation plays a crucial role in process control, system configuration an‎d producing products with the desired specifications. The run-out table is one of the main components in the hot rolling process, responsible for controlling the sheet temperature after it exits the rolling section an‎d reducing it temperature to a specified range. Adjusting the flow rate of each cooling shower directly affects the cooling rate an‎d the thermal profile of the sheet, bringing the stripʹs thermal profile closer to the target thermal profile to produce a product with the desired quality. Modeling the run-out table system at the second level of automation provides an intelligent infrastructure for adjusting water flow an‎d precisely managing sheet temperature before coiling. The run-out table consists of a series of showers an‎d spray nozzles that apply water at controlled flow onto the moving sheet surface. The target thermal profile on the run-out table is determined based on the cooling strategy, desired cooling rate an‎d target coiling temperature. The goal of the modeling thermal behavior on Level 2 system is to adjust the shower flow so that the actual thermal profile matches the target profile. Duo to the complexity of the process, interaction of various parameters an‎d changes in operating conditions of the run-out table, achieving the target cooling profile requires a fast an‎d intelligent mechanism that can align the actual thermal profile with the target while considering the system’s physical constraints. In this thesis, a neural network model of the run-out table is designed an‎d trained using simulation data. The trained model can estimate the thermal profile of the sheet along the run-out table based on the shower flow an‎d sheet specifications. Using the numerical model of run-out table allows faster computations compared to traditional computational models. Based on this numerical model, gradient-based optimization algorithm has been developed, taking into account the operational constraints of the cooling showers. To further accelerate flow calculations, an inverse model of the run-out table is also defined, which determines suitable initial values for the optimization algorithm based on the target thermal profile. The results of implementing this structure show that the proposed algorithm, considering time an‎d operational constraints, can accurately approach the target thermal profile. This method can be used as part of an advanced control system an‎d serves as an effective step toward smart industrial processes, reducing energy consumption, improving product uniformity, an‎d increasing efficiency.

جعفر قيصري

جواد عسگري مارناني , حامد جلالي بيدگلي