Abstract: In practical on-site flotation control, a problem with simple structure but difficult to control due to time-varying characteristics is often encountered. Classic PID parameters have a wide range of variation, and system coupling makes maintenance difficult. Advanced control theories usually require complex computer control software and abstract theories that are difficult for field technicians to understand. Therefore, this paper attempts to use a first-order model and the dynamic matrix control method in model predictive control theory to simplify the situation. By using a simple and easy-to-understand approach, a dynamic matrix predictive control scheme that can be implemented in PLC is designed. The problem is simplified to solve a quadratic equation and utilize the error correction and rolling optimization characteristics of the dynamic matrix control method to compensate for the error of the first-order model.