Multi-objective Optimization And Control Method For Semi-autogenous Grinding Process Operation

The current situation of mineral resources and metal demand makes improving the processing efficiency of grinding equipment an important development direction.Semi-autogenous grinding process is also showing its advantages in grinding process.However,the core automation technology of domestic semi-autogenous grinding process is not mature,and there are various difficulties in control.In order to make the operation control of semi-autogenous grinding process free from human participation and realize automation.In this paper,the semi-autogenous grinding process is taken as the research object and the multi-objective optimization and control of operation are studied.The main work is as follows:Through the analysis of operation principle and internal particle movement state of semi-autogenous mill,considering the influence of various factors on mill operation,the balance relationship between import and export materials of semi-autogenous mill is determined,and the control loop state equation model of mathematical relationship between important internal state and controllable quantity and observable output of semi-autogenous mill is established,which provides an important basis for parameter optimization and control.Based on the mathematic model of the important working parameters of semi-autogenous grinding operation and the analysis of the influencing factors,the multi-objective optimization model and constraints of semi-autogenous grinding operation are determined.The multi-objective optimization model is solved by using the non-inferior hierarchical sorting genetic algorithm NSGA2 with elite strategy.The Pareto solution set is determined by TOPSIS decision-making method.The optimal working state is taken as the control expectation of the control strategy.Aiming at the unmeasurable characteristics of the internal state of semi-autogenous grinding process,an extended state observer based on strong tracking filter(STF)is proposed,which observes the internal state variables of the system to determine the state errors,and estimates the combined real-time effect of the unmodeled part and disturbance.It can be used in subsequent control laws.Compensation is made in the determination.Aiming at the problem that the semi-autogenous grinding process is difficult to control because of its non-linearity,multi-variable coupling and unmeasurable internal information,an auto-disturbance rejection control(ADRC)method for semi-autogenous grinding process operation based on STF is proposed,which makes the semi-autogenous grinding process run stably and achieves many optimization objectives such as increasing output,saving energy and reducing consumption.The simulation results show that the STF-based ADRC control method adopted in this paper achieves the desired effect,and the algorithm is simple and easy to understand,and the amount of calculation is small.Compared with the 7%and 3%overshoot of the PID control method,the control method has no overshoot,and has more stability and shorter response time to achieve the control goal.So it has important application value.