Research On Application Of PSO Algorithm In Hot Rolling Coiler Tension Control

Hot rolling coiler is a very important equipment in hot rolling production line, its working performance mainly depends on tension control effect. Tension control directly affects the quality of the final products of steel coils, to ensure the control system has good performance, the key is to improve tension control system performance, maintain constant tension, avoid tension fluctuate.This paper takes the project as the background, researches on the tension control in the normal coiling section. Tension control has three ways:direct tension control, indirect tension control, compound tension control. Indirect tensile control is applied in most actual projects, the shortcoming of this kind of control is it cannot detect the tension, and the control precision is bad. So this paper proposes detecting tension with soft sensor, and adds the tension feedback on the foundation of the indirect tension control, forms the tension control closed loop, and establishes the compound tension control system, thus controls tension constant better.This paper establishes the mathematical model of compound tension control system, gives the computational method of forward-feed current compensation value, and designs the current regulator and the tension regulator. In view of using the general method to tune PID parameters cannot obtain the best control effect,this paper proposes to optimize the tension regulator PID parameters with fuzzy PSO algorithm, through the simulation analysis, the effect of this method is better.In view of the complexity, time-varying parameters of coiler system, the fixed PID parameters cannot satisfy the control requirement. This paper analyses the variation rule of the parameters of the controlled object, the analysis indicts that the main time-varying parameter of the controlled object is the electromechanical time-constant Tm, other parameters are nearly invariable.So this paper adopts the method of off-line to optimize, online to adjust, calculates system’s electromechanical time-constant through the coil diameter, then optimize multiple times of controlled object’s precise mathematical models with improved PSO algorithm, it obtains an array of PID parameters, and then establishes the corresponding relationship between the coil diameter and PID parameters using BP neural network, thus adaptively adjust the parameters of the tension controller through detecting coil diameter in the process of coiling. Through the simulation analysis, this method is a very good solution to solve the performance influence caused by system’s changing parameters.