A control performance evaluation method for chemical process with actuator valve viscous characteristics

Abstract

The invention discloses a method for evaluating the control performance of a chemical process with a viscous characteristic of an actuator valve. Obtain sequence y(t), collect yt at time t; (2) use radial basis neural network for nonlinear time series modeling on y(t), get initial predicted value (3) construct sequence and perform Gaussian / linear test , calculate the network evaluation index EHT; until the EHT reaches the set value, output the predicted value (4) loop operation steps (2) to (3), output the optimal value of Spread and the optimal output predicted value (5) construct the sequence and The modeling solution obtains the minimum variance performance lower limit of the control loop to be evaluated, and uses the minimum variance performance lower limit to evaluate the performance of the control loop. The invention is suitable for the control loop with the viscous characteristic of the executive valve, and can accurately evaluate the control performance benchmark of the chemical process.

Description

technical field [0001] The invention relates to the field of control performance evaluation of industrial control systems such as petrochemical industry, in particular to a control performance evaluation method of a chemical process with a viscous characteristic of an actuator valve. Background technique [0002] Industrial sites usually contain many control loops, and inspection and maintenance of these control loops is a tricky and arduous task in the operation of the production process. According to the survey, there are performance degradation problems in more than 60% of industrial controllers. How to use the daily operation data to measure the performance of the control loop, so as to optimize its settings to obtain more economic benefits, has become a hot spot for many scholars. In recent years, the research on performance monitoring and evaluation of control loops has become a topic of widespread concern in the field of industrial control. [0003] Control performa...

AI Technical Summary

Problems solved by technology

[0010] The performance evaluation method established based on the above two ideas has limitations in application: First, the method based on the spline smoothing function cannot completely remove the nonlinearity of the process output due to the interpolation function, resulting in performance evaluation results that are greater than the real value, and The error of the sticking characteristic of the actuator valve increases further; secondly, the method of searching for the equilibrium state of the actuator valve is more suitable for the control loop with a long enough stagnation stage of the actuator valve, and this control loop is the most important in practical industrial applications. hard to guarantee

Images