Pressure Vessel Alarm Threshold Valve Optimization Research Based On Data Driven Chemical

In current Pressure vessel alarm system, the biggest problem with alarm quantity is big, false alarm and alarm correlation is more, the alarm management confusion, alarm utility sell at a discount greatly.Its main reason is that the alarm threshold Settings is not reasonable, alarm system performance poor. So far,there are not unified, very effective methods to calculate alarm threshold, on-site equipment alarm threshold Settings are often depended on personal judgment and subjective experience and decide and lack of theoretical guidance, which is currently alarm system the main reason for the low performance. In order to improve the performance of the alarm system and improve the Pressure vessel operation safety and reliability, it is necessary to optimize the alarm threshold.This paper first introduces research results in the fields of the alarm at home and abroad, this paper analyzes the problems of traditional threshold settings,puts forward a new method for the threshold settings based on data drive technology, and applies the new method to the specific chemical machinery, obtains the satisfactory results.The main content of this paper includes four aspects:(1) The introduction of Theory.Introducing the nonparametric estimation of probability and statistics—kernel density estimation, and through the actual research sample, We obtained the influence of the sample size, the kernel function and the window width on the kernel density estimation results.(2) Method validation. We set up alarm threshold optimization target function based on the Bayesian decision criterion of the pattern recognition, and obtain the best value of the alarm threshold by solving the objective function.Then apply this method to industrial simulation platform TE process and DMF recovery process and verify the validity of the new method through the simulation analysis. (3) Software implementation. User interface concerning process alarm thresholds optimization is developed based on GUI in MATLAB to realize the function of threshold optimization.(4) Come to the conclusion.