Proposed multivariable process optimization of four tanks controlled by PID controller using paraconsistent analysis network

Vinícius Rodrigues dos Santos, Cláudio Rodrigo Torres

Abstract


Recovering the control of an industrial process, after an issue, failure or operational error, means financial savings, production gain, prevention of environmental impact or saving lives. In the industry field there are processes that yield products and by products, with their respective inherent risks. A multivariable process, in order to achieve stability, complying with a specification, depends on the influence of more than one variable. In this paper, issues were simulated, altering their dynamics in a “X” parameter, using a multivariable system of level control in four tanks, developed to study the limitations in the performances of controllers while changing the process dynamics. The objective was to promote the reestablishment of the control over this multivariable process and interconnect it into a PID controller (proportional-integral-derivative). In order to reestablish the control, an artificial intelligence network and paraconsistent annotated logic with annotation of two values (PAL2v) were developed and tested in an innovative way for this plant. Tests were performed with various parameters where conventional controllers weren’t effective in recovering control. These tests consisted of three simulations, a total of forty runs, twenty using   the  paraconsistent analysis network (PAN). Many favorable results were found due to the PAN’s detection when the plant’s control was lost, even with the dynamic variable of the process, without informing the PAN of any parameters over the changes in the dynamics (“X” parameter) and also without human interference. After detecting the loss of control, PAN and auxiliary circuits developed in this research could return the control to the PID controllers, complying with the specifications of the process. This paper aims towards the possibilities of the use of artificial intelligence in the recovery of control in industrial processes.

 

Keywords: Processes control, Control Techniques, PAL2v, Control Recovery, Automation

Full Text: PDF

Refbacks

  • There are currently no refbacks.