Simplified Discrete Binary PSO tuned Multivariable PID Controller for Binary Distillation Column plant

Abstract

In this paper a Multivariable PID Controller is designed for a Binary Distillation column plant. The Binary Distillation Column is mainly used for the separation of liquid mixtures containing volatile and non-volatile components. To maintain the mole fraction of the distillate the PID Controller regulates the vapour tlow rate and liquid tlow rate. Decoupler is used to avoid loop interactions as it may cause instability. The PID controller is tuned using the Particle Swarm Optimization (PSO) algorithm and the Discrete Binary Particle Swarm Optimization (DBPSO) algorithm and Simplitied Discrete Binary Particle Swarm Optimization (SDBPSO) algorithm to obtain reduced settling time. Simulations are carried out to minimize Integral of Absolute Error (IAE). Transient response analysis shows the better performance of SDBPSO algorithm over PSO and DBPSO algorithm. Stability analysis is done using pole-zero map and Nyquist plot.

A Survey on Multi Objective Optimization Challenges in Swarm Intelligence

“…The Discrete Binary PSO method was presented by Queen and Auxillia [12]. The Hamming distance between the particle at time t and time 𝑡 + 𝑙 is used to represent the particle's velocity in this procedure.…”

A New Method of particle swarm optimization approach for tuning PID controller

“…The control input will be feedback to electric valve to change the flow rate, and then the level of water in each tank can be controlled [31]. (3) (4) (5) From eq.(3),eq. ( 4) and eq.…”

“…Z-N method fails to provide an acceptable performance because it is always provides a large overshoot and settling time, so that the values of the PID parameters are often subsequently refined in accordance with the operator's experience [37]. many evolutionary computation methods such as genetic algorithm (GA), differential evolution (DE) and particle swarm optimization (PSO) have been employed to tune PID controller in various plants [2][3][4][5][6][7].However, the simplicity of PSO (i.e. it is straight forward and has less parameters to be tuned) and its low computational cost with high performance make it commonly used in the industrial applications [43].…”