its simplicity and robustness. PID controller is still favorableto be used with more than 90% of industrial controllers [35,36].However, finding the optimal parameters of PID controller is verydifficult especially in non-linear control system as in liquid levelcontrol systems. So, several methods have been proposed to tunePID controller. One of these methods is Ziegler and Nichols (Z-N)method [1]. It is the oldest method and simplest one. However, it isoften hard to determine optimal PID parameters with Z-N formulain many industrial plants. Z-N method fails to provide an accept-able performance because it provides a large overshoot and settlingtime, so that the values of the PID parameters are often subse-quently refined in accordance with the operator’s experience [37].Recently, many evolutionary algorithms such as differential evolu-tion (DE), genetic algorithm (GA) and particle swarm optimization(PSO) have been employed to tune PID controller in various plants[2–7]. However as PSO is straight forward, has less parameters tobe tuned and its low computational cost with high performancemakes it commonly used in the industrial applications [43].PSO is an evolutionary computation technique introduced in1995 [8]. PSO is inspired from imitating the behavior of flock of