<![CDATA[Active Queue Management (AQM) is a mechanism adapted for notifying senders with network congestion traffics before any overflow happens in the queue which is led to loss data. AQM technique can be applicable in different network size in different fields like industrial systems, colleges and government. In this paper, a nonlinear Fractional Order proportional Integral (NLFOPI) controller is proposed for controlling the Active Queue Management (AQM) system in a stable and robust behavior. An intelligent optimization algorithm called Pelican Optimization algorithm (POA) has been chosen for attain optimal system desired response based on tuning the proposed controller gains for minimizing the error depending on the use of Integral time absolute Error as a fitness function to maintain the whole tuning process based on Matlab program. The proposed NLFOPI controller is regarded as one of the fractional order controllers that depend on using one fractional variable for the integral term only, due to this the tuning parameter will be three instead of two also the nonlinear term will give an enhanced robustness that reflected clearly on system performance. The evaluation analysis represented by settling time, peak time, rise time and overshoot value appeared in system response are done, based on comparison with different classical controllers (PI-PID-FOPI) to show the performance of the proposed controller in different scenarios and then a robustness analysis is adopted by varying the desired queue number values in different time period and also by disturbance rejection when add disturbances signals with values ± 100 packets to desired number of queue in two different periods (15-35) sec., the results reflect how does the system faces these tests done efficiently. Based on simulation results, the NLFOPI controller is regarded as the best controller based on its faster peak time value (tp=3.8 sec) with stable response and a smooth rise time value (tr=1.8 sec.) also a fast-settling time (ts=3.4 sec.) is achieved with un noticeable overshoot (0.2%) if it is compared to other controllers then its robust response is appeared by achieving a satisfied stability and robustness.]]>
