Reduced Order Modelling of LTI Systems using Invasive Weed Optimization

Reduction of single-input single-output (SISO) continuous time systems into reduced order model (ROM) using invasive weed optimization (IWO) has been presented here. IWO is a novel population based numerical stochastic, derivative free optimization algorithm inspired from the biological growth of weed plants. The objective function is the integral square error (ISE), which has been minimized. The step and frequency responses of both low and high order systems have been compared. Also, the performance of IWO has also been compared with other existing techniques in terms of ISE and transient response parameters. It has been observed that IWO outperforms the existing techniques in literature by giving low value of ISE and better transient response's parameters.