Application of SCE Algorithm to determination of PID controller coefficients

Due to water resources shortage in our country, improvement of water distribution management is unavoidable matter for increase irrigation network performance. Control systems have an important role in water distribution irrigation network and network succesful depends on performance. Succesful modelling and running depends on tuning of control coefficients. Regarding the control, in general, the water level deviation from a pre-defined target level should be kept as small as possible. Aim of this study is determination of optimal control coefficients using SCE method.
In this study, PID controller (upstream control and downstream control) is developed for slide gate on the ICSS model. PID is the most commonly applied type of controller in control engineering. An error value is calculated as the difference between a measured water level and a desired target level. The controller attempts to minimize the error by adjusting the flow passing under the gates. The PID algorithm consists of three control terms: the proportional term (P) which depends on the present error; the integral term (I) which is based on accumulated errors; and the derivative term (D) which is based on rate of change of errors. For tuning PID coefficients, SCE algorithm is used for increment and decrement operation scenario in Mc canal of Alborz irrigation network with 12.6km length at 5hr time operation. The objective function consists of three performance indexes of MAE (Maximum Absolute Error), IAE (Integral of Absolute Error) and dimensionless form of SRT (System Response Time). Also Molden and gates indexes is used for performance assessment canal.
Optimal coefficients controller are gained in operation scenario. Using gained coefficients, model will be able to tune water level in target level in increment and decrement process at short time. Maximum errors are related to C1 structure with 10cm in upstream controller and C3 structure with 6cm in downstream controller. Upstream controller are not able to tuning water level in downstream, so water supply in TO14 and TO16 is disturbed and MPA index is fair and poor performance classes Respectively for TO14 and TO16. The MPA index have been improved in downstream controller Compared to upstream controller. MPA, MPF, MPD indexes are good performance classes in both controller and MPE index is fair.
Results show gained optimal coefficients of the developed model to the upstream and downstream controllers is Responsive to changes in operation in a short time.