Development of a Canal Automation Model: A Laboratory Experiment

Abstract

The main objective of the project was to study and develop a canal automation model using local material and equipment. The canal automation model was developed at the Department of Irrigation Engineering, Faculty of Engineering, Kasetsart University, Kamphaengsaen campus. The model consisted of 4 gate controllers (Robogates) installed in the canal model for upstream water level control and 1 Robogate installed at the head tank to control the discharge into the canal model. Each Robogate controller is an embedded system designed to monitor the water level, gate positioning and to control the regulator. A Robogate is designed to work in 3 modes: Mode 0 (telemetering mode), Mode 1 (automatic mode) and Mode 2 (remote control). An upstream control algorithm was used for self-regulating the check gate in automatic mode. The performance of the canal automation model and the Robogates was tested in 6 runs. The farm turnout (FTO) gate was adjusted randomly to create disturbance to the flow in the canal model. The results showed that generally, the Robogate was very capable at controlling the water level in the model. The coefficient of variation of the water level upstream of the Robogate was very small, being less than 0.06 in all experiments. Two indicators–namely the maximum control error and the unsteady period, were selected for the analysis of the performance of the canal automation model under disturbed conditions. The percent maximum error of water level from the target was smaller for the most upstream Robogate compared to the downstream Robogates. The maximum error increased from upstream to downstream. The average maximum errors were 6.6, 9.4, 20.5 and 29.2% for Robogates 1, 2, 3 and 4, respectively. Although the maximum error was rather high, particularly for the downstream Robogates, this was only for a short time during the model testing. The average percentages of maximum error in controlling the water level between the 4 Robogates were significantly (P < 0.05) different. This result confirmed one of the disadvantages of the upstream control algorithm. The deviation from target, either in terms of water excess or water shortage, will usually be passed to the downstream reach of the canal or any flow disturbance will be passed to the downstream canal reach. The unsteady period ranged between 2 and 8 min and was 6 min on average. The analysis of variance of the unsteady time due to the adjustment of the FTO gate showed that the average time period for each Robogate to stabilize the water level back to the target level was not significantly (P > 0.05) different. This laboratory experiment showed that the Robogates could remove the effect of flow disturbances within a reasonable period of 2–8 min. This experiment helped in determining that the Robogates can be used for the effective and automatic control of the upstream water level.