Artificial neural networks for simulating wind effects on sprinkler distribution patterns

Abstract

A new approach based on Artificial Neural Networks (ANNs) is presented to simulate the effects of wind on the distribution pattern of a single sprinkler under a center pivot or block irrigation system. Field experiments were performed under various wind conditions (speed and direction). An experimental data from different distribution patterns using a Nelson R3000 Rotator ® sprinkler have been split into three and used for model training, validation and testing. Parameters affecting the distribution pattern were defined. To find an optimal structure, various networks with different architectures have been trained using an Early Stopping method. The selected structure produced R 2 = 0.929 and RMSE = 6.69 mL for the test subset, consisting of a Multi-Layer Perceptron (MLP) neural network with a backpropagation training algorithm; two hidden layers (twenty neurons in the first hidden layer and six neurons in the second hidden layer) and a tangent-sigmoid transfer function. This optimal network was implemented in MATLAB ® to develop a model termed ISSP (Intelligent Simulator of Sprinkler Pattern). ISSP uses wind speed and direction as input variables and is able to simulate the distorted distribution pattern from a R3000 Rotator ® sprinkler with reasonable accuracy (R 2 > 0.935). Results of model evaluation confirm the accuracy and robustness of ANNs for simulation of a single sprinkler distribution pattern under real field conditions. Additional key words: backpropagation; distortion by wind; ISSP; MLP neural networks; simulation; sprinkler; water application. ResumenRedes neuronales artificiales para simular el efecto del viento sobre el patrón de distribución del agua de un aspersor Se presenta un nuevo modelo basado en la técnica de Redes Neuronales Artificiales (RNA) para simular el efecto del viento sobre la distribución de agua de un aspersor, en un sistema estacionario o en equipos pivote. Se han realizado una serie de ensayos experimentales con diferentes velocidades y direcciones de viento, para el emisor Rotator R3000 de Nelson. El conjunto de datos obtenidos para los diferentes patrones de distribución del agua han sido divididos en tres grupos, y utilizados en las correspondientes fases de entrenamiento, análisis y validación. Se han definido los parámetros que influyen sobre el patrón de distribución de agua. Con el fin de encontrar una estructura de red óptima, varias redes con diferente arquitectura han sido entrenadas usando un método supervisado. Con la estructura óptima se consiguió un R 2 = 0,929 y RMSE = 6,69 mL para el grupo de ensayos correspondientes a la red Multi-Layer Perceptron (MLP) mediante el algoritmo de aprendizaje supervisado con retroalimentación, dos niveles de capas ocultas (veinte neuronas en el primer nivel y seis neuronas en el segundo) y una función de transferencia tangente hiperbólica. Esta red optimizada fue implementada en MATLAB para desarrollar un modelo llamado ISSP (Simulador Inteligente del Modelo de Distribución). ISSP utiliza la velocidad y dirección de vie...

Effect of terrain slope on water distribution and application uniformity for sprinkler irrigation

“…Based on the given radial leg water distribution data of a single sprinkler, the software can predict water distribution of a single sprinkler or a solid-set sprinkler irrigation system under various sprinkler spacing and different environmental parameters [17] . Faria et al and Siyanda et al simulated the impact of wind on water distribution patterns [18,19] . Mantovani et al and…”

Prediction of wind drift and evaporation losses of a sprinkler irrigation system using principal component analysis and artificial neural network technique

“…The back-propagation algorithm has emerged as one of the most widely used learning procedures for multilayer networks in the irrigation field (Sayyadi et al, 2012;De Menezes et al, 2015). In Fig.…”

“…Several studies have been conducted to develop sprinkler irrigation simulation models that can be used to estimate water distribution patterns of irrigation systems under real or controlled conditions (Sayyadi et al, 2012). These models can minimize the use of field tests and can improve the design and management of sprinkler irrigation systems.…”

“…As an example, a multilayer feedforward neural network with 2 input parameters (wind speed and wind direction) was employed by Sayyadi et al (2012) to simulate the effects of wind on the distribution pattern of a single sprinkler under a centre pivot or block irrigation system. A multi-layer perceptron neural network with a back propagation-training algorithm was used; furthermore, 2 hidden layers (20 neurons in the first hidden layer and 6 neurons in the second hidden layer) and a tangent-sigmoid transfer function were utilized.…”

Prediction of global horizontal solar irradiance in Zimbabwe using artificial neural networks

“…They consist of numerous numbers of interconnected processing elements called neurons. ANNs consist of an input layer, a hidden layer and an output layer (Yalcintals and Akkurt, 2005;Sayyadi et al, 2012). They are able to learn, memorize and create relationships between inputs and outputs (Sozen et al, 2004), which is a capability absent from empirical methods (Notton et al, 2013).…”