Classification and evaluation of emitter clogging degree and prediction method of emitter clogging risk

An accurate and rapid evaluation has been a high demand for the clogging state of the emitter during the operation of the drip irrigation system. It is also necessary to quantify and predict the dynamic trend for the clogging degree of the emitter with the increase in irrigation time. In this study, a fuzzy comprehensive and quantitative evaluation was established to predict the clogging risk of emitter during drip irrigation, where the average relative flow of the emitter and the irrigation uniformity were taken as the evaluation indicators. The entropy weight method and the triangular membership function were selected to calculate the weight and membership of each Evaluation Index (EI). A gray GM (1, 1) prediction model was utilized to predict the dynamic change of the emitter clogging degree with the increase of the operation time in the drip irrigation system, according to the EI emitter clogging state. The results show that the emitter clogging presented a gradual fluctuating with the increase of irrigation time. Two stages were divided in this period, including stable fluctuation and rapid development. An excellent linear relationship was obtained between the average relative flow of the emitter and the uniformity of irrigation, indicating the better consistent clogging of different emitters. The EI value of the same emitter presented a better autocorrelation between the comprehensive EI values of the emitter under different irrigation times. It infers that the experimental data of the emitter in the early stage was used to predict the later blockage of the emitter with the increase of the irrigation time. Two stages were also divided in the comprehensive EI value of the emitter for the dynamic change trend with the increase of irrigation time, namely the stable fluctuation and rising stage. The grading standard of the emitter clogging degree was then proposed using comprehensive EI, where the emitter clogging degree was divided into five grades. The optimal number of initial sequences was determined when the gray GM (1, 1) model was used to predict the blockage of the emitter. The prediction accuracy of the model was the best when the number of initial sequences was 5. The overall prediction accuracy of the seven emitter clogging degrees was 85.7%, and the relative errors were all less than 15%. The reclaimed water and Yellow River water were selected to further verify the reliability of the gray GM (1, 1) prediction model. The overall prediction accuracy of the gray GM (1, 1) prediction model was 88.1% and 96.2%, respectively, and the average relative errors were both less than 11%. Consequently, the gray GM (1, 1) prediction model can be widely expected to predict the dynamic change trend of the comprehensive EI value of emitter clogging with the increase of irrigation time. In addition to the high prediction accuracy, it is necessary to extend the model for accurate prediction in the actual irrigation. This finding can provide a theoretical basis to evaluate the anti-clogging performance of emitters for the configuration of anti-clogging preventive measures in drip irrigation.