ANN Analysis of Injection Timing on Performance Characteristics of Compression Ignition Engines Running on the Blends of Tropical Almond Based Biodiesel

Abstract

In the present work, biodiesel prepared from Tropical almond oil (Terminalia Catappa) was used as fuel in C. I engine. Performance studies were conducted on a single cylinder four-stroke water-cooled compression ignition engine. Experiments were conducted for different percentage of blends of Tropical almond ester with diesel at different injection timings. Experimental investigations on the performance parameters from the engine were done. Artificial neural network (ANN) of back-propagation feed-forward Levenberg-Marquardt algorithm was used to predict the performance characteristics of the engine. An ANN model was developed for the performance parameters. To train the network, blend percentage, percentage load and injection timings were used as the input variables whereas engine performance parameters (brake thermal efficiency, exhaust gas temperature, and brake specific fuel consumption) were used as the output variables. The obtained experimental results were used to train the network structure. Results showed very good correlation between the ANN predicted values and the desired values for various engine performance values. Mean relative error values were less than 10 percent which by many standards is acceptable. The results show that ANN is an accurately reliable tool for the prediction of engine performance.