The Identification of Car Combustion Engine Category on Exhaust Emissions Data Pattern Base Using Sum Square Error Method

Abstract

There are two main categories of combustion condition in the car combustion engine chamber, they are complete and incomplete combustion condition. The complete combustion condition is obtained when a balance of fuel, airflow, and ignition in the combustion chamber occurred. It is characterized by the level of elements and compounds contained in the exhaust emissions. The dominant elements and compounds in exhaust emissions of the gasoline-fueled engine are Hydrocarbon (HC), Carbon Monoxide (CO), Carbon Dioxide (CO 2) and Oxygen (O 2). This study aims to establish a system that is capable of identifying the combustion category of automobile engine through exhaust emissions. This emissions data was procured using gas multi-sensor processed by a Fast Fourier Transform (FFT) process and constructed in the form of the data pattern. The identifying process of the combustion category was done by comparing the detected data pattern with a reference data pattern utilizing the Sum Square Error (SSE) method. Trials had been conducted on some gasoline and Pertalite cars for carburetor and injection systems. The trials result showed that the system accuracy in identifying the category of complete combustion was 87% and the incomplete combustion category was 77%.

Deep learning approach analysis model prediction and classification poverty status

“…The same explanation also explains that K-mean cluster is an unsupervised method that works efficiently in grouping big data [16]. To optimize K-means, the sum of squared error (SSE) method can to validate the cluster results [17], [18]. The K-means algorithm has been developed and used as a form of data grouping process on various problems [19].…”