IMPLEMENTASI ALGORITMA K-MEANS DAN ALGORITMA APRIORI OPTIMASI KINERJA ECU (STUDY KASUS MOBIL AVANZA DAN XENIA)

Abstract

Currently the vehicle system has been controlled using an electronic ECU (Engine Control Unit). ECU damage will affect engine performance, so a system that can handle problems accurately detects quickly in making decisions is needed. In data clustering, there are several algorithms that can be used, such the K-Means Algorithm and the A Priori Algorithm, which is an algorithm with a high level of accuracy and the best among these three algorithms by doing a comparison using Rapidminer. Comparison of algorithms aims to obtain results and predictions from research that has been done. The development of the Analysis System with K-Mains and Data Clustering is a solution to help analyze data in the process of analyzing the optimization of ECU performance on vehicle engine performance including data collection, processing data, detecting weaknesses in digital data changes so that they can quickly optimize ECU performance in data grouping using K-means clustering. From the results of the K-Means Clustering study, it was found that C1(781-784), C2(896-927), C3(1223-1321), C4(1460-1587), and C5(1689-2716) Engine and A-Priori AUB support an average of 20%, Support A an average of 80% and an average Confidence value of 80%. Based on the remapping variation of 3 degrees of forward ignition produces engine power and stable engine torque at low speed of 781 rpm to high speed 2176 rpm  with remapping engine conditions at the time of testing.