Peramalan Jumlah Persediaan Sparepart Carburetor Assy dengan Metode Exponential smoothing Brown dengan Genetic Algorithm (order cross over dan one point mutation) di PT Hitachi Astemo Bekasi Powertrain System

Abstract

Abstrac: One of the business people in the automotive sector who continues to move as a manufacturing company in the production of vehicle spare parts is PT ABC which is the largest manufacturer of spare parts for several vehicle companies such as Astra Honda Motor (AHM), Yamaha and so on. One type of production from PT ABC is the carburator assy which functions to mix fuel and air to produce propulsion for vehicles. Seeing its very important function, PT ABC places the carburator assy as a type of spare part that must be produced every month In research, a method is needed to solve the problems that occur. Research methods are scientific ways to obtain data / information as it is and not as it should be, with certain purposes and uses. This research is based on descriptive research. This study used the double exponential smoothing Brown method for forecasting carburator assy 052A with Alpha assessment using Genetic Algorithm order Cross over PT ABC.Calculations carried out for forecasting the Double exponential smoothing Brown method using Alpha values of 0.36 and 0.32 obtained randomly resulted in RMSE values that were greater than Alpha values obtained using Genetic Algorithms, namely 103.7234 for Alpha values of 0.36 and 117.6129 for Alpha values of 0.32. Therefore, it can be interpreted that determining the Alpha value using the Genetic Algorithm method can be a solution to reduce the RMSE error value in Double exponential smoothing Brown forecasting. In research conducted at PT ABC by analyzing the influence of Genetic Algorithms in determining α values, several conclusions can be obtained, namely as follows: Determination of α value for the calculation of Double exponential smoothing Brown using the Genetic Algorithm obtained an Alpha value of 0.43 with the smallest error result of 95.29633. Calculations MADe on Genetic Algorithms can determine α value that can minimize errors in forecasting compared to α values without Genetic Algorithms.