Rancang bangun model sistem penunjang keputusan cerdas untuk sistem rantai pasokan beras di provinsi DKI Jakarta

Abstract

DKI Jakarta is the region which has a very large population but it is not supported directly by rice field area that can meet the needs of rice for its population. Cipinang rice market center (PIBC) managed by PT. Food Station Tjipinang Jaya (FSTJ) is expected to be a party that can manage and control the rice for food security, especially in the Jakarta area. There are several important aspects to be considered by FSTJ in regulating and controlling food security. These aspects are the supply of rice, rice price, rice supplier selection, transportation and distribution aspects of rice as well as the performance aspect of the rice supply chain. The purpose of this study was to develop a model of intelligent decision support system for rice supply chain that covers all these aspects effectively and efficiently. There were several methods used in this study. Artificial neural network was used to analyze aspects of the supply and price of rice, TOPSIS was used to analyze the rice suppliers selection, simulated annealing was used to analyze the distribution and transportation of rice and fuzzy inference system was used to analyze the performance of the rice supply chain. Analysis results showed that the accuracy of forecasting models to forecast the supply of rice and to forecast rice prices have reached more than 90 percent. Rice supplier selection model is effectively able to sort the suppliers of rice based on predetermined criteria. Transportation and distribution model has effectively been able to make the shortest route to distribute the rice. This model also has been able to make the sequence assignment of vehicles to deliver the rice to its customers. Finally, the rice supply chain performance model also effectively has been able to measure the performance of supply chain that accommodates the input factors. According to the results of analysis and experts opinion, there were advantages and disadvantages of the model produced, but they can be verified and they were also valid.