2 research outputs found
Implementasi Metode PCA – K Nearest Neighbour untuk Deteksi Golongan Kendaraan Berdasarkan Jumlah Pasang Gandar
No full textDetection of vehicle type based axle allow to do classify vehicle based on the load that will carried by the vehicle. However at this time, classification of heavy vehicle types based on axles is still carried out by humans. In here research, author use image processing for classification of vehicles class with PCA-K Nearest Neighbour Method. The reason for choosing this algorithm is because previous research use K-Means Clustering have a same formula with K-Nearest Neighbour. The data used is the image captured from the video camera processed by background subtraction method to separate moving object namely vehicle with its background. PCA method is required to obtain eigen vektor from vehicle and reduce its dimensions. Meanwhile, for classification author tries to use classification algorithm of K-Nearest Neighbour. To detect vehicle axles author use Circular Hough Transform method. The Testing Result shows detect vehicle type use PCA-K Nearest Neighbour method with value of K=1 has an accuracy 90%.Deteksi golongan kendaraan berdasarkan gandar memungkinkan untuk melakukan klasifikasi kendaraan berdasarkan beban yang akan di bawa oleh kendaraan tersebut. Namun pada saat ini, klasifikasi jenis kendaraan berat berdasarkan gandar masih dilakukan oleh manusia. Dalam penelitian ini, penulis menggunakan pengolahan citra untuk klasifikasi golongan kendaraan dengan metode PCA-K Nearest Neighbour. Alasan pemilihan algoritma ini adalah karena penelitian sebelumnya menggunakan algoritma K-means Clustering memiliki rumus yang sama dengan K-Nearest Neighbour. Data yang digunakan adalah citra yang ditangkap dari kamera video diproses dengan metode background subtraction untuk memisahkan objek yang bergerak yaitu kendaraan dengan backgroundnya. Metode PCA diperlukan untuk mendapatkan vektor eigen dari kendaraan dan mereduksi dimensinya, sedangkan untuk klasifikasi penulis mencoba menggunakan algoritma klasifikasi K Nearest Neighbours. Untuk deteksi gandar kendaraan penulis menggunakan metode Circular Hough Transform. Hasil pengujian menunjukkan deteksi golongan kendaraan menggunakan metode PCA-K Nearest Neighbours dengan nilai K=1 mendapat akurasi 90%.84 HalamanSkripsi Sarjan
TRANSPORTATION STRATEGY DEVELOPMENT FOR OPTIMUM COAL MINING: A CASE STUDY OF BLOK-D MINING PROJECT PT. XYZ
No full textBlok-D included in PT.XYZ concession is blessed with large coal resources and has the potential to be the main pillars of the company's future sustainability. The low calorific value of the coal resources and their locations are believed to erode the value of the mining project. Evaluate whether mining Blok-D is a relevant strategy for the Company, and then the development of the transportation strategy is a must before the mining project started. Selected transportation strategy is according to those who can provide a return on investment at equal or higher than the shareholder expectation.The increase in demand for coal by India and other countries in Asia was one of the reasons why the coal industry still had a strong appeal. Supported by its several competitive advantages, PT.XYZ has the advantage to grow the business by doing Blok-D mining. Assessment of the financial aspect is a must since it becomes the major challenge for transportation strategy development.Cost leadership is a suitable competitive strategy for the Company according to the analysis of the Company’s capabilities and competitive advantages. SWOT and TOWS approach shows that mining Blok-D is an inline strategy with the company’s corporate strategy. There are 6 (sixth) alternatives solutions for the transportation strategy, which are distinguished by the location of the barge loading facilities and the land coal transport method. The DCF financial model indicates that Alternative B1.2 provides the highest project value. The sensitivity analysis shows that the coal price followed by the OPEX becomes the most sensitive input variable for the project value. Monte Carlo analysis results show that Alternative B1.2 is viable to be executed since it gives an expected NPV of 333.47 million USD and an IRR of 65.4% which is higher than 11.96% as the required rate of return (WACC). Based on the analysis result, PT.XYZ is recommended to do the Blok-D mining projects with Alternative B1.2 as the transportation strategy.Keywords: Transportation strategy, business strategy, coal mining, investment analysis, DCF model, Monte Carlo analysis
