تخمين عيار طلا در كانسار خوني با استفاده از بررسي رفتار عناصر طلا، آرسنيك و آنتيموان به روش خوشه بندي K-means.

Estimation of grade Gold in khooni deposit using the behavior of gold, Arsenic and Antimony elements by clustering k-means method.

محدوده اكتشافی خونی در270 كیلومتری استاناصفهان در منطقه انارك و در زون ایران مركزی واقع شده است. با توجه به وجود شواهدی حاكی از كانی‌سازی طلا در این ناحیه، شناسایی نواحی امید بخش معدنی در این منطقه ضروری است. بنابراین ضرورت یافتن اطلاعاتی در مورد ارتباط و چگونگی رفتار عناصر طلا، آرسنیك و آنتیموان نسبت به یكدیگر در این محدوده، برای پیدایش و تعیین وسعت هاله­های ژئوشیمیایی، تعیین امیدبخش بودن منطقه و تخمین عیار اهمیت می­یابد و باید مورد بررسی قرار گیرند. بدین منظوردر تحقیق حاضر برای تعیین رفتار عناصر نام برده، با استفاده از روش معروف و كارآمد K-means انجام می‌شود كه یكی از روش­های خوشه­بندی بوده و مبتنی بر كمینه نمودن مجموع فواصل اقلیدسی هر یك از نمونه­ها از مركز دسته­ای كه به آن تخصیص می‌یابد، است. در این پژوهش از تابع كیفیت خوشه­بندی و میزان مطلوبیت نمونه در خوشه مورد نظر (S(i))، برای تعیین تعداد خوشه بهینه استفاده شده است و در پایان با توجه به مراكز خوشه­ها و نتایج حاصل، معادله­ای جهت تخمین عیار عنصر طلا بر حسب چهار پارامتر عیار آرسنیك، عیار آنتیموان، طول و عرض نقاط نمونه‌برداری ارائه می‌شود.

Summary Khooni exploratory area is located at 270 kilometers of Esfahan, and belongs to Central Iran geological zone. According to some instances expressing gold mineral forming in this area, make reconnaissance of favorable area as an essential fact. A well-known algorithm of clustering is K-Means by which the data are divided into K clusters on the basis of distance. In this analysis, using the K-Means method to classify the sampling of khooni district for Gold, Arsenic and Antimony elements. The optimal K value was determined and then the data were clustered and the behavioral characteristics were analyzed, and at the end relationships and equations with correlation coefficients were identified and the grade of gold were estimated based on four parameters Arsenic, Antimony content; length and width of the sampling points.   Introduction There are many methods to estimate the grade because of mining projects have high dependency to determine the tonnage accurately. One of the new methods of estimating the grade is clustering. Thus, it is necessary to find information about the relationships and behavior elements of Gold, Arsenic and Antimony to each other. in this area to determine the extent of the geochemical halo and to identify favorable area as well as estimation.   Methodology and Approaches In this study fordetermination the behavior of the mentioned elements, the famous and efficient method of K-Means, which is one of the clustering methods, are used, that is based on minimizing total of Euclidean distance from the center each class. For this purpose, the quality function and silhouette criterion have been used to determine the optimal number of clusters. All data ranged the number of classes from k=3 to k=10 and afterwards the optimumnumber of clusters were selected by mentioned methods. At the end, according to the Center of clusters, the equation for estimating the grade of gold is provided. In this study, MATLAB and SPSS software are used to achieve results.   Results and Conclusions The relationship of gold, arsenic and antimony according to length and width of the sampling points were determined for estimation grade of gold, Results showed that clustering with k=5 in case of Au, As and Sb were better than other classes number in each case. Obtained equation for estimating grade of gold element based on four parameters Arsenic, Antimony content; length and width of the sampling points, is  and the correlation coefficient 73% is reported. Also to evaluate the accuracy of the estimator, the results were validated. Estimated gold element values almost match with actual values of gold element that represents the accuracy of the used method.