Yangi flotatsion texnologiyani oltin-mis-molibden rudalarini qayta ishlash Mikheev boyitish fabrikasining misolida tahlil qilish uchun yangi yo'li ko'rib chiqilgan. Flotatsiya jarayonini sanoatdagi tahlil qilishning ilmiy yondashuvi ionometrik xaritalash va neyron tarmoqlarini modellashtirish metodologiyasiga asoslangan. Ko'rsatilgan texnologik ko'rsatgichlarning katta farqlanishi sababli qoldiqli texnologiyadan foydalaniladi, bu esa flotatsion xomashyo zaxiralarida oltin va molibdenning yo'qotilishiga olib keladi. Harmonik tahlil usulini qo'llagan holda, namunaning bir xil emasligi va tadqiqot qilinayotgan kuzatuvlar ikkita “qo'shib qo'yilgan” taqsimot bilan hosil qilinganligi isbotlangan. Natriy sulfidni zavodda qabul qilingan texnologik jarayon bo'yicha jarayonning boshida faqat bitta nuqtaga etkazishga urinishlar noqonuniyligi isbotlangan. Neyron tarmoqlarini modellashtirish natijasida, mavjud bo'lgan texnologiya sharoitda amalga oshirish uchun tavsi-ya etilgan rejim parametrlarining maqbul qiymatlari aniqlandi. Mavjud texnologiyani takomillashtirish yo'nalishlari ifodalangan.

Tayanch iboralar: oltin-mis-molibden rudalari, flotatsiya, tasodifiy funktsiyalarning harmonik dekompoziyasi, boyi tishning egri chiziqlari, neyron tarmoq modellashtirish, flotatsiya obyektining ko'p omilligi va noaniqliligi, mis tarkibi va metallni qayta ishlash, natriy sulfidi, ksantogenat, ohak suti, elektrokimyoviy salohiyat.

A new approach to the analysis of the existing flotation technology is considered on the example of the Mikheev con-centrator processing gold-copper-molybdenum ores. The scientific approach to the analysis of the industrial flotation process is based on the methodology of ionometric mapping and neural network modeling. The reasons for the observed large variance of technological indices are associ-ated with the use of calcareous technology, which causes losses of gold and molybdenum in the tailings of flotation. Using the method of harmonic analysis, it is proved that the sample is not homogeneous and the observations under investigation are generated by two "superimposed" distributions. The illegitimacy of attempts to supply sodium sulfide only to one point in the head of the process according to the technological process adopted at the factory is proved. As a result of neural network simulation, optimal values of the regime parameters were found, which are recom-mended for implementation in the conditions of the current technology. Directions for improving the existing technology are formulated.

Key words: gold- copper-molybdenum ores, flotation, harmonic decomposition of random functions, enrichment curves, neural network modeling, multifactority and non-linearity of the flotation object, copper content and metal recovery, sodium sulphide, xanthate, lime milk, electrochemical potential.