Search Documents

By Ryan Lindmark, Guoxing Gu, Zhenghe Xu, Jacob H. Masllyah

Induction time is often referred to the time required to form a stable three-phase contact between a gas bubble and a flat or droplet surface in a liquid medium. It can be used to quantify the attachment process in aflotation system. Various techniques have been developed to measure this important parameter. Due to its importance in flotation systems and other processes, significant advances in refining the induction time measurement techniques have recently been made. This paper summarizes these recent advances and describes a novel method developed to monitor bubble attachment to a pendant bitumen drop or bitumen-coated flat surface. This visualization-based method offers a higher sensitivity and accuracy than the previously developed apparatuses in the induction time measurement. As an example, the induction time for oxygen bubble-bitumen attachment was measured at a fixed bubble size as a function of temperature in two different aqueous solutions. The results showed that higher process temperature is favorable for oxygen bubble-bitumen attachment and hence bitumen flotation. Addition of calcium to the aqueous solution slightly decreased induction time.

Jan 1, 2005

By R. C. Bryan, D. B. Jamison

This paper describes a procedure for minimizing the operational costs associated with blending sodium in coal at a surface mine in the Powder River Basin. Mine site blending is performed by adjusting daily production rates for high- and low-sodium pits. These adjustments are designed to produce a consistent sodium value called the blending target. The mine's contractual limit for sodium is sufficiently close to the field average value that any decrease in the sodium blending target may result in an opportunity cost, if the number of potentially minable tons is reduced. Conversely, any increase in the blending target may result in a higher sampling cost if more information is required to keep the blend within contractual limits. Blending parameters were defined using a geostatistical block model of sodium content that was based on analyses of hundreds of in-pit sample holes. Variograms for these data were used to compute the marginal costs of opportunity and sampling for a series of hypothetical in-pit drilling patterns. These marginal costs provided a basis for selecting optimum values for key parameters such as the blending target, production variance and in-pit sampling density. The benefit of implementing these optimized parameter values is that operational costs of sodium blending will be minimized over the long-term. At the mine where this procedure was developed, each dollar spent to increase the sampling density to the optimum value was shown to generate a $10 savings in opportunity cost.

Jan 1, 1991

By Davidtz. John C.

INTRODUCTION The effectiveness of collector type on sulfide mineral flotation is generally determined by the amount and type of functional groups of collector molecules that interact with surface and water at the mineral/collector/water/gas interface. The final interfacial state when bubble contact takes place is preceded by chemical and physical reactions that involve these gas/liquid/solid interactions. These interactions differ not only for different minerals, solution and gas composition, but also for different collector molecules. The purpose of this paper is primarily to summarize the developments and status of short (C2-C6) and long (C12) chain, aliphatic collectors that contain three sulfur atoms in their polar function. These are trithiocarbonates (TT'Cs). For purposes of comparison, similarities in molecules containing 0, N and S are compared. Mostly short chain ionic collectors molecules of dithiocarbonates (DTCs) and TT'Cs were compared. The studies, largely bench scale flotation and plant testing, were carried out on ores containing Cu, Ni, Au, and PGM's. Studies of long chain 'ITC collectors are reported on Ni, PGM and auriferous pyrite ores containing uranium. Under starvation dosage conditions, surface activity coefficients of short chain collector molecules and interacting water were calculated. Based on the UNIFAC group contribution method of activity coefficient calculations, the Gibbs Excess Free Energy

Jan 1, 2005

By F. Woodcock, A. R. Laplante

A test to characterize the amount of gravity-recoverable gold (GRG) in an ore is described. Typical GRG results are given and possible uses are discussed. The overall database, which includes 68complete tests, is presented. The ore samples treated had an average GRG content of 63% of the total gold in the sample, with a standard deviation of 20% and a range of 3% to 97%. Both the F8(, of gold and the GRG content at a grind of 100% finer than 850 pm correlate well with the overall GRG content (at 80%finer than 75 pm). The primary use of the test is to determine whether gravity recovery should be used and how. Results can also be used to predict gravity recovery, as many of the ores tested are from plants employing gravity recovery. Gravity-circuit simulation can be used to refine recovery estimates or for optimization studies.

Jan 1, 2000

By George E. Aiken, B. R. Jr. Waples

INTRODUCTION In this paper the author will attempt to bridge from the paper work of design and engineering to the reality of implementation. Certain facets of construction are constant to a successful project, no matter the level of technical complexity. Basic to timely, low cost construction are the people assigned to the project. Construction is not accomplished by robots, computers, or speeches; but rather by people. People do however require the avail- ability of equipment and communications to properly complete the work. ORGANIZATION Good organization is the basic requirement of a well implemented construction project. For this paper the author will suggest an organization chart as well as an activity listing that together support the construction activities. There is little difference in the activities or functions needed to construct a small plant or a large one. It is a matter of how much time an individual has to spend on an activity that determines the structure of an organization. The difference between a large and small plant construction job is often described as the multiple hat versus the single hat philosophy. As an example, a resident construction manager on a small job might be the manager, the field engineer, the scheduler, the estimator, and the procurement manager. At some time during a construction

Jan 1, 1986

By Raymond E. Connell

From a macro business process perspective, if what we do in the Mining Industry is find material, dig it out of the ground, convert it to a product and sell it, we are in what might be considered a "simple business". If it is so simple, where does information technology fit? How does it affect our industry? Is it a strategic tool for our industry? What approach should we be employing to ensure we are utilizing information technology effectively? How is information technology integrated or tied together with other areas in the organization such as process control? What kind of people and skill sets are required to support the whole area of information technology? What effect does information technology and automation have on streamlining and rethinking the way we do business in the 90's. This paper covers how these areas have been addressed within the Noranda Group of Companies. It reviews how information technology has evolved, the approach, and what effect it has had on a diversified natural resources company in the mining and metals, forest products, and oil and gas industries.

Jan 1, 1993

By Euler De Souza, Jason Chew, Aldrich Philip Dirige

Sillmats are structural elements upon which the safety and economics of underfill mining or sill pillar recovery in stopes of moderate width with steeply dipping ore zones, depend. The traditional stability design of cemented sillmats appears to have developed from experience and from standard static physical model tests such as uniaxial and triaxial compression. Such standard static physical model testing approach arrives at some approximate parameters, normally used to describe or predict the behaviour of sillmats in analytical or numerical models. This approach offers inherent limitations and generally results in conservative designs. Centrifuge modeling studies are combined with numerical modeling in this paper to provide insights into sillmat strength behaviour. The results of both modeling techniques suggest that, in stopes with smooth rock wall conditions, sillmat failure is driven by the fill self-weight and has minimum dependency on the fill binder content. In stopes with rough rock wall conditions, however, the wall roughness contributes significantly to the stability of the sillmat during underfill mining. This study indicates that the stability of paste fill sillmats can be designed with confidence using centrifuge physical modeling in combination with numerical modeling.

Jan 1, 2001

At first glance the staid Joralemon background does not seem to equate with the life of adventure that irresistibly called me for over 70 years. But my ancestors had their adventures too. My father's family were Dutch Reformed Protestants who had energy or daring enough to come to America to escape the Catholic persecution that followed the Massacre of St. Bartholemew. The first Joralemon came from Friesland to New Amsterdam m 1636, just after Peter Stuyvesant. My father's mother's family, the Berdans, were French Huguenots who came to northern New Jersey only a little later to escape the Catholic domination in France. Some say that the first Berdan to come to America was given all the land between the Passaic and the Hacken-sack Rivers by King James. Whatever, the Berdan family settled there and was able to make a scanty living by farming and selling their produce in Manhattan. The spirit of adventure in my mother's family was just as tenuous. Gamaliel Beaman was pastor of a shipload of Puritans who came to Massachusetts in 1636. Gamaliel's descendents were peaceful farmers and some ministers in New England and later New York State. They had got all of the adventure out of their systems in the long and dangerous voyage across the Atlantic to escape religious persecution. The Joralemons and the Berdans married into families that had been in America almost as long as they had. On my father's side the Berdans, Doremus, DeGreys, Vanderbeeks, and Varicks showed by their names their Dutch or French Huguenot stock. The only one who achieved any fame was Peter Varick, who was George Washington's military secretary during the American Revolution. After this he was nine times mayor of New York and was one of the founders of Kings College, later Columbia University. He was responsible for one of my few ancestral claims to distinction. When the Marquis de Lafayette made a fare-

Jan 1, 1976

By Peter K. Achireko, Samuel Frimpong

Design and optimization of open pit limits are of paramount importance because they provide information for evaluating the economic potential of a mineral deposit and for developing short and long range mine plans. Many algorithms and their modifications have been used to design and optimize open pit limits. These algorithms have provided mine planning engineers with pertinent information in designing, optimizing and extracting ore reserves by open pit technology. However, they do not address the random field properties associated with the ore grades and reserves and commodity prices, and thus, fail to yield the truly optimized pit limits in any time horizon. In this paper, a new algorithm, MCSIMFNN, which overcomes these limitations is proposed and used to optimize open pit limits. The random field properties of the ore grade and reserves have been modelled using the modified conditional simulation based on the best linear unbiased estimation and local average subdivision techniques. Artificial neural networks are used to classify the blocks into classes based on their conditioned values. The error back propagation algorithm, in the neural networks, is used to optimize the pit limits by minimizing the desired and actual outputs error in a multi- layer perceptron under the pit wall slope constraints. The optimized pit value obtained by this algorithm has been compared to that from the Lerchs-Grossman's algorithm in a case study. The results are the same, but in random multivariable states, the MCSI MFNN algorithm is the most suitable for pit optimization. The method is also fast since it mimics the operation of the brain in the solution of the pit optimization problem.

Jan 1, 1996

By J. Linhart

In vertical airways, falling material - solids-or water - may cause undesirable disturbances in the ventilation in the airway itself and also in neig¬bouring airways unless precautions are taken to stabilize the air flow. Such cases can occur, for example, - in shaft sinking by drilling from a pilot hole, - during the filling up of shafts, and - during the introduction of water into shafts in the case of mine fires. In an airway with upward ventilation the falling material will reduce the air flow and may even re¬verse its direction. In an airway with downward ventilation the air flow can only increase, but a reversal of the direction of ventilation may result in a parallel airway. These disturbances in the normal ventilation are undesirable as it must be the aim to have ventilation that is stable in quantity and direction throughout a mine. In order to be able to prepare suitable stabilization measures one must know the pressure generated by the falling material; this pressure will henceforth be called "aeromotive" pressure. The main factors determining the influence of falling material on the ventilation are - the amount of falling material and its height of fall, - the cross sectional area of the airway, - the pressure. difference prevailing in it, - the-position of the. airway in the ventilation. network, - the air velocity, and - other parameters, e. g. the density and grain size of the falling material. To determine the aeromotive pressure-measurements have been made underground during the sinking and filling of shafts and systematic studies have been carried out during the supply of water into a staple shaft. The present paper deals first with some fundamental theoretical considerations and then describes the measuring arrangements and the results of the series of experiments that have been carried out.

Jan 1, 1980

By Victor F. Hollister

Porphyry copper deposits can be categorized into three models for purposes of prospect evaluation: Breccia-pipe, modified Lowell and Guilbert, and diorite models. The modified Lowell and Guilbert model is normally associated with a quartz-bearing intrusive and a potassic, phyllic, argillic, and propylitic alteration sequence. Variations can occur in the alteration zones of this model that range from orthoclase-rich (Kalamazoo, Silver Bell, and Mineral Park) to ortho- clase-deficient (Petaquilla and Bougainville) potassic zones. The unifying features of the model are the presence of a phyllic zone and a quartz-bearing intrusion. The breccia-pipe model (Toquepala) is characterized by structure and alteration related to or controlled by breccia. Where tourmaline occurs in the breccia, the identity of the model is usually clear. However, if tourmaline is not commonly developed in the breccia (La Caridad), the deposit can be misidentified as to model. Porphyry copper deposits associated with quartz-deficient intrusives, whether these are calc-alkalic diorites or alkalic plutons, are characterized by alteration suites distinct from those of the modified Lowell and Guilbert model. These have been generalized as diorite model porphyry deposits. Because alteration and mineralogical development in diorite model deposits is distinct from that of the modified Lowell and Guilbert, ore guides and exploration tactics differ for the two models. The reprint that follows describing the three prospecting models summarizes their features. It is to be emphasized that these are models used in exploration for specific deposit types and are not genetic, tectonic, or geographic models. Porphyry copper deposits can be metamorphosed. Gibralter is an example of a Lowell and Guilbert model deposit that has undergone regional metamorphism.

Jan 1, 1991

By B. Fitch, E. J. Roberts

Definitions and Terminology Classification is defined within this section as the separation of solid particles into size or weight fractions or classes by differential settling through a fluid. It is usually thought of as a size separation, but is actually made on the basis of particle settling rate, and so is affected by any variations in particle density and shape. The fluid may be liquid (wet classification) or gas (dry or air classification). This section will treat only wet classification. There are two basic types of classification; pool and hydraulic. In pool classification a suspension of feed particles is fed into and out of a pool of some kind at such a rate that only part of them¬ the coarser and faster-settling part-has time to settle out. The remain¬der overflows as fines. In hydraulic classification the feed suspension is introduced into or above one or more columns or pockets through which water is rising at a controlled velocity. Coarse particles subside through the pockets and are removed through spigots. Those which settle slower than the rising velocity in any column or pocket are (at least ideally) prevented from subsiding and so must overflow. Hydraulic classifiers or sizers are capable of making far sharper sepa¬rations than pool classifiers. The force causing settling may be gravity, or alternatively may be from centrifugal acceleration as developed in a centrifuge or hydra cyclone. Two things are needed to characterize the separation produced by classification: first, the size, D„ or equivalently the screen mesh at which it is made; second the sharpness of separation. Separation size may be defined in various ways. In closed-circuit grinding it is commonly specified by some point in the overflow screen analysis: For example, 5% on 65 mesh. In other applications, notably

Jan 1, 1985

By A. J. Basu

A prototype expert system for planning and designing underground coal mines has been developed in a doctoral study (Basu, 1991) at the University of Kentucky (U.K.). The objective of the study was to provide a design tool for configuring layout parameters, production systems, and for determining economic feasibility. Details of knowledge engineering and the prototype knowledge-based system for integrated and intelligent planning and design have been covered in the dissertation. The prototype demonstrates its robustness by allowing interaction among major design components, most notably among ventilation, ground control, and production planning. A plan for testing and evaluating the prototype, an important phase of software development, is presented here. Existing evaluation procedures for expert system prototypes are discussed and the evaluations for the underground coal mine design expert system prototype are included. The initial testing plan used senior-level mining engineering students to test the acceptance of the prototype. This test suggested the usefulness of such a system for training and instruction. This initial acceptance test was performed with the feasibility prototype. The other test objective, i.e., testing the model's accuracy, involved extensive participation of domain experts. For overall accuracy evaluation, an existing underground coal mine design was compared with the output from a design session using the expert system prototype. The testing procedure is not an isolated, independent development phase. Much of the accuracy evaluation is done at the time of knowledge base development. Therefore, the knowledgebase itself does not need a formal test phase. The major purpose of the field testing is to determine probable user acceptance. The acceptance test determines the completeness of the prototype and also will determine efficiency of user interfaces.

Jan 1, 1992

By Peter W. Scott

The Caribbean islands, with their very diverse geology, are a source of several industrial minerals both for their own domestic consumption and for export. The geology of the Greater Antilles, which comprises Cuba, Hispaniola, Puerto Rico, and Jamaica has a Mesozoic or pre-Mesozoic basement overlain by sediments and volcanics, the latter being dominantly island arc in character. Later granitoid intrusions were emplaced. Obduction of ophiolites also occurred variously from the mid- Cretaceous to the Palaeogene. Thick accumulations of limestones surround the older strata, along with Neogene and younger sediments. The Lesser Antilles islands form an arcuate chain, mainly of volcanic rocks of calc-alkali or sub-alkaline affinity with some younger limestones and siliciclastic sediments. Trinidad and Tobago have closer similarities with the South American continent in their geology. Tobago has a metamorphic complex, juxtaposed with a volcanic-plutonic complex, and limestones. Trinidad has very thick Cretaceous to Recent exposed sedimentary sequence. The current exploitation of industrial minerals in the Caribbean is dominated by construction raw materials (mainly sand and gravel, limestone, and cement), although there is extraction of gypsum, silica sand, marble, clay (almost entirely for construction use), pumice, and natural asphalt. Salt is obtained from solar evaporation of seawater. Limestone of high purity and high brightness, gypsum, pumice, and high-density igneous rock have potential for further development of their markets, particularly ex- ports. Potential new and undeveloped industrial-mineral resources include residual and sedimentary kaolin, bentonite, zeolites, magnesite, and talc. Cuba could be a source for many industrial minerals in the future.

Jan 1, 1999

By E. C. Grunsky

Spatial auto- and crosscorrelation assessment of variables can be carried out using quadratic function estimates. From the function estimates, multivariate relationships of the variables can be determined in the form of spatial factors. The calculation of the spatial factors requires computation of the corresponding amplitude vectors from the eigenvalue solution. This vector reflects the relative amplitudes by which the variables follow the spatial factors. Instability of some of the eigenvalue solutions requires that caution be used in interpreting the resulting factor pat- terns. A measure of the predictive power of the spatial factors can be determined from the auto- correlation coefficients and squared multiple correlation co-efficients which indicate which of the variables are significant in any given factor. The use of unconditional simulations provides a basis by which geological variables of varying spatial ranges can be modelled using the spatial factor technique. Simulations were generated for a variety of ranges and variables were created that form single or multiple combinations of the spatial patterns. The success of the spatial factor technique depends on the amount of associated noise, and the size of the neighbourhood relative to the spatial range of the variable. The predictive power of the components is dependent on these two factors. Problems with collinearity of variables can result in unrealistic eigenvalues and result in the failure of the method. Application of this method to spatial data in mineral exploration can assist in outlining spatially continuous zones of alteration associated with mineralization.

Jan 1, 1989

By R. Norman Holme

Sulfonate promoters and collectors were originally developed as relatively inexpensive anionic collectors to selectively concentrate the iron-bearing minerals in low grade iron ores and to remove iron and other mineral contaminants by flotation in the processing of glass sands, feldspar and other ceramic raw materials. The first reagents produced were effective collectors for such iron oxide minerals as goethite, hematite, limonite, magnetite and siderite. Further development has resulted in additional reagents with a broadened application to the selective flotation of a variety of nonsulfide and nonmetallic minerals. The products in this group are anionic sulfonate-type promoters, natural petroleum sulfonates, and synthetic modified sulfonates. Beyond their applications to iron-bearing minerals, the sulfonate collectors have gained commercial use in the flotation treatment of barite, celestite, fluorite, apatite, chromite, industrial garnet, kyanite, mica, rhodochrosite and cassiterite. in addition, extensive laboratory studies indicate their potential suitability for flotation of the following minerals: Anatase Anhydrite Bauxite Beryl Calcite Chlorite Some Copper Carbonates Corundum Dolomite Hornblende Huebnerite Ilmenite Lepidolite Magnesite Phosphate Minerals Rutile Scheelite Spodumene Sillimanite Talc Tourmaline Wolframite Zircon Sulfonate promoters are water soluble or dispersible liquids, syrups or pastes of varying viscosity with colors ranging from dark blackish-brown to clear. Customarily, they are fed to the conditioner or flotation as 5% to 30% water solutions or dispersions. In addition to their normal collector activity, certain of these products display notable frothing characteristics. Improved collector efficiency and adequate froth control can frequently be obtained by employing mixtures of sulfonate reagents, of ten with results quite markedly superior to those achieved with equivalent dosages of the individual promoter alone. Reagent requirements will vary with the type of mineral being treated; however, a general overall range of 0.13 to 2.0 kg/metric ton (0.25 to 4.0 lb/ton) of feed should cover most applications.

Jan 1, 1986

By C. R. Stephan

There are a large number of facilities throughout the world which handle coal, such as preparation plants. Many other facilities use coal as a fuel, such as cement and lime factories. Although coal can be handled safely and can be an efficient fuel, there are explosion hazards which are accentuated as the particle size is reduced. Particle sizes of coal which can fuel a propagating explosion occur within thermal dryers, cyclones, baghouses, pulverized-fuel systems, grinding mills, and other process or conveyance equipment. This paper discusses how explosions can occur within these facilities.

Jan 1, 1998

By Jeffery Denit, Steve Hoffman, Van Housman

The Environmental Protection Agency (EPA) is in the early stages of developing a national mining waste program. Our goal is to develop a mining waste program that will be state implemented, multimedia in nature, flexible, utilize and enhance existing state programs, and protect the environment. The Agency appreciates the cooperation the mining industry is providing through its participation in the Policy Dialogue committee on Mining and by providing the Agency with technical data. The following is a discussion of what led to the development of a national mining waste program, the agenda initiative to evaluate Toxic Substances Control Act authority in phosphate manufacturing, and the important steps being taken through the Policy Dialogue process to frame and discuss issues. EXTRACTION AND BENEFICIATION OF ORES The Resource Conservation and Recovery Act of 1976 (RCRA) and subsequent amendments created a framework for the regulation of hazardous wastes under Subtitle C and nonhazardous waste under Subtitle D. In 1978, initial regulations were proposed for hazardous waste management units under RCRA. These identified a universe of special wastes that were large volume, low hazard and less amenable to Subtitle C control techniques. EPA noted that waste materials from the "extraction, beneficiation and processing of ores and minerals m fit these criteria. In October 1980, Section 3001 of RCRA was amended. This Bevill amendment excluded solid waste from the extraction, beneficiation and processing of ores and minerals from regulatory control under RCRA's hazardous waste Subtitle C. This was done pending a report to Congress and a regulatory determination regarding the Subtitle C status of these wastes. Thus in November 1980, EPA published an interim final amendment to the RCRA Subtitle C regulations to reflect the Bevill amendment. RCRA requires EPA to perform studies and make reports to Congress regarding the management methods and potential hazards of managing waste. RCRA also calls for a regulatory determination regarding whether Subtitle C (hazardous) or D (non- hazardous) controls are warranted for these wastes. In December 1985, EPA submitted a report to congress entitled, “Wastes from the Extraction and Beneficiation of Metallic Ores, Phosphate Rock, Asbestos, Overburden from Uranium Mining and Oil Shale." In that report, EPA found that some mining wastes exhibit hazardous characteristics, that the range of risk from mining waste is broad, and that waste management practices have caused environmental damage. In July 1986, EPA published a regulatory determination based on information gathered in the 1985 report to Congress and comments from interested parties. In the report, EPA concluded that RCRA Subtitle C regulation of extraction and beneficiation wastes was unwarranted because mining wastes tend to be disposed of in arid climates, waste volumes are high, and facilities and wastes are located in sparsely

Jan 1, 1992

By T. M. Karppanen

Shotcrete is used both in mining and tunnelling for supporting rock. Both dry and wet processes are used for shotcreting, and each produces good quality concrete. Thick shotcrete layers function as a concrete arch supporting loose rock. Thin layers prevent relative movements between the blocks of rock due to the fine concrete in the cracks and joints. No progressive fallouts occur since key blocks are kept in place. Rockbolts are often used together with shotcrete in order to create an arch around the tunnel making the rock to support itself. Shotcrete is sprayed either manually or by means of mechanized equipment. The aim is to spray required the layer thickness in one pass and with a minimum amount of rebound. Sprayed layers should also be as even as possible. The tip of the nozzle should be moved in circular movements. The shotcrete should be sprayed at a right angle to the rock surface; the optimum distance from the nozzle to the application surface must be found by experiment. In a mine shotcreting forms a materials handling system which includes batching, transporting and the actual shotcreting. The equipment used vary depending on the spraying method, mining conditions and the location of the mine. Shotcreting is performed with a purpose-designed unit in most cases where operation is mechanized. The shotcreting unit consists of a spraying machine with a hose and a nozzle, an additive system, a manipulator and a mobile carrier with all the necessary controls and accessories. The key element of the unit is the shotcreting manipulator which holds and guides the nozzle as required. The manipulators currently in use are hydraulically operated, and their remote controlled booms are also often called robots. These manipulators have enabled users to increase their output capacities, to remove the operator away from unsafe areas and from the influence of dust and rebound. Furthermore, they improve the shotcrete quality and reduce shotcreting costs.

Jan 1, 1989

By Qiong Wu, Li Guo, Qinghua Gu

BP neural network is a multilayer feedforward network trained by error back-propagation algorithm, which is one of the most widely used neural network models. However, BP neural network has exposed more and more shortcomings and deficiencies with the expansion of the application scope. In the prediction of ore quantity, BP neural network has the characteristics of slow convergence and easy to fall into local minimum point. In order to obtain the global optimal solution, and to improve the defects of BP neural network, this paper proposes combination optimization algorithm of genetic algorithm (GA) and BP neural network to improve the speed and accuracy of forecasting the main design flow chart and the analysis of the sort distinguish algorithm are offered, and then some problem in the design and debugging of the algorithm are discussed. On this basis, the GA-BP neural network model is constructed and applied to optimize the initial weights and threshold value of BP neural network. This model choices the floating point coding method to encode the connection weights and thresholds, and divides subjects into several populations. Through the introduction of selection, mutation, crossover, initial weight and other operators, making operational synergies between the various groups. This study selects 30 geological units, 8 quantitative variables (Pb, Zn, Cu, Mo, Si, Ni, Co, V) and 12 qualitative variables to carry out empirical analysis. Then the simulation of the algorithm is carried out in MATLAB and the parameters are analysed. By normalizing the input samples, 22 groups of observation data are used as the training data for prediction, and the latter 8 groups of observation data are used as the test data to be verified. The results show that when the ore quantity characteristics are not very significant, the model will produce prediction bias. But the improvement of the algorithm increases the efficiency of the function approach capacity of BP neural network and conquer the BP neural network system’s instability. It provides an auxiliary guide for ore prediction, which have higher reference value.

Jan 1, 2017