AMS Stope Shape Optimiser. Ver 5.02 Reference manual / Оптимизатор формы очистного забоя AMS. Справочное руководство версии 5.02

This manual documents the methods for generating stope-shapes using Version 5 of the Stope Shape Optimiser developed by Alford Mining Systems.

The Stope Shape Optimiser engine (also referred to as "SSO" in this manual) is available in commercial products that are marketed and supported by the mining software suppliers - Datamine, Maptek and Deswik have an established client base and several new suppliers have been added with the release of Version 3 (Hexagon, Geovia, RPM Global). The look and feel of the Stope Shape Optimiser product on each vendor platform will conform to the interface standards and functionality available in each environment. Additional capabilities may have been added by the vendor to fully integrate the Stope Shape Optimiser with the proprietary platform.

This manual does not describe the user interface, data conventions, proprietary input and output file formats, or mechanisms for maintaining scenarios and processing capabilities for single or multiple scenario execution. For further information on these supplier package specific topics, please consult the User and/or Training manuals supplied by your preferred vendor.

This reference manual aims to explain the optimisation techniques and clarify the input parameters for the optimisation methods and provide guidelines for effective and efficient processing of data sets.

In some cases, the XML parameter file formats will be referenced to highlight the options available. For a complete description of the XML format and parameters used, consult the separate document “Version 5 Parameter Summary”. The parameter summary will be of more interest to interface builders and expert users who manipulate XML parameter files in a text editor.

With Version 5 a very significant improvement in run times has been achieved by refactoring the StopeOpt engine as multi-threaded. Performance improvement can be a factor of 3, 5, 10, even up to 20, but it is very dependent of the data set, model size, parameters selected, and the hardware available.