This paper reports on the results of a three year research project into the development of a novel open pit scheduling methodology based on a genetic algorithm (GA) approach. GA systems have been used in a variety of complex optimisation problems with a high degree of success. The AIMS (Artificial Intelligence in the Minerals Sector) Research Unit have been investigating their application to the problem of long term production scheduling for open pit mining systems. The technique developed is based around splitting the scheduling problem into two distinct elements. The first part is concerned with the development of a 3-dimensional genetic algorithm that can be used to evolve a solution to the pit scheduling problem. The second part is concerned with the method of evaluating the fitness of particular solutions. As the basic optimisation engine is de-coupled from the fitness evaluation method it is possible to use the system in a number of different modes such that flexible scheduling can take place. The paper describes the development of the 3-dimensional optimising engine and highlights the major genetic operators that are used to evolve the optimum solution. These operators include; randomisation, mutation, cross-over and local optimisation. The paper also presents the results of studies into selection of appropriate control parameters for each of the genetic operators and suggests ranges for practical control of the GA system. The paper also presents details of a variety of fitness functions which can be used to assess the suitability of a produced solution under different types of constraint. For example it will show how the same system can be used to produce 'optimal' solutions for the same deposit with differing production objectives. Typical objectives include; maximising net present value, minimising early stripping, balancing stripping and balancing ore production for multiple minerals. The paper presents two case studies which illustrate the application of the system and highlight the flexible nature of the approach.
展开▼
