Developing an Agent Based Model for Underground Dispatching


Fleet management and dispatch systems have been used for open pit mining since the early 1980’s but similar systems do not widely exist for underground mining. There is a need for developing dispatch systems for underground mines, particularly highly automated operations such as block caving. A block cave underground dispatch system must be integrated with the cave management system and daily draw control system. Developing a better dispatch system for mobile fleets is key for improving the efficiency and safety of block cave mining. Agent Based Modeling (ABM) is a powerful modeling technique that has already been applied to several real-world problems with significant successes and promise for other applications. Agent based models are generally robust and very flexible. These models are designed to provide the best solution at the time of decision making and also to react to any changes in the operational environment, such as equipment breakdowns, hang-ups at draw points, or poor ground conditions. ABM models can react to these upset conditions and provide stability to the overall system. Unlike other dispatch systems based on very complicated mathematical algorithms, an agent based dispatch system uses simple functions and queues to describe the complexity of a problem. This paper describes the key parameters of a block cave mining method and presents the agent based dispatch model along with simulation results of a generic block cave mine.