273 Figura 7- Economic Analysis by period It shows in Figure 7 a consistent increase in NPV, indicating the agent is approaching the maximum. The RL agent adjusted extraction and processing based on ore grade and capacity, learning only from its environment. It found both the optimal and other feasible solutions useful for engineering analysis. Cash flow analysis of the optimized mine production schedule. (Top) Period-by-period cash flows with cumulative trajectory showing NPV convergence to $967.85M at 8% discount rate. The Q-Learning agent's policy generates front-loaded cash flows, with 58.7% of total value captured in the first 10 periods. Peak cash flow of $135.05M occurs at t=3, coinciding with maximum copper grade extraction (1.08%). The horizontal dashed line indicates the final NPV benchmark. (Bottom-left) Comparison of undiscounted versus discounted cash flows demonstrates the time-value impact, with later periods contributing progressively less to NPV despite stable production. (Bottom-right) Phase-wise NPV contribution reveals the strategic importance of early extraction: the Early Phase (t=1–10) contributes 58.7% of total NPV, Mid Phase (t=11–25) contributes 31.2%, and Late Phase (t=26–38) contributes only 10.1%, validating the agent's learned policy of prioritizing high-grade material extraction in early periods. With the power of create many mining plans in parallel way, is possible to have a variety of solutions like it shows in Figure 8. This variety of solutions permit us to identify different solutions, risk and opportunities to enhance the strategy of mining and processing.
RkJQdWJsaXNoZXIy MTM0Mzk2