Statistical modeling for lime kiln control

Automatic control is often absent or improperly applied in complex processes with large amounts of disturbance and long response times. The reasons for this stem from the difficulty in identifying the individual process dynamics when many are simultaneously effecting the output and from the frustration that occurs in waiting for results from slow processes. Statistical modeling methods on large samples of pre-collected data can make this task more feasible. In this paper we will examine an application of these methods.