Multi dimensional statistics industrial process control technique takes control parameter outputs and uses multi statistical matrix techniques deriving parameters within/outside statistical functions and outputting changes

摘要

The process takes the output of control parameters and points of observation, transferring the values obtained such that they are compatible with a multidimensional gaussian distribution. The transformed observations are placed in multidimensional space, associating dimension and value magnitude. The observations under control, and out of control can then be separated, and observation points out of control investigated to find the direction from which the regulation is incorrect, providing possible anomaly causes, switching to correct derivatives.