Method for estimating disturbances and giving recommendations for improving process performance

An arrangement to estimate disturbances and to give recommendations in order to improve process performance, the arrangement having measurement devices and receiving interfaces to measure variables of a process, and receive process data, and a pre-processing arrangement to pre-process measurement and process data from the measuring devices and the receiving interfaces, wherein the arrangement comprises:

• a first unit to estimate disturbances, and a second unit to form recommendations, which first unit, in order to estimate the disturbances for each disturbance estimation of a parameter of the process, is arranged to receive the pre-processed process and measuring data from a pre-selected group of the variables of the process,
• a third unit to form machine learning prediction values, the machine learning prediction values being outputs of a trained machine learning model formed from the pre-processed measurement and process data, and the first unit is arranged to also receive the machine learning prediction values from a pre-selected group of the machine learning prediction values to estimate disturbances,
• a fourth unit to form explanation values formed from the machine learning prediction values from the third unit, and quantifying, for each machine learning prediction value, a contribution of each input variable to that prediction, from respective pre-selected groups of
• machine learning prediction values and explanation values,
• wherein the first unit is arranged to also receive the explanation values from a pre-selected group of the explanation values to estimate disturbances, normalize the received pre-processed measuring data, the received machine learning prediction values and the received explanation values, operate the normalized data, comprising the normalized pre-processed measurement and process data, the normalized machine learning prediction values and the normalized explanation values, and scale the operated normalized data, an output of the scaling being the disturbance estimation of the parameter of the process,
• wherein the second unit, for each recommendation forming, is arranged to:
• receive the disturbance estimations from a pre-selected group of the outputs of the first unit, map each received disturbance estimation to one of status categories, and form each recommendation utilizing the mapped disturbance estimations, and apply the formed recommendations as setpoints to controllers of the water intensive industrial process.