System and Method for Predicting Industrial Equipment Motor Behavior

A computer implemented method for predicting an indicator data value for a motor, the method comprising:

• Defining, by a server, a data collection time period;
• defining an initial period of the data collection time period as zero;
• initializing an input vector for the defined initial period;
• inserting stored indicator data as initial values for the defined initial period;
• predicting the indicator data value at the current time;
• outputting the predicted indicator data value;
• calling the input vector at the current time;
• inputting the input vector into a machine learning model;
• and
• predicting the indicator data value for at least a unit time stamp ahead of the current time.

A computer implemented method for predicting the behavior of a motor, the method comprising:

• using a multiple predictive machine learning model, comprising the following steps:
• calculating, by a computer, a percentage of input values that are correct;
• removing, at random, a plurality of input values corresponding to the percent that are incorrect;
• calculating a mean square prediction error value from the input values not removed;
• calculating a weight of each of a plurality of single predictive machine learning models;
• determining an upper bound and a lower bound of the input values;
• calculating a counter value for each single machine learning model;
• updating the weight of each machine predictive learning model;
• calculating a normal probability;
• updating the upper bound and lower bound of the input values;
• updating the normal probability based on the upper bound and lower bound of the input values;
• and
• calculating the probability of a plurality of operational outcomes of the motor.