Analysis method and devices for same

1. A method for predicting process deviations in an industrial-method plant, the method comprising:

automatically generating a prediction model, wherein, for generating the prediction model, process values and/or status variables measured by a sensor are stored during operation of the industrial-method plant for a predetermined period, and wherein the predetermined period for which process values and/or status variables are stored during operation of the industrial-method plant is predetermined in dependence at least one of:

(i) the industrial-method plant is in an operation-ready state, in particular for a production operation, for at least 60% of the predetermined period,

(ii) the industrial-method plant is in a production-ready state for at least 60% of the predetermined period,

(iii) a predetermined number of process deviations and/or disruptions in the predetermined period; and

predicting process deviations during operation of the industrial-method plant, using the prediction model, wherein the method for predicting process deviations is carried out in an industrial supply air plant, a pre-treatment station, a station for cathodic dip coating and/or a drying station.

2. The method according to claim 1 , wherein process deviations of production-critical process values in the industrial-method plant are predicted by the prediction model, on the basis of changing process values during operation of the industrial-method plant.

3. The method according to claim 1 , wherein for generating the prediction model, a machine learning method is utilized, and wherein the process values and/or status variables that are stored for the predetermined period are used for generating the prediction model.

4. The method according to claim 3 , wherein the machine learning method is carried out on the basis of features that are extracted from the process values and/or status variables stored for the predetermined period.

5. The method according to claim 4 , wherein one or more of the following is used for extracting features:

statistical key figures;

coefficients from a principal component analysis;

linear regression coefficients; and

dominant frequencies and/or amplitudes from a Fourier spectrum.

6. The method according to claim 1 , wherein a selected number of prediction data sets with process deviations and a selected number of prediction data sets with no process deviations are used for training the prediction model.

7. The method according to claim 6 , wherein selection of the number of prediction data sets with a process deviation is made on the basis of one or more of:

a minimum time interval between two prediction data sets with process deviations;

an automatic selection on the basis of defined rules; and

a selection by a user.

8. The method according to claim 6 , wherein prediction data sets with process deviations are characterised as such if a process deviation occurs within a predetermined time interval.

9. The method according to claim 8 , wherein the process values and/or status variables that are stored for the predetermined period are grouped into prediction data sets by pre-processing.

10. The method according to claim 9 , wherein the pre-processing includes the following:

regularisation of the process values stored for the predetermined period; and

grouping the process values and/or status variables into prediction data sets by classifying the process values and/or status variables into time windows with a time offset.

11. The method according to claim 1 , further including displaying or providing the process deviations.

12. The method according to claim 11 , wherein the process deviations are displayed on a diagnostic window.

13. The method according to claim 1 , further including:

determining a fault of the industrial-method plant based on the process deviations; and

indicating the fault on a diagnostic window.

14. The method according to claim 1 , wherein generation of the prediction model includes:

determining a time series into time frames, and

determining key variables of the time frames.

15. The method according to claim 1 , further including: determining a fault cause; and determining relevant process values to be associated with the determined fault cause.

16. A prediction system for predicting process deviations in an industrial-method plant, wherein the prediction system takes a form and is constructed for carrying out the method for predicting process deviations in an industrial-method plant, according to claim 1 .

17. An industrial control system that includes the prediction according to claim 16 .

18. A system for predicting process deviations in an industrial-method plant, the system comprising:

a sensor to measure a process value; and

an industrial controller to:

generate a prediction model, wherein to generate the prediction model, process values and/or status variables are stored during operation of the industrial-method plant for a predetermined period, and wherein the predetermined period for which process values and/or status variables are stored during operation of the industrial-method plant is predetermined in dependence at least one of:

(i) the industrial-method plant is in an operation-ready state, for a production operation, of at least 60% of the predetermined period,

(ii) the industrial-method plant is in a production-ready state for at least 60% of the predetermined period,

(iii) a predetermined number of process deviations and/or disruptions in the predetermined period; and

predict, based on the process value, process deviations during operation of the industrial-method plant, using the prediction model, wherein the system for predicting process deviations is carried out in an industrial supply air plant, a pre-treatment station, a station for cathodic dip coating and/or a drying station.

19. The system according to claim 18 , wherein the generated prediction model corresponds to an occurrence probability of respective process values.