Method and device for efficient fault analysis through simulated faults in a digital twin

Patent metadata

Field	Value
Publication number	US-11567489-B2
Application number	US-202017638622-A
Country	US
Kind code	B2
Filing date	Aug 4, 2020
Priority date	Aug 28, 2019
Publication date	Jan 31, 2023
Grant date	Jan 31, 2023

Abstract

A method, device for identifying causes of faults in automated systems and an automated system which forms the device for identifying causes of faults in automated systems, wherein within a digital twin of the automated system, at least one element of the digital twin is assumed to be faulty and then simulated using the digital twin until a fault time, and where at least one faulty element of the automated system is identified as the cause of a fault based on the at least one element assumed to be faulty.

First claim

The invention claimed is:

1. A method for identifying causes of faults in automated installations, the method comprising:
a) capturing continuously a real installation state of an automated installation, the real installation state comprising at least one physical variable of the automated installation;
b) checking continuously whether a fault in the captured real installation states exists and, if the fault exists in the captured real installation states, determining a most recent fault-free installation state in the captured real installation states and determining a fault time;
c) initializing a digital twin of the automated installation with the determined most recent fault-free installation state;
d) simulating the installation state via the initialized digital twin until the determined fault time, at least one element of the digital twin being assumed faulty; and
e) comparing the simulated installation state at the determined fault time with the real installation state at the fault time and one of e1) identifying (5.1) at least one faulty element of the automated installation as the cause of the fault in the automated installation on the basis of the at least one element of the digital twin that is assumed faulty if the simulated installation state at the fault time matches the real installation state at the fault time or e2) repeating (5.2) steps c) to e) if the simulated installation state at the fault time fails to match the real installation state at the fault time, at least one other element of the digital twin being assumed to be faulty during step d).

2. The method as claimed in claim 1 , wherein a type of fault from a plurality of different types of faults is assumed for the at least one element assumed to be faulty in step d).

3. The method as claimed in claim 2 , wherein at least one of (i) a one-off fault, (ii) a temporarily occurring fault and (iii) a permanent fault is assumed for the at least one element assumed to be faulty in step d).

4. The method as claimed in claim 1 , wherein a failure probability for at least one of (i) the at least one identified faulty element and (ii) the automated installation is derived from the at least one identified faulty element in sub step e1).

5. The method as claimed in claim 1 , wherein said determining the most recent fault-free installation state in step b) comprises:
b1) selecting a provisional most recent fault-free installation state in the captured real installation states;
b2) initializing the digital twin with the selected provisional fault-free installation state provisionally;
b3) simulating the installation state provisionally via the provisionally initialized digital twin until the determined fault time; and
b4) checking whether no fault in the provisionally simulated installation state exists, substeps b1) to b4) being reperformed if a fault in the provisionally simulated installation state exists and an earlier real installation state being provisionally selected from the captured real installation states as the previously provisionally selected installation state.

6. An apparatus for identifying causes of faults in automated installations, comprising:
at least one sensor configured to continuously capture at least one physical variable of the automated installation; and
a controller communicatively connected to the at least one sensor, the controller being configured to:
capture a real installation state of an automated installation continuously, the real installation state comprising the at least one physical variable captured by the at least one sensor;
check continuously whether a fault exist in the captured real installation states and, if there a fault exists in the captured real installation states, determine a most recent fault-free installation state in the captured real installation states and determine a fault time;
initialize a digital twin of the automated installation with the determined most recent fault-free installation state;
simulate the installation state via the initialized digital twin until the determined fault time, at least one element of the digital twin being assumed faulty;
compare the simulated installation state at the determined fault time with the real installation state at the fault time; and
at least one of (i) identify at least one faulty element of the automated installation as the cause of the fault in the automated installation based on the at least one element of the digital twin assumed to be faulty if the simulated installation state at the fault time matches the real installation state at the fault time and (ii) re-initialize the digital twin, simulate the installation state and compare the simulated installation state with the real installation state if the simulated installation state at the fault fails to match the real installation state at the fault time, at least one other element of the digital twin being assumed faulty.

7. The apparatus as claimed in claim 6 , wherein the control is further configured to assume a type of fault from a plurality of different types of faults for the at least one element assumed to be faulty during said simulation of the installation state.

8. The apparatus as claimed in claim 7 , wherein the controller is further configured to assume at least one of (i) a one-off fault, (ii) a temporarily occurring fault and (iii) a permanent fault for the at least one element assumed to be faulty during said simulation of the installation state.

9. The apparatus as claimed in claim 7 , wherein the controller is further configured to derive a failure probability for at least one of (i) the at least one identified faulty element and (ii) the automated installation from the at least one identified faulty element during said identification of at least one faulty element of the automated installation as the cause of the fault in the automated installation.

10. The apparatus as claimed in claim 7 , wherein controller is further configured, during said continuously checking to determinate the most recent fault-free installation state, to:
b1) select a provisional most recent fault-free installation state in the captured real installation states;
b2) initialize the digital twin with the selected provisional fault-free installation state provisionally;
b3) simulate the installation state provisionally via the provisionally initialized digital twin until the determined fault time; and
b4) check whether no fault in the provisionally simulated installation state exists, said select, initialize, simulation and checking of b1) to b4) being respectively reperformed if a fault in the provisionally simulated installation state exists and an earlier real installation state being provisionally selected from the captured real installation states as the previously provisionally selected installation state.

11. The apparatus as claimed in claim 6 , wherein the apparatus is implemented in an automated installation.

Assignees

• Siemens Ag

Inventors

• Baudisch Thomas
• Metzner Maximilian
• Montrone Francesco
• Peschke Jörn
• Rossgoderer Ulrich

Classifications

• G05B19/4183

  characterised by data acquisition, e.g. workpiece identification · CPC title

• G05B23/0275 Primary

  Fault isolation and identification, e.g. classify fault; estimate cause or root of failure · CPC title

• G05B19/41885

  characterised by modeling, simulation of the manufacturing system · CPC title