Optimizing sensor placement for structural health monitoring

• US 9,964,468 B1
• Filed: 12/08/2014
• Issued: 05/08/2018
• Est. Priority Date: 12/08/2014
• Status: Active Grant

• Current Assignee: Bentley Systems Incorporated
• Original Assignee: Bentley Systems Incorporated
• Inventors: Wu, Zheng Yi, Yi, Xiaohua
• Primary Examiner(s): Tran, Michael
• Application Number:
  US14/563,668
• Time in Patent Office: 1,247 Days
• Field of Search: 702 85, 702182, 702 61
• CPC Class Codes: G01M 5/00 Investigating the elasticity of structures, e.g. deflection of bridges or air-craft wings G01M9/00 takes precedence
• CPC Class Codes: G01M 99/00 Subject matter not provided for in other groups of this subclass
• CPC Class Codes: G01R 21/133 by using digital technique

Abstract

In one example embodiment, an analysis application is used to optimize sensor placement by implementing a two-part optimization solution procedure, involving generating a contribution database, and determining an optimized sensor location set using the contribution database. The optimized sensor location set may indicate locations that maximize coverage of dynamic integrity, which is quantified by as a ratio of detectable damage scenarios to all damage scenarios used by the analysis application.

Claims

1. A method of optimizing sensor placement for structural health monitoring of a structure, comprising:

representing the structure by a finite element (FE) model having a plurality of elements connected at nodes;

generating a contribution database that is stored on a non-transitory electronic device readable medium by;

generating a plurality of damage scenarios representing structural damage to the structure, using a Monte Carlo simulation algorithm which produces random damage scenarios, each random damage scenario involving structural damage to one or more randomly selected elements of the FE model, and

analyzing each damage scenario to determine sensitivity at possible sensor locations;

receiving a user'"'"'s selection of a number of sensors to be applied to the structure;

optimizing placement of the number of sensors to produce an optimized sensor location set that maximizes coverage of dynamic integrity for the number of sensors, wherein dynamic integrity is measured as a ratio of detectable damage scenarios to the plurality of damage scenarios, by;

determining a candidate sensor location set by an optimization module that utilizes a genetic algorithm executing on an electronic device,

computing a performance indicator for the candidate sensor placement solution based on the contribution database, by a sensor placement evaluation module executing on the electronic device or another electronic device, wherein the performance indicator is computed as a ratio of a number of damage scenarios covered by the candidate sensor placement solution to a total number of the plurality of damage scenarios,

using, by the optimization module, the performance indicator as a fitness value for the candidate sensor location set to search for a subsequent sensor location set, and

repeating the determining, computing and using until the optimized sensor location set is produced; and

applying sensors at locations on the structure indicated by the optimized sensor location set.

2. The method of claim 1, wherein the possible sensor locations coincide with nodes of the FE model.

3. The method of claim 1, wherein the generating further comprises:

simulating each damage scenario by changing model parameters to indicate a stiffness reduction.

4. The method of claim 3, wherein the generating further comprises:

in response to the changed model parameters, determining a mode shape difference (MSD) matrix; and

from the mode shape difference matrix, determining a contribution matrix which is stored in the contribution database, the contribution matrix having a plurality of rows and columns which represent damage scenarios and degrees of freedom (DOFs).

5. The method of claim 1, wherein a number damage scenarios of the plurality of damage scenarios is greater than or equal to 1000.

6. The method of claim 1, wherein the number of sensors is less than or equal to 15.

7. The method of claim 1, wherein the sensors comprise accelerometers.

8. A system comprising:

an electronic device having

a display screen;

one or more processors configured to execute executable instructions; and

a memory configured to store a finite element (FE) model that represents a structure, the model having a plurality of elements connected at nodes which coincide with possible sensor locations, and a contribution database describing a plurality of random damage scenarios, each random damage scenario involving structural damage to one or more randomly selected elements of the FE model, and executable instructions for a plurality of software modules that are executable on the one or more processors, the plurality of software modules including;

a user interface module configured to receive a user'"'"'s selection of a number of sensors to be applied to the structure,

a scenario generation module that uses a Monte Carlo algorithm configured to generate the random damage scenarios,

a user interface module configured to receive a user'"'"'s selection of a number of sensors to be applied to the structure,

a sensor placement evaluation module configured to compute performance indicators of candidate sensor location sets based on the contribution database, the performance indicators computed as a ratio of a number of damage scenarios covered by the candidate sensor placement solution to a total number of the random damage scenarios, and

an optimization module that utilizes a genetic algorithm configured to maximize coverage of dynamic integrity for the number of sensors, wherein coverage of dynamic integrity is measured as a ratio of detectable damage scenarios to the plurality of damage scenarios, by determining candidate sensor location sets using the performance indicators as fitness values for the candidate sensor location sets and producing an optimized sensor location set; and

a plurality of sensor applied at locations on the structure indicated by the optimized sensor location set that operate to monitor structural health of the structure.

9. The system of claim 8, wherein the scenario generation module is configured to simulate each damage scenario by changing model parameters to indicate a stiffness reduction.

10. The system of claim 9 wherein the scenario generation module is configured to, in response to the changed model parameters, determine a mode shape difference (MSD) matrix, and, from the mode shape difference matrix, determine a contribution matrix which is stored in the contribution database, the contribution matrix having a plurality of rows and columns which represent damage scenarios and degrees of freedom (DOFs).

11. The system of claim 8, wherein a number damage scenarios of the plurality of damage scenarios is greater than or equal to 1000, and the number of sensors is less than or equal to 15.

12. The system of claim 8, wherein the sensors comprise accelerometers.

13. A method of optimizing sensor placement for structural health monitoring of a structure, comprising:

accessing a finite element (FE) model of a structure;

producing a contribution database by generating a plurality of random damage scenarios representing structural damage to the structure, each random damage scenario involving structural damage to one or more randomly selected elements of the FE model, and analyzing each random damage scenario to determine sensitivity at possible sensor locations;

receiving a user'"'"'s selection of a number of sensors to be applied to the structure;

optimizing placement of the number of sensors to produce an optimized sensor location set that maximize coverage of dynamic integrity for the number of sensors, wherein coverage of dynamic integrity is measured as a ratio of detectable damage scenarios to the plurality of damage scenarios, by determining with a genetic algorithm candidate sensor location sets of possible sensor locations that coincide with nodes of the FE model, computing performance indicators for the candidate sensor location sets based on the contribution database, the performance indicators computed as a ratio of a number of damage scenarios covered by the candidate sensor placement solution to a total number of the random damage scenarios, and using the performance indicators as fitness values to search for subsequent sensor location sets, until the optimized sensor location set is produced;

displaying locations for sensors indicated by the optimized sensor location set; and

applying sensors at locations on the structure indicated by the optimized sensor location set.

14. A method of optimizing sensor placement for structural health monitoring of a structure, comprising:

generating, by an electronic device, a contribution database that is stored on a non-transitory electronic device readable medium by generating a plurality of scenarios, each scenario involving changing one or more model parameters of one or more selected elements of a model of the structure, and analyzing each scenario to determine sensitivity at possible sensor locations;

optimizing placement of a number of sensors on the structure at selected ones of the possible sensor locations maximizing coverage for the number of sensors, wherein coverage is measured as a ratio of detectable scenarios to the plurality of scenarios, the optimizing performed by the electronic device by determining with a genetic algorithm candidate sensor location sets, computing performance indicators for the candidate sensor location sets based on the contribution database, and using the performance indicators as fitness values to search for subsequent sensor location sets, until an optimized sensor location set is produced;

displaying locations for sensors indicated by the optimized sensor location set on a display screen; and

applying sensors at locations on the structure indicated by the optimized sensor location set.

15. The method of claim 14, wherein the sensors are structural health monitoring sensors and the scenarios are damage scenarios.

16. The method of claim 14, wherein the model is a finite element (FE) model.

17. The method of claim 14, wherein the generating is performed using a Monte Carlo simulation algorithm that produces random scenarios, each random scenario involving changing one or more model parameters of one or more randomly selected elements of the model of the structure.