Identifying true positive data within a set of blast exposure data

Patent metadata

Field	Value
Publication number	US-2025067610-A1
Application number	US-202418944513-A
Country	US
Kind code	A1
Filing date	Nov 12, 2024
Priority date	Nov 9, 2020
Publication date	Feb 27, 2025
Grant date	—

Abstract

Methods, systems, and computer-readable media for identifying true positive data within a set of blast exposure data. An equation fit is applied to generate one or more equations corresponding to portions of pressure data within the set of blast exposure data. The one or more equations are compared to the pressure data to determine if respective portions of the blast exposure data relates to true positive data.

First claim

Having thus described various embodiments of the invention, what is claimed as new and desired to be protected by Letters Patent includes the following:

1 . A method of filtering a set of blast exposure data, the method comprising:
receiving the set of blast exposure data comprising pressure data over time including at least one pressure trace received from one or more blast sensors operable to detect blast pressure associated with a blast exposure; determining a false positive score based on one or more features within the set of blast exposure data; identifying false positive data within the set of blast exposure data based on the false positive score; responsive to identifying the false positive data, removing the false positive data from the set of blast exposure data; applying an equation fit to a remaining portion of the set of blast exposure data to generate one or more equations; comparing one or more equation parameters from the one or more equations to one or more parameters from the pressure data over time; identifying true positive data within the remaining portion of the set of blast exposure data based at least in part on comparing the one or more equation parameters to the one or more parameters from the pressure data over time; and responsive to identifying the true positive data within the remaining portion of the set of blast exposure data, extracting the true positive data from the remaining portion of the set of blast exposure data.

2 . The method of claim 1 , further comprising:
wherein the false positive data is identified by applying a false positive flag to a portion of the set of blast exposure data.

3 . The method of claim 1 , wherein the false positive data is identified based on the false positive score exceeding a predetermined threshold.

4 . The method of claim 1 , further comprising:
prior to determining the false positive score, determining a bias value associated with the pressure data over time; and debiasing the pressure data over time based on the bias value.

5 . The method of claim 4 , further comprising:
prior to determining the false positive score, applying a filter to the pressure data over time to thereby remove noise from the pressure data.

6 . The method of claim 3 , further comprising:
after debiasing the pressure data over time, integrating the pressure data over time to produce impulse data.

7 . The method of claim 1 , further comprising:
identifying the one or more features within the set of blast exposure data using a machine learning algorithm trained with historic blast exposure data.

8 . A method of filtering a set of blast exposure data, the method comprising:
receiving the set of blast exposure data comprising pressure data over time including at least one pressure trace received from one or more blast sensors operable to detect blast pressure associated with a blast exposure; applying an equation fit to the set of blast exposure data to generate one or more equations; comparing one or more equation parameters from the one or more equations to one or more parameters from the pressure data over time; identifying true positive data within the set of blast exposure data based at least in part on comparing the one or more equation parameters to the one or more parameters from the pressure data over time; responsive to identifying the true positive data within the set of blast exposure data, classifying the true positive data within the set of blast exposure data; determining a false positive score based on one or more features within the set of blast exposure data; identifying false positive data within the set of blast exposure data based on the false positive score; and responsive to identifying the false positive data, removing the false positive data from the set of blast exposure data.

9 . The method of claim 8 , further comprising:
deriving the pressure data over time to produce pressure change data.

10 . The method of claim 9 , wherein the one or more features within the set of blast exposure data are identified within the pressure change data.

11 . The method of claim 8 , further comprising:
filtering the set of blast exposure data using a Savitzky-Golay filter.

12 . The method of claim 8 , wherein the equation fit includes a Friedlander waveform fit.

13 . The method of claim 8 , further comprising:
identifying a blast source associated with the blast exposure based on the set of blast exposure data.

14 . The method of claim 8 , further comprising:
determining a plurality of slope values associated with respective portions of the pressure data over time; and identifying the one or more features within the set of blast exposure data based on the plurality of slope values.

15 . A method of filtering a set of blast exposure data, the method comprising:
receiving the set of blast exposure data comprising pressure data over time including at a pressure trace received from one or more blast sensors operable to detect blast pressure associated with a blast exposure; determining a false positive score based on one or more features within the set of blast exposure data, the false positive score associated with at least one predefined false positive class of a plurality of predefined false positive classes; identifying false positive data within the set of blast exposure data based on the false positive score; responsive to identifying the false positive data, removing the false positive data from the set of blast exposure data; applying an equation fit to the set of blast exposure data to generate one or more equations; comparing one or more equation parameters from the one or more equations to one or more parameters from the pressure data over time; identifying true positive data within the set of blast exposure data based at least in part on comparing the one or more equation parameters to the one or more parameters from the pressure data over time; and responsive to identifying the true positive data within the set of blast exposure data, extracting the true positive data from the set of blast exposure data.

16 . The method of claim 15 , wherein the plurality of predefined false positive classes comprises:
a noise class; a physical impossibility class; and a sensor error class.

17 . The method of claim 15 , wherein the false positive score is determined using a machine learning algorithm trained with historical blast exposure data.

18 . The method of claim 17 , further comprising:
retraining the machine learning algorithm based in part on the set of blast exposure data.

19 . The method of claim 15 , further comprising:
determining a bias associated with the pressure data over time by averaging at least a portion of a waveform in the pressure trace.

20 . The method of claim 19 , further comprising:
debiasing the set of blast exposure data by subtracting the bias from the pressure data over time.

Assignees

• Applied Res Associates Inc

Inventors

• Wiri Suthee
• Ortley David J

Classifications

• G01L5/14 Primary

  for measuring the force of explosions; for measuring the energy of projectiles · CPC title

• G06F16/9035

  Filtering based on additional data, e.g. user or group profiles · CPC title

• G06N20/00

  Machine learning · CPC title

• G06F18/2113

  by ranking or filtering the set of features, e.g. using a measure of variance or of feature cross-correlation · CPC title

• A42B3/046 Primary

  Means for detecting hazards or accidents · CPC title