Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – Vehicle diagnosis or maintenance indication
Reexamination Certificate
2000-03-31
2002-11-12
Cuchlinski, Jr., William A. (Department: 3661)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
Vehicle diagnosis or maintenance indication
C701S013000, C701S014000, C701S035000
Reexamination Certificate
active
06480770
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates generally to a system for analyzing data of a repetitive occurrence, such as aircraft flights. More particularly, the invention relates to a system for determining atypical flight characteristics and for using these atypical flight characteristics to determine pre-cursors to certain events, the risk of the occurrence of an event, and the consequences resulting from a certain event. Mitigating factors that avoid the occurrence of an event, or the consequences of an event, are also determined. The information obtained in the PAR (Pre-cursors, Atypicality, and Risk analysis) system is used to identify aircraft flights that are at risk of being in an accident or incident, or of having some other phenomenon occur.
In accordance with the present invention, in a defined set of flights, or domain, an atypicality vector is determined for each flight by evaluating parameters of a flight for deviation from the norm (atypicality). For a set of flights within the domain having a certain “exceedance event,” i.e., a detected violation of standard operating procedure, common vector components of the atypicality vectors indicate pre-cursors to the exceedance event. Flights that most closely resemble the precursor pattern are at a higher risk of encountering the exceedance event. Mitigating factors are determined by examining the atypicality vectors of flights closely matching the pre-cursor pattern but not encountering the exceedance event. Consequences of an event are determined by evaluating the portion of the atypicality vector corresponding to measurements made after the occurrence of the event. Mitigating factors for the consequences are determined by examining the atypicality vectors of flights closely matching the consequences pattern but not encountering the consequences.
2. Description of the Related Art
Efforts to identify flights that are at risk of being in an accident or incident are known. In analyzing data for repetitive occurrences, for example aircraft flights, pre-defined criteria for judging a flight are normally used to detect an “exceedance event.” However, those criteria frequently do not measure all hazards to the successful completion of the flight.
A known technique to help identify unknown hazards to the successful completion of a flight is to identify flights that are atypical. These atypical flights have characteristics that may not be obvious, but which separate them from the majority of typical flights. Once atypical flights are identified, it is standard practice to identify factors that made those flights atypical, and to decide if those factors are symptoms of a hazard to flight or are symptoms of some other phenomenon. Other flights that are at risk can be identified by using those symptoms.
Singular value decomposition (SVD) and principal axis analysis are techniques that are currently known in the art to evaluate data in relation to the predefined criteria discussed above. These techniques can find data points that are within the normal range of individual tests but that are outliers when viewed with an appropriate coordinate transformation. However, the present state of the art has numerous limitations. While these techniques are able to determine if data is different, they are not able to tell you why the difference occurred or whether the difference matters. The techniques mentioned above are also slow, expensive, and tedious. For example, singular value decomposition requires the collaboration of a research scientist and a knowledge domain expert to identify resources in both a knowledge domain and an analysis domain and to interpret the data. The process requires time to format the data for analysis and to manually iterate, generate, identify, and classify phenomena as benign or dangerous. The results may then require further manual transcription. Additionally, SVD is expensive because of the amount of expertise, time, and computational power required.
SUMMARY OF THE INVENTION
It is a general object of the present invention to provide a system for analyzing data of a repetitive occurrence that overcomes the disadvantages of the conventional methods.
It is another object of the present invention to provide a system for analyzing data of a repetitive occurrence that determines atypical data within a given domain.
It is still another object of the present invention to provide a system for analyzing data of a repetitive occurrence where atypical data is used to determine pre-cursors to exceedance events.
It is yet another object of the present invention to provide a system for analyzing data of a repetitive occurrence where the risk of an exceedance event is determined by analyzing how closely a given pattern matches the pre-cursor pattern.
It is yet another object of the present invention to provide a system for analyzing data of a repetitive occurrence where mitigating factors are determined by analyzing data sets that had pre-cursors to an exceedance event but in which the exceedance event did not occur.
It is another object of the present invention to provide a system for analyzing data of a repetitive occurrence where the consequences of an event are determined by analyzing the atypicality vectors of data sets after the occurrence of the event.
It is yet another object of the present invention to provide a system for analyzing data of a repetitive occurrence where mitigating factors are determined by analyzing data sets that had an event occurrence but in which specific consequences of that event did not occur.
In accordance with the objects described above, one aspect of the present invention includes a system for analyzing aircraft data that includes the steps of identifying a domain comprising sets of data, calculating ranges of typical values for components of the sets of data within the domain, and determining atypical components or each set of data within the sets of data based on the typical values calculated in the calculating step. In this aspect, the domain may be limited by identifying specific sets of data within the domain that relate to a specific event.
In another aspect of the present invention, the system for analyzing aircraft data can further include the step of determining atypical sets of data by (a) calculating a weight of the atypical components for each set of data within the sets of data, and (b) comparing the weight of each set of data to a distribution of the weights of all of the sets of data within the domain. A set of data is atypical if its weight is high with respect to the distribution of the weights of the sets of data within the domain.
In yet another aspect of the present invention, the system for analyzing aircraft data can also include the step of determining a pre-cursor pattern to an exceedance event by (a) identifying exceedance sets of data within the domain corresponding to an exceedance event, and (b) identifying common atypicalities of the exceedance sets of data. In this aspect, all components of the exceedance sets of data occurring after the exceedance event are ignored.
In still another aspect of the present invention, the system for analyzing aircraft data can further include the step of determining the risk of the occurrence of the exceedance event for a risky set of data within the domain by comparing atypicalities of the risky set of data with the pre-cursor pattern determined in the step of determining a pre-cursor pattern. In this aspect, all components of the exceedance sets of data occurring after the exceedance event are ignored, and the risk is proportional to the amount of correlation between the atypicalities of the risky set of data and the pre-cursor pattern.
In still another aspect of the present invention, the system for analyzing aircraft data can include the additional step of determining mitigating factors for an exceedance event by (a) identifying non-event sets of data within the domain in which atypicalities resembling the pre-cursor pattern exist but in which the exceedance event does not occur, and (b) id
Cuchlinski Jr. William A.
Honeywell International , Inc.
Marc-Coleman Marthe Y.
LandOfFree
Par system for analyzing aircraft flight data does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Par system for analyzing aircraft flight data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Par system for analyzing aircraft flight data will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2985885