Smart sensor continuously adapting to a data stream in real time using both permanent and temporary knowledge bases to recognize sensor measurements
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06F-017/00
G06F-003/08
G06N-005/00
출원번호
UP-0362595
(2006-02-25)
등록번호
US-7624080
(2009-12-02)
발명자
/ 주소
Morales, Miguel A.
Haas, David J.
출원인 / 주소
The United States of America as represented by the Secretary of the Navy
대리인 / 주소
Kaiser, Howard
인용정보
피인용 횟수 :
6인용 특허 :
40
초록▼
As the present invention's adaptation process is typically practiced, an observation made by one or more sensing element(s) is classified as being either recognized or unrecognized in the context of a knowledge base. If the observation is classified as being recognized and consistent, then the obse
As the present invention's adaptation process is typically practiced, an observation made by one or more sensing element(s) is classified as being either recognized or unrecognized in the context of a knowledge base. If the observation is classified as being recognized and consistent, then the observation is assimilated into the knowledge base; otherwise, it is not assimilated. If the observation is classified as being unrecognized, then the observation is classified as being uncorroborated in the context of the knowledge base. Prior to being classified as being uncorroborated, the unrecognized observation is categorized in the context of the knowledge base and is associated with an outcome in terms of relationship between/among physical parameters. At the time that corroboration is determined, the observation (originally unrecognized) and its categorization-related and association-related information are assimilated into the knowledge base.
대표청구항▼
What is claimed is: 1. A method for effecting smart sensing, the method comprising: making a series of sensor measurements constituting a data stream in real time; and continually determining, in real time, useful collections of values relating to said sensor measurements; said continually determin
What is claimed is: 1. A method for effecting smart sensing, the method comprising: making a series of sensor measurements constituting a data stream in real time; and continually determining, in real time, useful collections of values relating to said sensor measurements; said continually determining being performed so as to recognize and assimilate new information, corroborate previous information, and identify inconsistencies in said data stream; said continually determining including classifying each current said sensor measurement as being either recognized or unrecognized in the context of a permanent knowledge base; wherein if the current said sensor measurement is classified as being recognized, then said continually determining further includes determining whether an outcome associated with the recognized current said sensor measurement is consistent with an outcome of at least one other said sensor measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition; wherein if said outcome associated with the recognized current said sensor measurement is consistent with an outcome of at least one other said sensor measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, then said continually determining further includes assimilating, into said permanent knowledge base, information including the recognized current said sensor measurement and at least one characteristic of the recognized current said sensor measurement; wherein if the current said sensor measurement is classified as being unrecognized, then said continually determining further includes categorizing the unrecognized current said sensor measurement in a hierarchal structure in a temporary knowledge base, associating the unrecognized current said sensor measurement with an outcome, and determining whether the unrecognized current said sensor measurement is subsequently corroborated by at least one future said sensor measurement; wherein if the unrecognized current said sensor measurement is determined to be corroborated by at least one future said sensor measurement, then said continually determining further includes assimilating, into said permanent knowledge base, information including the corroborated current said sensor measurement and at least one characteristic of the corroborated current said sensor measurement; and wherein if the current said sensor measurement is classified as being unrecognized, then: the unrecognized current said sensor measurement is made during a current event run; said at least one future said sensor measurement is made during at least one future said event run. 2. The method for effecting smart sensing as recited in claim 1, wherein if the current said sensor measurement is classified as being unrecognized, said continually determining further includes placing the unrecognized current said sensor measurement in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 3. The method for effecting smart sensing as recited in claim 1, wherein said corroboration includes: recognition of the unrecognized current said sensor measurement in view of at least one future said sensor measurement; and consistency of said outcome associated with the unrecognized current said sensor measurement with an outcome of at least one future said sensor measurement. 4. The method for effecting smart sensing as recited in claim 3, wherein if the current said sensor measurement is classified as being unrecognized, said continually determining further includes placing the unrecognized current said sensor measurement in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 5. A smart sensing device comprising at least one sensing element and a machine having a memory, said machine containing a data representation pertaining to observations performed via said at least one sensing element, said observations constituting a data stream in real time, said data representation continually adapting, in real time, to recognize and assimilate new information, to corroborate previous information, and to identify inconsistencies in said data stream, said data representation thereby continually determining, in real time, useful collections of values relating to said at least one sensing element, said data representation being generated, in real time, for availability for containment by said machine, by the method comprising: classifying a current said observation as being either recognized or unrecognized in the context of a permanent knowledge base; if the current said observation is classified as being recognized, determining whether an outcome associated with the recognized current said observation is consistent with an outcome of at least one other said observation that exists in said permanent knowledge base and that constitutes a basis for said recognition; if said outcome associated with the recognized current said observation is consistent with an outcome of at least one other said observation that exists in said permanent knowledge base and that constitutes a basis for said recognition, assimilating, into said permanent knowledge base, information that includes the recognized current said observation and at least one characteristic of the recognized current said observation; if the current said observation is classified as being unrecognized, categorizing the unrecognized current said observation in a hierarchal structure in a temporary knowledge base, associating the unrecognized current said observation with an outcome, and determining whether the unrecognized current said observation is subsequently corroborated by at least one future said observation; if the unrecognized current said observation is determined to be corroborated, assimilating, into said permanent knowledge base, information that includes the corroborated current said observation and at least one characteristic of the corroborated current said observation; wherein, if the current said observation is classified as being unrecognized, then: the unrecognized current said observation is made during a current event run; said at least one future said observation is made during at least one future said event run. 6. The smart sensing device of claim 5, wherein if the current said observation is classified as being unrecognized, then the unrecognized current said observation is placed in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 7. The smart sensing device of claim 5, wherein said corroboration includes: recognition of the unrecognized current said observation in view of at least one future said observation; and consistency of said outcome associated with the unrecognized current said observation with an outcome of at least one fixture said observation. 8. The smart sensing device of claim 7, wherein if the current said observation is classified as being unrecognized, then the unrecognized current said observation is placed in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 9. A computer program product comprising a computer readable storage medium having computer readable program code portions recorded thereon for enabling a computer, in an ongoing manner, to gain knowledge originating with measurements taken by a sensor, the computer readable program code portions comprising: a first executable portion, for enabling said computer to classify a current measurement as being either recognized or unrecognized in the context of a permanent knowledge base; a second executable portion, for enabling said computer, if the current said measurement is classified as being recognized, to determine whether an outcome associated with the recognized current said measurement is consistent with an outcome of at least one other measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition; a third executable portion, for enabling said computer, if said outcome associated with the recognized current said measurement is consistent with an outcome of at least one other said measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, to assimilate information in said permanent knowledge base, said assimilated information including the recognized current said measurement and at least one characteristic of the recognized current said measurement; a fourth executable portion, for enabling said computer, if the current said measurement is classified as being unrecognized, to categorize the unrecognized current said measurement in a hierarchal structure in a temporary knowledge base, to associate the unrecognized current said measurement with an outcome, and to determine whether the unrecognized current said measurement is subsequently corroborated by at least one fixture said measurement; and a fifth executable portion, for enabling said computer, if the unrecognized current said measurement is determined to be corroborated, to assimilate, into said permanent knowledge base, information including the corroborated current said measurement and at least one characteristic of the corroborated current said measurement; wherein, if the current said measurement is classified as being unrecognized, then: the unrecognized current said measurement is made during a current event run; said at least one future said measurement is made during at least one fixture said event run; and wherein said measurements constitute a data stream in real time, said computer continually adapting, in real time, to recognize and assimilate new information, to corroborate previous information, and to identify inconsistencies in said data stream, said computer thereby continually determining, in real time, useful collections of values relating to said sensor. 10. The computer program product according to claim 9, wherein said computer readable program code portions further include a sixth executable portion, for enabling said computer, if said current measurement is classified as being unrecognized, to place the unrecognized current said measurement in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 11. The computer program product according to claim 9, said corroboration including: recognition of the unrecognized current said measurement in view of at least one future said measurement; and consistency of said outcome associated with the unrecognized current said measurement with an outcome of at least one fixture said measurement. 12. The computer program product according to claim 11, wherein said computer readable program code portions further include a sixth executable portion, for enabling said computer, if the current said measurement is classified as being unrecognized, to place the unrecognized current said measurement in provisional storage at least until the conclusion of the current said event run, said categorizing being performed after the conclusion of the current said event run. 13. A method for effecting smart sensing, the method comprising: making a series of sensor measurements constituting a data steam in real time; and continually determining, in real time, useful collections of values relating to said sensor measurements, said continually determining being performed so as to recognize and assimilate new information, corroborate previous information, and identify inconsistencies in said data stream; said continually determining including classifying each current said sensor measurement as being either recognized or unrecognized in the context of a permanent knowledge base: wherein if the current said sensor measurement is classified as being recognized, then said continually determining further includes determining whether an outcome associated with the recognized current said sensor measurement is consistent with an outcome of at least one other said sensor measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition; wherein if said outcome associated with the recognized current said sensor measurement is consistent with an outcome of at least one other said sensor measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, then said continually determining further includes assimilating, into said permanent knowledge base, information including the recognized current said sensor measurement and at least one characteristic of the recognized current said sensor measurement; wherein if the current said sensor measurement is classified as being unrecognized, then said continually determining further includes categorizing the unrecognized current said sensor measurement in a hierarchal structure in a temporary knowledge base, associating the unrecognized current said sensor measurement with an outcome, and determining whether the unrecognized current said sensor measurement is subsequently corroborated by at least one future said sensor measurement; wherein if the unrecognized current said sensor measurement is determined to be corroborated by at least one future said sensor measurement, then said continually determining further includes assimilating, into said permanent knowledge base, information including the corroborated current said sensor measurement and at least one characteristic of the corroborated current said sensor measurement; and wherein if said outcome associated with the recognized current said sensor measurement is not consistent with any outcome of at least one other said sensor measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, then said continually determining further includes flagging, but not assimilating into said permanent knowledge base, the recognized current said sensor measurement. 14. A computer program product comprising a computer readable storage medium having computer readable program code portions recorded thereon for enabling a computer, in an ongoing manner, to gain knowledge originating with measurements taken by a sensor, computer readable program code portions comprising: a first executable portion, for enabling said computer to classify a current measurement as being either recognized or unrecognized in the context of a permanent knowledge base; a second executable portion, for enabling said computer, if the current said measurement is classified as being recognized, to determine whether an outcome associated with the recognized current said measurement is consistent with an outcome of at least one other said measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition; a third executable portion, for enabling said computer, if said outcome associated with the recognized current said measurement is consistent with an outcome of at least one other said measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, to assimilate information in said permanent knowledge base, said assimilated information including the recognized current said measurement and at least one characteristic of the recognized current said measurement; a fourth executable portion, for enabling said computer, if the current said measurement is classified as being unrecognized, to categorize the unrecognized current said measurement in a hierarchal structure in a temporary knowledge base, to associate the unrecognized current ad measurement with an outcome, and to determine whether the unrecognized current said measurement is subsequently corroborated by at least one future said measurement; a fifth executable portion, for enabling said computer, if the unrecognized current said measurement is determined to be corroborated, to assimilate, into said permanent knowledge base, information including the corroborated current said measurement and at least one characteristic of the corroborated current said measurement; and a sixth executable portion, for enabling said computer, if said outcome associated with the recognized current said measurement is not consistent with any outcome of at least one other said measurement that exists in said permanent knowledge base and that constitutes a basis for said recognition, to flag, but not assimilate into said permanent knowledge base, the recognized current said measurement; wherein said measurements constitute a data stream in real time, said computer continually adapting in real time, to recognize and assimilate new information, to corroborate previous information, and to identify inconsistencies in said data stream, said computer thereby continually determining, in real time, useful collections of values relating to said sensor.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (40)
Chen Bor-Dong (Dearborn MI), Adaptive air/fuel ratio control method.
Lee Ronald A. (Brownsburg IN) McCoy William L. (Mooresville IN) McQuinn Ted M. (Carmel IN), Adaptively calibrated sensing mechanism for an engine demand device.
Chandy Ashok ; Pattok Kathryn Lynn ; Chhaya Deepak, Apparatus and method for producing a desired return torque in a vehicle power steering system having a rotational steering position sensor.
Lind, Michael A.; Priddy, Kevin L.; Morgan, Gary B.; Griffin, Jeffrey W.; Ridgway, Richard W.; Stein, Steven L., Application specific intelligent microsensors.
Hoffberg Steven Mark ; Hoffberg-Borghesani Linda Irene, Ergonomic man-machine interface incorporating adaptive pattern recognition based control system.
Anastasio, Thomas J.; Ray, Sylvian R.; Huang, Thomas S., Method for determination of spatial target probability using a model of multisensory processing by the brain.
YangQuan Chen SG; LeeLing Tan SG; KianKeong Ooi SG; Qiang Bi SG; KokHiang Cheong SG, Repeatable runout compensation using a learning algorithm with scheduled parameters.
Dickinson John E. ; Carson Douglass L. ; Salman Mutasim Abdurrahman ; Pastor Stephen Robert ; Tierney Gordon Leo ; Hu Hong Xing ; Brown Todd ; Lee Alan James ; Leaphart Eldon Gerrald, Sensor-responsive control method and apparatus.
Rose-Pehrsson Susan L. (Alexandria VA) Di Lella Daniel (Lorton VA) Grate Jay W. (West Richland WA), Smart sensor system and method using a surface acoustic wave vapor sensor array and pattern recognition for selective tr.
Lipchak John B. (Forest Hills PA) Oblak Tod A. (Belle Vernon PA) Bednar Fred H. (Millvale PA) Ciaramitaro William (Murrysville PA) Hughes Francis R. (Verona PA) Smith John R. (Monroeville PA) Gisoni , Smart sensor system for diagnostic monitoring.
Jolly, Mark R.; Fuhrer, Zachary; Fogle, Robert E.; Catanzarite, David M., Aircraft rotary wing motion control and instrumented motion control fluid device.
Gagnon, Ghyslain; Bouchard, Tommy; Tremblay-Munger, Olivier; Paquette, Remi; Gagnon, Francois, Energy consumption evaluation system having reduced energy consumption.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.