Method and system of monitoring, sensor validation and predictive fault analysis
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G05B-013/02
G01F-001/00
G01F-007/00
출원번호
US-0793162
(2004-03-04)
등록번호
US-7451003
(2008-11-11)
발명자
/ 주소
Chester,Daniel L.
Daniel,Stephen L.
Fickelscherer,Richard J.
Lenz,Douglas H.
출원인 / 주소
Falconeer Technologies LLC
대리인 / 주소
Phillips Lytle LLP
인용정보
피인용 횟수 :
25인용 특허 :
55
초록▼
The present invention provides an improved method and system for real-time monitoring, validation, optimization and predictive fault analysis in a process control system. The invention monitors process operations by continuously analyzing sensor measurements and providing predictive alarms using mod
The present invention provides an improved method and system for real-time monitoring, validation, optimization and predictive fault analysis in a process control system. The invention monitors process operations by continuously analyzing sensor measurements and providing predictive alarms using models of normal process operation and statistical parameters corresponding to normal process data, and generating secondary residual process models. The invention allows for the creation of a fault analyzer directly from linearly independent models of normal process operation, and provides for automatic generation from such process models of linearly dependent process models. Fuzzy logic is used in various fault situations to compute certainty factors to identify faults and/or validate underlying assumptions. In one aspect, the invention includes a real-time sensor data communications bridge module; a state transition logic module; a sensor validation and predictive fault analysis module; and a statistical process control module; wherein each of the modules operates simultaneously.
대표청구항▼
What is claimed is: 1. A method of monitoring, validation and analysis for a process control system having normal process data and sensors, and represented by a plurality of primary process models and a plurality of process variables, comprising: generating a plurality of primary residual process m
What is claimed is: 1. A method of monitoring, validation and analysis for a process control system having normal process data and sensors, and represented by a plurality of primary process models and a plurality of process variables, comprising: generating a plurality of primary residual process models derived from said primary process models, said normal process data, and one or more statistical parameters corresponding to said normal process data; measuring real-time sensor data; computing primary residual values of said primary residual process models corresponding to said real-time sensor data; deriving an expected value for one or more of said primary residual values from historical normal process data; determining deviations of said primary residual values from said corresponding expected values; mapping one or more of said deviations of said primary residual values from said corresponding expected values into one or more compressed values; automatically generating one or more fuzzy logic diagnostic rules; computing a certainty factor for a possible fault by applying said fuzzy logic diagnostic rules to said compressed values; and displaying said certainty factor on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 2. The method as set forth in claim 1, further comprising: measuring said normal process data. 3. The method as set forth in claim 1, further comprising: calculating said statistical parameters. 4. The method as set forth in claim 1, further comprising: computing a certainty factor corresponding to each of said primary residual values. 5. The method as set forth in claim 1 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 6. The method as set forth in claim 1 wherein said fuzzy logic is defined in a diagnostic rule as follows; description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 7. The method as set forth in claim 6, further comprising: determining a direction of deviation of said primary residual values from said expected values; and determining evidence-for-fault and evidence-against-fault by comparing said direction of deviation to an expected direction of deviation consistent with a fault. 8. The method as set forth in claim 6, further comprising: determining neutral-evidence by comparing the magnitude of a deviation of said primary residual values to zero. 9. The method as set forth in claim 1, further comprising: generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common process variable. 10. The method as set forth in claim 9, further comprising: computing secondary residual values of said secondary residual process models corresponding to said real-time sensor data; and comparing said secondary residual values to expected values. 11. The method as set forth in claim 10, further comprising: determining a direction of deviation of said secondary residual values from said expected values; comparing said direction of deviation to an expected direction of deviation consistent with a fault. 12. The method as set forth in claim 11 wherein said determination is made by calculating a first partial derivative of said secondary residual process model. 13. The method as set forth in claim 10, further comprising: computing a certainty factor for a possible fault as a function of one or more of said secondary residual values using fuzzy logic. 14. The method as set forth in claim 13, further comprising: computing a certainty factor corresponding to each of said secondary residual values. 15. The method as set forth in claim 13 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 16. The method as set forth in claim 15, further comprising: determining neutral-evidence from one or more of said secondary residual process models which do not relate to relevant process variables. 17. The method as set forth in claim 9 wherein said secondary residual process models are functions of both measured process variables and unmeasured process variables. 18. The method as set forth in claim 9, further comprising: regenerating said secondary residual process models. 19. The method as set forth in claim 1 wherein said statistical parameters comprise a mean and a standard deviation. 20. The method as set forth in claim 1 wherein said primary residual process models are functions of both a measured process variable and an unmeasured process variable. 21. The method as set forth in claim 1, further comprising: regenerating said primary residual process models. 22. The method as set forth in claim 1, further comprising: computing a first partial derivative and a second partial derivative of said primary residual process models. 23. The method as set forth in claim 1, further comprising: reporting said certainty factor if it exceeds a predetermined threshold. 24. The method as set forth in claim 1 wherein said certainty factor is between 0.0 and 1.0. 25. The method as set forth in claim 1, further comprising: evaluating said real-time sensor data using statistical process control charting techniques to determine if a corresponding sensor is in control. 26. The method as set forth in claim 1, further comprising: calculating the exponentially weighted moving average of said real-time sensor data to determine if a corresponding sensor is in control. 27. The method as set forth in claim 1, further comprising: computing a plurality of certainty factors for a corresponding plurality of possible faults. 28. A method of monitoring, validation and analysis for a process control system having normal process data and sensors, and represented by a plurality of primary residual process models and a plurality of process variables, comprising: measuring real-time sensor data; computing primary residual values of said primary residual process models corresponding to said real-time sensor data; deriving an expected value for one or more of said primary residual values from historical normal process data; determining deviations of said primary residual values from said corresponding expected values; mapping one or more of said deviations of said primary residual values from said corresponding expected values into one or more compressed values; automatically generating one or more fuzzy logic diagnostic rules; computing a certainty factor for a possible fault by applying said fuzzy logic diagnostic rules to said compressed values; and displaying said certainty factor on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 29. The method as set forth in claim 28, further comprising: computing a certainty factor corresponding to each of said primary residual values. 30. The method as set forth in claim 28 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 31. The method as set forth in claim 28 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 32. The method as set forth in claim 31, further comprising: determining a direction of deviation of said primary residual values from said expected values; and determining evidence-for-fault and evidence-against-fault by comparing said direction of deviation to an expected direction of deviation consistent with a fault. 33. The method as set forth in claim 31, further comprising: determining neutral-evidence by comparing the magnitude of a deviation of said primary residual values to zero. 34. The method as set forth in claim 28, further comprising: generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common process variable. 35. The method as set forth in claim 34, further comprising: computing secondary residual values of said secondary residual process models corresponding to said real-time sensor data; and comparing said secondary residual values to expected values. 36. The method as set forth in claim 35, further comprising: computing a certainty factor for a possible fault as a function of one or more of said secondary residual values using fuzzy logic. 37. The method as set forth in claim 36, further comprising: computing a certainty factor corresponding to each of said secondary residual values. 38. The method as set forth in claim 36 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 39. The method as set forth in claim 38, further comprising: determining neutral-evidence from one or more of said secondary residual process models which do not relate to relevant process variables. 40. The method as set forth in claim 35, further comprising: determining a direction of deviation of said secondary residual values from said expected values; comparing said direction of deviation to an expected direction of deviation consistent with a fault. 41. The method as set forth in claim 40 wherein said determination is made by calculating a first partial derivative of said secondary residual process model. 42. The method as set forth in claim 34, further comprising: regenerating said secondary residual process models. 43. The method as set forth in claim 28, further comprising: regenerating said primary residual process models. 44. The method as set forth in claim 28, further comprising: computing a first partial derivative and a second partial derivative of said primary residual process models. 45. The method as set forth in claim 28, further comprising: evaluating said real-time sensor data using statistical process control charting techniques to determine if a corresponding sensor is in control. 46. The method as set forth in claim 28, further comprising: computing a plurality of certainly factors for a corresponding plurality of possible faults. 47. A method of monitoring, validation and analysis for a process control system having normal process data and sensors, and represented by a plurality of primary process models and a plurality of process variables, comprising: generating a plurality of primary residual process models derived from said primary process models, said normal process data, and one or more statistical parameters corresponding to said normal process data; automatically generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common variable by combining said primary residual process models to remove any terms containing said common variable; mapping one or more residual values into one or more compressed values; using one or more of said primary residual process models, one or more of said secondary residual process models and said compressed values to predict a possible fault; and displaying said possible fault on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 48. The method as set forth in claim 47, further comprising: measuring said normal process data. 49. The method as set forth in claim 47, further comprising: calculating said statistical parameters. 50. The method as set forth in claim 47, further comprising: measuring real-time sensor data. 51. The method as set forth in claim 50, further comprising: computing primary residual values and secondary residual values of said primary residual process models and said secondary residual process models corresponding to said real-time sensor data; comparing said primary residual values and said secondary residual values to expected values; and computing a certainty factor for a possible Fault using fuzzy logic. 52. The method as set forth in claim 51 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 53. The method as set forth in claim 51, further comprising: computing a certainty factor corresponding to each of said primary residual values. 54. The method as set forth in claim 51, further comprising: computing a certainty factor corresponding to each of said secondary residual values. 55. The method as set font in claim 47, further comprising: regenerating said secondary residual process models. 56. The method as set forth in claim 47, further comprising: regenerating said primary residual process models. 57. The method as set forth in claim 47, further comprising: using one or more of said primary residual process models and one or more of said secondary residual process models to predict a plurality of possible faults. 58. A method of monitoring, validation and analysis for a process control system having normal process data and sensors, and represented by a plurality of primary process models and a plurality of process variables, comprising: generating a plurality of primary residual process models derived from said primary process models, said normal process data, and one or more statistical parameters corresponding to said normal process data; automatically translating said primary residual process models into executable pseudo-code; measuring real-time sensor data; executing said pseudo-code to compute primary residual values corresponding to said real-time sensor data; deriving an expected value for one or more of said primary residual values from historical normal process data; determining deviations of said primary residual values from said corresponding expected values; mapping one or more of said deviations into one or more compressed values; computing a certainty factor for a possible fault by executing said pseudo-code; and displaying said certainty factor on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 59. The method as set forth in claim 58 wherein said pseudo-code uses fuzzy logic. 60. The method as set forth in claim 59 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 61. The method as set forth in claim 59 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 62. The method as set forth in claim 61, further comprising: determining a direction of deviation of said primary residual values from said expected values; and determining evidence-for-fault and evidence-against-fault by comparing said direction of deviation to an expected direction of deviation consistent with a fault. 63. The method as set forth in claim 61, further comprising: determining neutral-evidence by comparing the magnitude of a deviation of said primary residual values to zero. 64. The method as set forth in claim 58, further comprising: generating pseudo-code for computing a certainty factor; computing a certainty factor corresponding to each of said primary residual values. 65. The method as set forth in claim 58, further comprising: generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common process variable. 66. The method as set forth in claim 65, further comprising: translating said secondary residual process models into pseudo-code; measuring real-time sensor data; executing said pseudo-code to compute secondary residual values corresponding to said real-rime sensor data; and comparing said secondary residual values to expected values. 67. The method as set forth in claim 66, further comprising: generating pseudo-code for computing a certainty factor; computing a certainty factor corresponding to each of said secondary residual values. 68. The method as set forth in claim 67, wherein said certainty factor is computed using fuzzy logic. 69. The method as set forth in claim 65, further comprising: determining neutral-evidence from one or more of said secondary residual process models which do not relate to relevant process variables. 70. The method as set forth in claim 58, further comprising: computing a plurality of certainty factors for a corresponding plurality of possible faults. 71. A computer-readable medium having computer-executable instructions for performing a method for a process control system having normal process data and sensors, and represented by a plurality of primary process models and a plurality of process variables, said method comprising: generating a plurality of primary residual process models derived from said primary process models, said normal process data, and one or more statistical parameters corresponding to said normal process data; measuring real-time sensor data; computing primary residual values of said primary residual process models corresponding to said real-time sensor data; deriving an expected value for one or more of said primary residual values from historical normal process data; determining deviations of said primary residual values from said corresponding expected values; mapping one or more of said deviations of said primary residual values from said corresponding expected values into one or more compressed values; automatically generating one or more fuzzy logic diagnostic rules; computing a certainty factor for a possible fault by applying said fuzzy logic diagnostic rules to said deviations; and displaying said certainty factor on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 72. The computer-readable medium as set forth in claim 71, said method further comprising: measuring said normal process data. 73. The computer-readable medium as set forth in claim 71, said method further comprising: calculating said statistical parameters. 74. The computer-readable medium as set forth in claim 71, said method further comprising: computing a certainty factor corresponding to each of said primary residual values. 75. The computer-readable medium as set forth in claim 71 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 76. The computer-readable medium as set forth in claim 71 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 77. The computer-readable medium as set forth in claim 76, said method further comprising: determining a direction of deviation of said primary residual values from said expected values; and determining evidence-for-fault and evidence-against-fault by comparing said direction of deviation to an expected direction of deviation consistent with a fault. 78. The computer-readable medium as set forth in claim 76, said method further comprising: determining neutral-evidence by comparing the magnitude of a deviation of said primary residual values to zero. 79. The computer-readable medium as set forth in claim 71, said method further comprising: generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common process variable. 80. The computer-readable medium as set forth in claim 79, said method further comprising: computing secondary residual values of said secondary residual process models corresponding to said real-time sensor data; and comparing said secondary residual values to expected values. 81. The computer-readable medium as set forth in claim 80, said method further comprising: determining a direction of deviation of said secondary residual values from said expected values; comparing said direction of deviation to an expected direction of deviation consistent with a fault. 82. The computer-readable medium as set forth in claim 81 wherein said determination is made by calculating a first partial derivative of said secondary residual process model. 83. The computer-readable medium as set forth in claim 80, said method further comprising: computing a certainty factor for a possible fault as a function of one or more of said secondary residual values using fuzzy logic. 84. The computer-readable medium as set forth in claim 83, said method further comprising: computing a certainty factor corresponding to each of said secondary residual values. 85. The computer-readable medium as set forth in claim 83 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 86. The computer-readable medium as set forth in claim 85, said method further comprising; determining neutral-evidence from one or more of said secondary residual process models which do not relate to relevant process variables. 87. The computer-readable medium as set forth in claim 79 wherein said secondary residual process models are functions of both measured process variables and unmeasured process variables. 88. The computer-readable medium as set forth in claim 79, said method further comprising: regenerating said secondary residual process models. 89. The computer-readable medium as set forth in claim 71 wherein said statistical parameters comprise a mean and a standard deviation. 90. The computer-readable medium as set forth in claim 71 wherein said primary residual process models are functions of both a measured process variable and an unmeasured process variable. 91. The computer-readable medium as set forth in claim 71, said method further comprising: regenerating said primary residual process models. 92. The computer-readable medium as set forth in claim 71, said method further comprising: computing a first partial derivative and a second partial derivative of said primary residual process models. 93. The computer-readable medium as set forth in claim 71, said method further comprising: reporting said certainty factor if it exceeds a predetermined threshold. 94. The computer-readable medium as set forth in claim 71 wherein said certainty factor is between 0.0 and 1.0. 95. The computer-readable medium as set forth in claim 71, said method further comprising: evaluating said real-time sensor data using statistical process control charting techniques to determine if a corresponding sensor is in control. 96. The computer-readable medium as set forth in claim 71, said method further comprising: calculating the exponentially weighted moving average of said real-time sensor data to determine if a corresponding sensor is in control. 97. The computer-readable medium as set forth in claim 71, said method further comprising: computing a plurality of certainty factors for a corresponding plurality of possible faults. 98. A computer-readable medium having computer-executable instructions for performing a method for a process control system having normal process data and sensors, and represented by a plurality of primary residual process models and a plurality of process variables, said method comprising: measuring real-time sensor data; computing primary residual values of said primary residual process models corresponding to said real-time sensor data; deriving an expected value for one or more of said primary residual values from historical normal process data; determining deviations of said primary residual values from said corresponding expected values; mapping one or more of said deviations of said primary residual values from said corresponding expected values into one or more compressed values; automatically generating one or more fuzzy logic diagnostic rules; computing a certainty factor for a possible fault by applying said fuzzy logic diagnostic rules to said deviations; and displaying said certainty factor on a monitor, computer screen or control console used for monitoring, validation and analysis for a process control system. 99. The computer-readable medium as set forth in claim 98, said method further comprising: computing a certainty factor corresponding to each of said primary residual values. 100. The computer-readable medium as set forth in claim 98 wherein said fuzzy logic is defined in a diagnostic rule that is regenerated each time said diagnostic rule is used. 101. The computer-readable medium as set forth in claim 98 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND) ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 102. The computer-readable medium as set forth in claim 101, said method further comprising: determining neutral-evidence by comparing the magnitude of a deviation of said primary residual values to zero. 103. The computer-readable medium as set forth in claim 98, said method further comprising: determining a direction of deviation of said primary residual values from said expected values; and determining evidence-for-fault and evidence-against-fault by comparing said direction of deviation to an expected direction of deviation consistent with a fault. 104. The computer-readable medium as set forth in claim 98, said method further comprising: generating one or more secondary residual process models, wherein each of said secondary residual process models is derived from two primary residual process models having at least one common process variable. 105. The computer-readable medium as set forth in claim 104, said method further comprising: computing secondary residual values of said secondary residual process models corresponding to said real-time sensor data; and comparing said secondary residual values to expected values. 106. The computer-readable medium as set forth in claim 105, said method further comprising: computing a certainty factor for a possible fault as a function of one or more of said secondary residual values using fuzzy logic. 107. The computer-readable medium as set forth in claim 106, said method further comprising: computing a certainty factor corresponding to each of said secondary residual values. 108. The computer-readable medium as set forth in claim 106 wherein said fuzzy logic is defined in a diagnostic rule as follows: description="In-line Formulae" end="lead"FAULT-IS-PRESENT=SOME (evidence-for-fault) AND ALL (neutral-evidence) AND NOT (SOME (evidence-against-fault)).description="In-line Formulae" end="tail" 109. The computer-readable medium as set forth in claim 108, said method further comprising: determining neutral-evidence from one or more of said secondary residual process models which do not relate to relevant process variables. 110. The computer-readable medium as set forth in claim 105, said method further comprising: determining a direction of deviation of said secondary residual values from said expected values; comparing said direction of deviation to an expected direction of deviation consistent with a fault. 111. The computer-readable medium as set forth in claim 110 wherein said determination is made by calculating a first partial derivative of said secondary residual process model. 112. The computer-readable medium as set forth in claim 104, said method further comprising: regenerating said secondary residual process models. 113. The computer-readable medium as set forth in claim 98, said method further comprising: regenerating said primary residual process models. 114. The computer-readable medium as set forth in claim 98, said method further comprising: computing a first partial derivative and a second partial derivative of said primary residual process models. 115. The computer-readable medium as set forth in claim 98, said method further comprising: evaluating said real-time sensor data using statistical process control charting techniques to determine if a corresponding sensor is in control. 116. The computer-readable medium as set forth in claim 98, said method further comprising: computing a plurality of certainty factors for a corresponding plurality of possible faults.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (55)
Woods David D. (Murrysville PA) Elm William C. (Monroeville PA) Lipner Melvin H. (Monroeville PA) Butterworth ; III George E. (Monroeville PA) Easter James R. (Pittsburgh PA), Alarm management system.
Brunemann ; Jr. George A. (Cincinnati OH) Reismiller Paul R. (West Chester OH), Apparatus for selecting a valid signal from a plurality of redundant measured and modelled sensor signals.
Li, Rongsheng; Wu, Yeong-Wei; Augenstein, David L., Attitude determination system and method with outer-loop gyro scale-factor non-linearity calibration.
Gross Kenny C. (Bolingbrook IL) Singer Ralph M. (Naperville IL) Humenik Keith E. (Columbia MD), Expert system for online surveillance of nuclear reactor coolant pumps.
Takeuchi Kenji (Murrysville PA) Gagnon Andre F. (Monroeville PA) Cheung Augustine C. (Murrysville PA) Meyer Philip E. (Penn Hills PA), Expert system for surveillance, diagnosis and prognosis of plant operation.
Chen, Zhongping; Ren, Hongwu; Nelson, J. Stuart, Imaging transverse flow velocity using spectral bandwidth of the doppler frequency shift in phase-resolved optical doppler tomography.
Bankert Raymond J. (Clifton Park NY) Imam Imdad (Schenectady NY) Rajiyah Harindra (Clifton Park NY), Integrated model-based reasoning/expert system diagnosis for rotating machinery.
Gregory D. Martin ; Eugene Boe ; Stephen Piche ; James David Keeler ; Douglas Timmer ; Mark Gerules ; John P. Havener, Kiln thermal and combustion control.
Wang Hsu-Pin (Tallahassee FL) Huang Hsin-Hao (Kaohsiung TWX) Knapp Gerald M. (Baton Rouge LA) Lin Chang-Ching (Tallahassee FL) Lin Shui-Shun (Tallahassee FL) Spoerre Julie K. (Tallahassee FL), Machine fault diagnostics system and method.
Spoerre Julie K. (Tallahassee FL) Lin Chang-Ching (Tallahassee FL) Wang Hsu-Pin (Tallahassee FL), Machine performance monitoring and fault classification using an exponentially weighted moving average scheme.
Qin S. Joe (Austin TX) Dunia Ricardo H. (Austin TX) Hayes Randall L. (Georgetown TX), Method and apparatus for detecting and identifying faulty sensors in a process.
Impink, Jr., Albert J.; Rumancik, Joseph A.; Easter, James R., Method and apparatus for on-line monitoring of the operation of a complex non-linear process control system.
Halverson Galen Dean ; Chou Youn-Min, Method and apparatus for statistical process control of machines and processes having non-constant mean of a response va.
Munday George (London GB2) Slater David H. (Berkshire GB2) Tyley Leonard R. T. (London GB2) Berenblut Brian J. (London GB2) Whitehouse Harry B. (Sevenoaks GB2), Method of and apparatus for the detection and analysis of hazards.
Hopkins Robert W. (Rochester NY) Miller Paige (Rochester NY) Swanson Ronald E. (Rochester NY) Scheible John J. (Fairport NY), Method of controlling a manufacturing process using multivariate analysis.
Borchers Hans-Werner,DEX ; Otte Ralf,DEX ; Speh Rainer,DEX ; Weisang Claus,DEX, Process diagnosis system and method for the diagnosis of processes and states in an technical process.
Petsche Thomas (Neshanic Station NJ) Garrett Charles (Plainsboro NJ), Radial basis function neural network autoassociator and method for induction motor monitoring.
White Andrew M. (Skokie IL) Gross Kenny C. (Bolingbrook IL) Kubic William L. (Sante Fe NM) Wigeland Roald A. (Olympia Fields IL), Surveillance of industrial processes with correlated parameters.
Gross Kenneth C. (Bolingbrook IL) Hoyer Kristin K. (Chicago IL) Humenik Keith E. (Columbia MD), System for monitoring an industrial process and determining sensor status.
Keeler James D. (Austin TX) Havener John P. (Austin TX) Godbole Devendra (Austin TX) Ferguson Ralph B. (Austin TX), Virtual continuous emission monitoring system.
Keeler James D. (Austin TX) Havener John P. (Austin TX) Godbole Devendra (Austin TX) Ferguson Ralph B. (Austin TX), Virtual continuous emission monitoring system with sensor validation.
Do, Minh Binh; Verdes, Hector Luis Palacios; Zhou, Rong; Kuhn, Lukas Daniel; de Kleer, Johan, Methods and systems for active diagnosis through logic-based planning.
de Kleer, Johan; Kuhn, Lukas Daniel; Price, Robert; Do, Minh Binh; Zhou, Rong; Liu, Juan; Saund, Eric, Methods and systems for continuously estimating persistent and intermittent failure probabilities for production resources.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.