Data sensor validation system and method

• US 6,687,637 B2
• Filed: 06/18/2002
• Issued: 02/03/2004
• Est. Priority Date: 06/18/2001
• Status: Expired due to Fees

• Current Assignee: GlobVision Inc.
• Original Assignee: GlobVision Inc.
• Inventors: Garabedian, Arminch
• Primary Examiner(s): SHAH, KAMINI S
• Application Number:
  US10/173,267
• Publication Number:
  US 20030097230A1
• Time in Patent Office: 595 Days
• Field of Search: 702/107, 706/20, 342/41, 342/59, 342/357.09, 701/301, 701/21, 701/275, 455/456.5
• US Class Current: 702/107
• CPC Class Codes: G05B 9/02 electric

Abstract

A method of data sensor validation is disclosed. The method comprises the steps of pre-processing data sensor from each sensor from a plurality of sensors for at least segmenting the data sensors into a plurality of groups, each group for grouping sensors for sensing highly relevant data one to another; providing the pre-processed data sensor to a correlation processor, the correlation processor for determining from pre-processed data sensor, pre-processed data that is other than correlated, the determination made in dependence upon redundant pre-processed data other than pre-processed data from two sensors for sensing an identical parameter; and, when pre-processed data that is other than correlated is detected, providing an indication to an operator that the sensor data is other than correlated. Advantageously the method is applicable for use in geographically remote sensor applications wherein the sensors are adapted to provide geographic location information using a global positioning system to the correlation processor through wireless communication with satellites.

Claims

1. A method of data sensor validation comprising the steps of:

pre-processing data sensor from each sensor from a plurality of sensors for at least segmenting the data sensors into a plurality of groups, each group for grouping sensors for sensing highly relevant data one to another;

providing the pre-processed data sensor to a correlation processor, the correlation processor for determining from pre-processed data sensor, pre-processed data that is other than correlated, the determination made in dependence upon redundant pre-processed data other than pre-processed data from two sensors for sensing an identical parameter; and

,

when pre-processed data that is other than correlated is detected, providing an indication to an operator that the sensor data is other than correlated.

2. A method according to claim 1, wherein the step of pre-processing data sensor comprises the steps of:

generating a prediction model for each sensor from the plurality of sensors, the prediction model for forming input patterns and output patterns based upon sets of continuous sensed data;

iteratively training the prediction model, the iteration performed when a pre-determined amount of data are sensed, the iteration performed until a pre-determined level is reached;

generating a regression model for each sensor from the plurality of sensors, the regression model for forming input patterns and output patterns based upon sets of continuous sensed data;

iteratively training the regression model, the iteration performed when a pre-determined amount of data are sensed, the iteration performed until a pre-determined level is reached; and

,

providing pre-processed data in dependence upon at least the regression model to the correlation processor.

3. A method according to claim 2 wherein the step of iteratively training the prediction model comprises the steps of:

creating connection weights; and

storing the connection weights such that they are indicative of the prediction model.

4. A method according to claim 2, wherein the step of iteratively training the regression model comprises the steps of:

creating connection weights; and

storing the connection weights such that they are indicative of the regression model.

5. A method according to claim 2, wherein the pre-determined level is when an acceptable minimum mean square error is achieved.

6. A method according to claim 2, wherein the pre-determined level is when a maximum pre-specified iteration is reached.

7. A method according to claim 1, wherein the step of segmenting the data sensors into a plurality of groups comprises the step of determining one of physical and statistical relationship between data sensors in dependence upon a dynamic data point.

8. A method of data sensor validation according to claim 7, wherein the step of segmenting the data sensors comprising the steps of:

sizing each group from the plurality of groups such that a size corresponds to a pre-determined number of sensor data within a predetermined period of time; and

,

performing a moving average of each group from the plurality of groups provided a group size is greater than twice the pre-determined number of sensor data within a predetermined period of time.

9. A method of data sensor validation according to claim 7, wherein the step of segmenting the data sensors into a plurality of groups comprises the step of generating an input pattern and an output pattern for each group of the plurality of groups, each of the input pattern and of the output pattern comprising continuous sets of sensor data.

10. A method of data sensor validation according to claim 9, wherein the step of generating an input pattern and an output pattern comprises the step of merging the input pattern and the output pattern from each group for forming a training input and output pattern for use with a prediction model.

11. A method of data sensor validation according to claim 7, wherein the correlation processor is coupled for receiving the pre-processed data and for processing the pre-processed data to determine a correlation between pre-processed data from each sensor within a same group;

and,

wherein the step of determining one of physical and statistical relationship comprises the step of performing training of the correlation processor based on a plurality of different segmentations of the plurality of data sensors to determine a significant grouping.

12. A method of data sensor validation according to claim 2, wherein the step of generating a prediction model comprises the step of filling in a missing value.

13. A method of data sensor validation according to claim 11, wherein the correlation processor is for processing correlations between different groups as well.

14. A method of data sensor validation according to claim 13, wherein the correlation processor is for processing correlations data immediately after the step of providing data transformed in dependence upon at least the regression model to the correlation processor.

15. A method according to claim 13, wherein the correlation processor is a neural network.

16. A method of data sensor validation according to claim 1, wherein the correlation processor is a neural network.

17. A method of data sensor validation according to claim 11, wherein the pre-processed data that is other than correlated is pre-processed data that represents a physical parameter that is inconsistent with other sensor data, the other sensor data received from data sensors segmented within a same group of data sensors.

18. A method of data sensor validation according to claim 17, wherein the pre-processed data that is other than correlated is pre-processed data that represents a physical parameter that is inconsistent with other pre-processed data, the other pre-processed data determined from data received from data sensors segmented within a different group of data sensors.

19. A method of data sensor validation according to claim 1, wherein the step of pre-processing data sensor from each sensor from a plurality of sensors comprising the step of suggesting a most probable data for use when data sensor are off a predictable range of data sensor according to data sensor from other sensors from the plurality of sensors.

20. A method of data sensor validation according to claim 17, wherein sensors are environmental sensors.

21. A method according to claim 20, wherein the environmental sensors include a hydrosensor for sensing information and providing data relating to at least one of waterflow and waterlevels.

22. A method according to claim 21, wherein the data is correlated for sensor validation in a water level control system including a plurality of dams and interconnected waterways.

23. A sensor for use in geographically remote sensor applications comprising:

a sensing circuitry for sensing data;

a transmitter for transmitting sensed data to a correlation processor, the correlation processor for determining from pre-processed sensed data, pre-processed data that is other than correlated, the determination made in dependence upon redundant pre-processed data other than pre-processed data from two sensors for sensing an identical parameter at an approximately same geographic location; and

a wireless transceiver circuit for wirelessly determining a geographic location of the sensor, for transmitting the determined geographic location of the sensor to the correlation processor, and for transmitting the sensed data to the correlation processor for allowing the correlation processor to associate the received sensed data with the determined geographic location.

24. A sensor for use in geographically remote sensor applications according to claim 23, wherein the correlation processor is a neural network.

25. A sensor for use in geographically remote sensor applications according to claim 23 wherein data sensed by the sensor is environmental data.

26. A sensor for use in geographically remote sensor applications according to claim 25, wherein the environmental data sensed by a sensor is determined upon the geographical location of the sensor.

27. A sensor according to claim 23, wherein the environmental sensor is a hydrosensor for sensing information and providing data relating to at least one of waterflow and waterlevels.

28. A sensor according to claim 27, wherein the correlation processor is for sensor validation in a water level control system including a plurality of dams and interconnected waterways.

29. A sensor for use in geographically remote sensor applications according to claim 23, wherein the wireless transceiver circuit comprises a global positioning system for determining geographic location of the sensor according to coordinates receive from satellites.

30. A method of data sensor validation comprising the steps of:

pre-processing data sensor from each sensor from a plurality of sensors;

providing the pre-processed data sensor to a correlation processor, the correlation processor for determining from pre-processed data sensor, pre-processed data that is other than correlated, the determination made in dependence upon redundant pre-processed data other than pre-processed data from two sensors for sensing an identical parameter; and

,

when pre-processed data that is other than correlated is detected, providing an indication to an operator that the sensor data is other than correlated.