Features of the present invention will become apparent to those skilled in the art from the following description with reference to the drawings. Understanding that the drawings depict only typical embodiments of the invention and are not therefore to be considered limiting in scope, the invention will be described with additional specificity and detail through the use of the accompanying drawings, in which:
The present invention relates to a method and system for detection and remediation of sensor degradation in a monitoring device. The monitoring device can have one or more sensors and is typically mounted on a platform such as a vehicle. In general, the present method utilizes real-time predictive filtering implemented by an algorithm that outputs high integrity proxy signals to replace signal measurements lost during temporary shut down of the monitoring device. For example, real-time present sensor measurements are combined with predictions from past sensor measurements. The present sensor measurement is replaced with the predicted sensor measurement when the present sensor measurement lies outside of a threshold it could not possibly exceed given the platform on which the monitoring device is mounted.
The method and system can be implemented in any monitoring device mounted to a platform that is subject to a shock event or other environmental influence causing the device to degrade or shut down temporarily. The present method is particularly useful in monitoring devices having one or more sensors such as MEMS-based inertial measurement units (IMUs) used on guided munitions and launch vehicles. While the present method and system particularly improve guidance and navigation performance in MEMS IMUs by mitigating shock events, the present method and system may be used in other applications as well. Examples of such applications include the guidance module in directional drilling tools used for oil and gas exploration, or multi-sensor fusion for autonomous navigation in which the prompt detection of degradation in sensor performance is important for continued integrity of the sensor fusion solution.
The present method utilizes a combination of sensor signal measurement and computational signal prediction, which allows the monitoring device to blend the two signals or switch back and forth between them, depending upon whether the monitoring device is within or outside of a nominal operating condition and how far the monitoring signal and its time derivatives are from nominal.
While the present method and system are generally described in the context of a MEMS IMU, it should be understood that the method can be implemented in other microprocessor operational software associated with monitoring devices that can temporarily shut down because of a platform shock event or other environmental influence.
The system 100 includes a threshold generator module 110, a degradation and shutdown detector/proxy weighting module 120, a signal history generator module 130, a sensor output predictor module 140, and sensor output phase-in/phase-out module 150. In one embodiment, these module components can be implemented in firmware and loaded on the gyro cluster board of a MEMS IMU device.
Outputs from a monitoring device 160 such as a MEMS IMU device provide an actual sensor output signal 162 and a monitor signal 164. The sensor output signal represents the primary signal of interest typically measured using the IMU device (for instance, voltage signals that capture platform attitude rate information). This is the signal affected adversely by shock and the signal that the present system provides a proxy signal for during shock-induced shut down of the primary IMU device.
The monitor signal 164 is another signal output from the IMU device that provides information about the ‘health’ of the IMU device. The monitor signal provides the most accurate information about the effect of shock on the IMU device. For some IMU devices, the monitor signal can be the same as the sensor output itself. In other cases, the monitor signal can be derived from a multiplicity of available sensor signals, and characteristics of the platform on which the sensor is mounted and the environment in which the sensor operates. An example of a simulated monitor signal including its first and second time derivatives is depicted in the graphs of
During operation of system 100 in
The degradation and shutdown detector/proxy weighting module 120 has a first output 122 that provides a degradation flag signal indicating the degrading of sensor output signal 162. A second output 124 provides a shutdown flag signal that indicates whether or not the IMU device is shut down and how far from recovery it is (in the form of a weighting on the proxy signal). The degradation and shutdown detector/proxy weighting module 120 receives as input monitor signal 164, sensor output signal 162, and monitor threshold signals 112. Essentially, when all three signals depicted in
The signal history generator module 130 receives an input comprising the sensor output signal 162 and has an output comprising a sensor output history signal 172 that is communicated to the sensor output predictor module 140. The signal history generator module 130 basically keeps a running history of sensor output signal 162. The time horizon is set so that sufficient past information is available for sensor output predictor module 140 to compute an appropriate proxy signal.
The sensor output predictor module 140 receives historical sensor output information, platform and noise characteristics, shutdown flag and proxy weights, as well as all available domain knowledge, and applies a real-time predictive filter to this information. There is no restriction on the specific form of the predictive filter, so most forms of recursive filters whose predictions are based on probabilistic and/or estimation-theoretic concepts will work.
For example, a real-time predictive filtering algorithm can be used to generate highly reliable proxy signals during temporary sensor shut down. The algorithm can use real-time predictive filtering techniques, such as standard Kalman filtering, finite impulse response (FIR) filtering, infinite impulse response (IIR) filtering, and simple non-linear extrapolation, for example. If noise characteristics are known, optimal filters, such as discrete FIR filters, discrete Kalman filters, or continuous Kalman filters and predictors can be used. The output from sensor output predictor module 140 is a proxy signal 142, which is communicated to sensor output phase-in/phase-out module 150. The proxy signal 142 represents a prediction of what the sensor output would be if shutdown had not occurred.
The sensor output phase-in/phase-out module 150 also receives an input comprising sensor output signal 162 as well as the shutdown flag and proxy weight signals from the degradation and shutdown detector/proxy weighting module 120, and has an output comprising a processed sensor output signal 180. The sensor output phase-in/phase-out module 150 provides the function of a smooth phase-in of the proxy signal during shut down and a corresponding smooth phase-out when the IMU device recovers. Essentially, the shutdown flag and how far the IMU device is from recovery are used dynamically to compute an appropriate weighting on the proxy signal. The complement of this weight is then applied to the actual sensor output signal (a form of complementary filtering) so that, when the IMU device recovers, there is little or no transient in the processed sensor output signal from the IMU device.
A sensor signal range bounded by upper and lower thresholds 310 is predetermined for a particular measured sensor output signal 320 (dotted line). A predicted proxy sensor signal 330 (solid line) is generated concurrently with the measured sensor output signal in real-time during normal operation of the sensor until a signal interrupting event occurs such as a shock. Such an event results in sensor output signal 320 crossing a threshold 310 at a threshold crossing point 312, during which a sensor shut-off period 340 is indicated. The predicted proxy sensor signal 330 continues to be used by the monitoring device during shut-off period 340. The predicted proxy sensor signal 330 used during the shut-off period can be generated based on a previous average of the measured sensor output signal 320. When shut-off period 340 ends at a threshold crossing point 314, sensor output signal 320 is once again used by the monitoring device. The predicted proxy sensor signal 330 can be blended with the measured sensor output signal 320 at the onset of shut-off period 340 and just prior to full recovery from shut-off period 340.
In implementing the method of the invention in a particular monitoring device, an initial determination is made of the most reliable monitor signal among candidate signals for selection as the monitor signal to be used. This should be the signal that provides the most reliable indication of impending sensor shut-off during a signal interrupting event such as a shock event. Alternatively, the monitor signal can be defined from other available signals by combining these signals in a suitable mathematical expression.
A determination is then made of the sources and nature of platform dynamics information. The monitor signal thresholds, timing requirements, and prediction horizons for proxy signal generation are established using a characterization of the noise and dynamics of the platform on which the monitoring device is mounted. For example, in applying the present method to an IMU device, various parameters are considered, such as IMU measurements prior to shock, shock characteristics, and the dynamics of the platform on which the IMU device is mounted, to provide the proxy signal during sensor shut-off.
The noise level in the monitor signal measurement can be used to determine the threshold setting. Thresholds can also be set on the rate of change of the measurement, in addition to the measurement itself. The noise level and therefore the threshold can be pre-computed for a particular sensor. Establishing the thresholds for phase-in/phase-out of the proxy signal can be accomplished with statistical characterization of monitor signal bounds beyond which there is the greatest probability of sensor shut-off. For example, a sensor signal range can be set with a 99.9% confidence level that the measured sensor output signal will not go beyond a threshold value of the sensor signal range until sensor shut-off occurs or is impending.
The shock magnitude expected to be encountered by the sensor during operation is used to determine the confidence level for the threshold setting. For a given shock magnitude in g's, the procedure to set the threshold includes: 1) setting an initial threshold value; 2) checking performance of the predictive sensing system with an ideal sensor, and also performance of the system in which it is used (if possible); and 3) repeating steps 1 and 2 until performance is acceptable.
Instructions for carrying out the various process tasks, calculations, control functions, and the generation of signals and other data used in the operation of the method and systems described herein can be implemented in software, firmware, or other computer readable instructions. These instructions are typically stored on any appropriate computer readable media used for storage of computer readable instructions or data structures. Such computer readable media can be any available media that can be accessed by a general purpose or special purpose computer or processor, or any programmable logic device.
Suitable computer readable media may comprise, for example, non-volatile memory devices including semiconductor memory devices such as EPROM, EEPROM, or flash memory devices; magnetic disks such as internal hard disks or removable disks; magneto-optical disks; CDs, DVDs, or other optical storage disks; nonvolatile ROM, RAM, and other like media. Any of the foregoing may be supplemented by, or incorporated in, specially-designed application-specific integrated circuits (ASICs). When information is transferred or provided over a network or another communications connection (either hardwired, wireless, or a combination of hardwired or wireless) to a computer, the computer properly views the connection as a computer readable medium. Thus, any such connection is properly termed a computer readable medium. Combinations of the above are also included within the scope of computer readable media.
The method of the invention can be implemented in computer readable instructions, such as program modules or applications, which are executed by a data processor. Generally, program modules or applications include routines, programs, objects, data components, data structures, algorithms, etc. that perform particular tasks or implement particular abstract data types. These represent examples of program code means for executing steps of the method disclosed herein. The particular sequence of such executable instructions or associated data structures represents examples of corresponding acts for implementing the functions described in such steps.
The present invention may be embodied in other specific forms without departing from its essential characteristics. The described embodiments and methods are to be considered in all respects only as illustrative and not restrictive. The scope of the invention is therefore indicated by the appended claims rather than by the foregoing description. All changes that come within the meaning and range of equivalency of the claims are to be embraced within their scope.