The invention pertains to sensors. Particularly, the invention pertains to calibration of sensors, and more particularly to self calibration of sensors.
The invention is a system for static and dynamic calibration of sensors.
A system may exist for the automation of static and dynamic calibration for a certain class of sensors. Static calibration may be performed via a slope seeking loop or algorithm. Dynamic calibration may be performed with both the slope seeking loop and a variation of the slope seeking set point.
The system may remove a need for manual adjustment of sensors to account for sensor drift due to ambient condition changes, and also the need for recalibration for operating condition changes. This may be significant if the sensors are wireless and there is a desire to minimize the need for manual adjustments.
The sensors may retain accuracy under changing ambient conditions and also recalibrate themselves to adjust to operating conditions and sensor aging. This may be done through using feedback control via sensing of the ambient conditions and the operating conditions.
A drift control feedback loop may be designed using identified empirical or semi-empirical constitutive models of sensor material. Similarly, a feedback loop may sense changes in operating conditions and use either a lookup table or a sensor model to change the gains at the sensor output.
Self-calibration of the sensors may be compelling for one or more of the following reasons. Accurate sensor calibration is time-consuming, expensive, and often manual. Aging of sensor parts compels periodic recalibration. Changes of operating conditions also necessitate recalibration. Sensor accuracy is compromised when operating conditions are not same as the calibration conditions. It is difficult to manufacture affordable sensors that do not need calibration or recalibration.
Self-tuning of the sensors may be compelling for one or more of the following reasons. The settling time of a sensor is different at various calibration points. The manufacture of sensors with uniform setting times is typically unaffordable. Thus, sensors generally require specific settings of calibration points. Aging and changes of operating conditions may change the settling time.
There may be a large class of or many sensors that transduce a difference between a control signal and an environmental signal to produce an output (typically electrical). Examples may include microphones, flow control valves, thermocouples, gimbaled mechanisms, and other relative-type measuring devices.
The present approach may be to adapt the operating point of a sensor for self-calibration. There is no need to do multiple sensors. A rough estimate of the operating curve may serve to eliminate individual calibration of manufactured sensors. The accuracy of a sensor may be well characterized through knowledge of the uncertainty in the transducer dynamics. Slope seeking is a key to an application of the present system.
The system may reduce sensor costs with the elimination or reduction of calibration tasks. The automatic calibration mechanism may be autonomous from the subject sensors.
In
f′ref(k+1)=f′ref(k)+g1(τs(k)), and
K1(k+1)=K1(k)+g2(τs(k)).
“k+1” and “k” indicate time steps. “τs” indicates settling time, “g1” and “g2” indicate a function of settling time relative to the set slope f′ref and amplifier gain K1, respectively.
A set slope f′ref 39 may be sent as an input to a slope setting processor 42 of a slope seeker for static calibration module 43. For the same sensitivity, the product of K1 and f′ref may be a constant. The slope setting processor 42 may reflect the following equation,
−(a/2)R{e−jφjωFo(jω)Co(jω)Fi(jω)}.
“a” may indicate a magnitude, “R” may indicate the real part, Fi may indicate the exciting dynamics, Fo may indicate the settling dynamics, and Co may indicate the washout filter (if ω is small, Fi(jω) and Fo(jω) behave as constant gains). The slope setting processor 42 may output a slope setting to a summer or adder 44. An output of washout filter sCo(s) 35 may multiplied with a phase shift signal 46 represented by sin(ωt−φ), at multiplier 45. The output of multiplier 45 may go to adder 44 where it is summed with the slope setting from processor 42. The output of adder 44 may include a signal representing a tracking error which is proportional to the difference between where one is and where one should be. This output of adder 44 may go to a tracking compensator 47 which may have a signal transformation aspect that is represented by Ci(s)/s. The output of compensator 47 may be a setting signal that goes to adder 22 to be combined with the perturbation signal from the low frequency forcing generator 21. The forcing may be an additive to the input of the exciting dynamics 23. The sinusoidal signal may be added to perturb the current setting. The output path of adder 22, along with the other processes, may be noted above.
In the present specification, some of the matter may be of a hypothetical or prophetic nature although stated in another manner or tense.
Although the invention has been described with respect to at least one illustrative example, many variations and modifications will become apparent to those skilled in the art upon reading the present specification. It is therefore the intention that the appended claims be interpreted as broadly as possible in view of the prior art to include all such variations and modifications.