0% found this document useful (0 votes)
50 views

Signals and Sensors: Lecture 2 - April 7th Taha Khan

This document discusses the static and statistical characteristics of measurement system elements. It defines key characteristics like range, span, sensitivity, nonlinearity, hysteresis, resolution, and tolerance. Environmental factors like interfering and modifying inputs are also examined. The document outlines how to identify these characteristics through calibration experiments involving measuring the output for varying inputs and environmental conditions. Repeated measurements allow determining the statistical distribution and repeatability of a system element. Overall, the document provides an overview of systematically and statistically quantifying the performance of sensors and measurement systems.

Uploaded by

Pavan T
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
50 views

Signals and Sensors: Lecture 2 - April 7th Taha Khan

This document discusses the static and statistical characteristics of measurement system elements. It defines key characteristics like range, span, sensitivity, nonlinearity, hysteresis, resolution, and tolerance. Environmental factors like interfering and modifying inputs are also examined. The document outlines how to identify these characteristics through calibration experiments involving measuring the output for varying inputs and environmental conditions. Repeated measurements allow determining the statistical distribution and repeatability of a system element. Overall, the document provides an overview of systematically and statistically quantifying the performance of sensors and measurement systems.

Uploaded by

Pavan T
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 38

Signals and Sensors

EL4009 - 2020

Static Characteristics of Measurement System Elements

Slides inspired from books Bentley, John P., Principles of Measurement Systems, Lecture 2 - April 7th
4th Ed., Pearson/Prentice Hall, 2005 and Chalmers, D., Sensing and Systems in
Pervasive Computing Engineering Context Aware Systems, Springer, 2011. Taha Khan
Recap
Quantified
mathematically

Cannot be exactly
quantified
Systematic characteristics
• Range

• Range of Input I: [IMIN IMAX]

• Range of Output O: [OMIN OMAX]

• E.g. Thermocouple I: [100°C 250°C] O: [4mV 10mV]


Systematic characteristics
• Span

• Span of Input I: SI = IMAX - IMIN

• Span of Output O: SO = OMAX - OMIN

• E.g. Thermocouple SI = 250-100 = 150°C SO = 10-4 = 6mV


Systematic characteristics
• Ideal Straight line
• An element is said to be linear if corresponding values of I and O lie on a straight line.

• The ideal straight line connects the minimum point A(I , O ) to maximum point B(I , O ) and has the equation
MIN MIN MAX MAX

That forms the ideal straight-line equation given as

With K as the slope

And a as the intercept


Systematic characteristics
• Non-linearity
• Defined in terms of a function N(I) which is the difference
between actual output O(I) and ideal straight-line behavior

producing the actual output

expressed in polynomial in I:
Systematic characteristics
• Non-linearity

• Non-linearity is often quantified in terms of the


maximum non-linearity ; expressed as a percentage
of full-scale deflection (f.s.d) i.e. as a percentage of
span, given as
Systematic characteristics
• Sensitivity
𝚫𝑶
• is the ratio
𝚫𝑰
• As the limit ΔI tends to zero, the ratio becomes the derivative
𝒅𝑶
𝒅𝑰
which is the rate of change of O with respect to I.
• For a linear element, the derivative is equal to the slope or
gradient K of the straight line.
𝒅𝑶 𝑑𝑁
• For a non-linear element =k+
𝒅𝑰 𝑑𝐼
Thermocouple sensitivity
i.e. sensitivity is the slope or gradient of the output
versus input characteristics O(I).
Systematic characteristics
• Environmental effects
• Modifying Input Im
• changes the linear sensitivity of an element
• At standard conditions when Im = 0, sensitivity = K
• Otherwise, Im = new value – standard value
• Sensitivity = K + KMIM Modifying Input
Systematic characteristics
• Environmental effects
• Modifying Input Im
• changes the linear sensitivity of an element
• At standard conditions when Im = 0, sensitivity = K
• Otherwise, Im = new value – standard value
• Sensitivity = K + KMIM Modifying Input

• Interfering Input II
• Changes the straight line intercept a of an element
• At standard conditions when II = 0, a = zero bias
• Otherwise, II = new value – standard value
• Zero bias = a + KIII

• KM and KI are referred to as environmental coupling


constants or sensitivities
Interfering Input
Hence, the generalized model of the actual
output is given as ...

The generalized equation: O=KI+a+N(I)

K is replaced with (K + KM IM)


a is replaced with a+KI II Which is a generalized model of a system element …. only if other
characteristics are not present
Other systematic characteristics
• Hysteresis

• For a given value of I, the output O may be different


depending on whether I is increasing or decreasing.
Hysteresis is the difference between these two values
of O

• Quantified in terms of the maximum hysteresis


expressed as a percentage of f.s.d
Other systematic characteristics
• Hysteresis

Backlash in gear system


Other systematic characteristics
• Resolution
• Some elements are characterized by the output increasing
in a series of discrete steps or jumps in response to a
continuous increase in input
• Resoluton is defined as the largest change in I that can occur
without any corresponding change in O, given as

• Example: ADC/encoders, where the output signal responds


in discrete steps to a continuous increase in input voltage
and change by least significant bit (read: Quantization).
Other systematic characteristics

• Wear and ageing


• effects K and a over time

• E.g. stiffness of a spring k(t) that


decreases due to wear

• where k0 is the initial stiffness, t is


time and b is a constant
Other systematic characteristics

• Error bands
• Sensor manufacturers define the
performance in terms of error bands i.e.

• For any value of I, the output O will be


within ±h of the ideal straight-line value
OIDEAL

• Here an exact systematic statement of


performance is replaced by a statistical
statement in terms of a probability
density function (PDF) p(O).
Statistical characteristics
Statistical characteristics
• Repeatability
• Repeatability is the ability of an element to produce the same output for the same
input when repeatedly applied to it.
• Lack of repeatability in the output O is due to random fluctuations with time in the
environmental inputs IM, II, i.e.
• if the constants KM, KI are non-zero, then there will be corresponding time variations in O.

• PDF for I, IM and II is the normal or Gaussian distribution function

where: 𝑥ҧ = mean or expected value (specifies center of distribution)


σ = standard deviation (specifies spread of distribution).
Statistical characteristics
• Repeatability

If a dependent variable y is a linear combination of independent


variables x1, x2 and x3, i.e.

and if x1, x2 and x3 have normal distributions with standard deviations


σ1, σ2 and σ3 respectively, then the probability distribution of y is also
normal with standard deviation σ given by:
Statistical characteristics
• Repeatability

• Consider the general model

• if ΔO is a small deviation in O from the mean value 𝑂, ത caused by deviations ΔI,


ΔI and ΔI from respective mean values 𝐼,ҧ 𝐼𝑀 ,and 𝐼ഥ𝐼 , then
M I

• Thus ΔO is a linear combination of the variables ΔI, ΔIM and ΔII


Statistical characteristics
• Repeatability
• Given that I, IM and II follow a normal distribution, standard deviation of O for
a single element can be given as

• And the corresponding mean is

• And the corresponding PDF p(O)


Statistical characteristics
• Tolerance
• is the statistical variation among a batch of similar elements

• E.g. assume that the resistance Ro at 0o C of 100 temperature


sensors is quoted by the sensor manufacturer to be 100Ω

• However, small random variations in manufacturing results in a


different value of Ro for different sensors

• This effect can be given as normal PDF

• σ𝑅𝑜 is the tolerance limit generally quoted by the


manufacturers
Identification of statistic characteristics

• Calibration

• Measuring values of corresponding input I, output O and the environmental inputs IM and II
experimentally, when I is constant or changed slowly, is referred to as Calibration

• Instruments and methods to quantify these variables are referred to as standards

• Accuracy = true value – measured value (Lecture 3)


Identification of statistic characteristics
In order to detect any
• Calibration hysteresis, separate regressions
should be performed on the
Calibration experiment is divided into 3 parts two sets of data to yield two
polynomials:
1. O versus I with IM = II = 0

Significant hysteresis Insignificant hysteresis


Identification of statistic characteristics

• Calibration

2. O versus IM and II at constant I


• Keeping constant I = IMIN, identify which II affects zero bias a.
• If ΔO > 0, then II is interfering, otherwise not interfering

• Compute KI as
• Now, keeping constant I = , identify which IM affects sensitivity.
• If ΔO > 0, then IM is interfering, otherwise not interfering

• Compute K asM
Identification of statistic characteristics

• Calibration

3. Repeatability test

for k measurements in a normal working environment,


we compute p(O) using

• Mean

• Standard deviation

• A histogram of the values Ok should then be plotted


and compared with p(O)
Statistical representation of signals
Statistical representation of signals
• Deterministic signal
• is one whose value at any future time can be exactly
predicted.

• Random signal
• is one whose behavior is not known once the observation
period TO is over.

• Behavior of a random signal is estimated using five


statistical quantities: 1) Mean, 2) Standard deviation, 3)
Probability density function, 4) Power Spectral density and
5) Autocorrelation

• Unwanted random signals are usually referred to as noise signals

• Unwanted deterministic signals as interference signals.


Statistical representation of random signals
• Mean of sample signal

• Standard deviation of sample signal


Statistical representation of random signals
• Probability Pj
Statistical representation of random signals
• Probability density function (p.d.f)

P.d.f is the derivative of P( y)

Probability P that the signal will lie between y and y is given by


y1, y2 1 2

• i.e. the area under the probability density curve between y and y .
1 2

• The total area under the probability density curve is equal to unity, corresponding
to the total probability of the signal having any value of y. The normal probability
density function provides an adequate description of the amplitude distribution
of random noise signals
Statistical representation of random signals
• Power Spectral Density φ (ω)
• is a stationary quantity which is a measure of how the
power in a random signal is distributed amongst different
harmonic frequencies ω.

• In order to explain the meaning of φ (ω), we approximate


the random signal by a periodic signal in which the
waveform recorded during the first observation period TO
is exactly repeated during each subsequent observation
period.

• Power spectral density is given as


Power Cumulative
Where W is the cumulative power and wn is the power at nth
harmonic spectrum Power
Statistical representation of random signals
• Example:
• For a 1 ohm resistance the power is given by P = V2/R
• The Fourier series for a voltage signal V with period TO is

where the fundamental frequency ω1 = 2π/TO, a0 = mean y

• the instantaneous power in the resistor at time t is a2ncos2nω1t watts, and the average power
over period TO is:

• power due to the nth harmonic at frequency nω1


Statistical representation of random signals

•Autocorrelation
Autocorrelation coefficient Ryy of
signal y of length N can be given
as

where m is the integer multiple of


the sampling interval ΔT

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy