U.S. patent application number 15/863102 was filed with the patent office on 2018-07-12 for system and method of identifying path of residual current flow through an intelligent power strip.
The applicant listed for this patent is Liebert Corporation. Invention is credited to Kevin R. FERGUSON.
Application Number | 20180196098 15/863102 |
Document ID | / |
Family ID | 61074549 |
Filed Date | 2018-07-12 |
United States Patent
Application |
20180196098 |
Kind Code |
A1 |
FERGUSON; Kevin R. |
July 12, 2018 |
SYSTEM AND METHOD OF IDENTIFYING PATH OF RESIDUAL CURRENT FLOW
THROUGH AN INTELLIGENT POWER STRIP
Abstract
A system and method is disclosed for detecting a specific
voltage phase, from a multiphase voltage source, and a specific
outlet of an intelligent power strip, that is associated with a
residual current flow. The method accomplishes this by using a
system that employs a statistical time series analysis using a
Pearson's correlation coefficient calculation to measure the linear
dependence between the discretely sampled residual current waveform
and each phase and outlet's discretely sampled current waveforms,
in turn. A residual current as low as 1 mA can be accurately
measured and its associated voltage phase source, as well as which
outlet of an intelligent power strip it flows out of, can be
reliably determined.
Inventors: |
FERGUSON; Kevin R.; (Dublin,
OH) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Liebert Corporation |
Columbus |
OH |
US |
|
|
Family ID: |
61074549 |
Appl. No.: |
15/863102 |
Filed: |
January 5, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
62443308 |
Jan 6, 2017 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01R 31/42 20130101;
G01R 31/50 20200101; G06F 1/266 20130101; G01R 21/133 20130101;
G01R 19/2513 20130101; H02H 3/042 20130101; G01R 31/52 20200101;
G06F 2200/261 20130101; G01R 19/02 20130101; H02H 3/32 20130101;
G01R 19/2509 20130101; H02H 3/16 20130101 |
International
Class: |
G01R 31/02 20060101
G01R031/02; G01R 19/02 20060101 G01R019/02 |
Claims
1. A method for detecting a residual current flowing through an
intelligent power strip having at least one alternating current
(AC) power outlet, the method comprising: obtaining current sensing
information; performing an analog-to-digital (ADC) conversion of
the current sensing information to obtain ADC samples; managing a
time series collection of the ADC samples as residual and outlet
current waveform samples; calculating residual current (RC) RMS
values for the obtained residual current samples; calculating phase
current from the aggregation of the outlet current waveform samples
having the same phase; calculating Pearson's correlation
coefficients for variables relating to residual current and the
phase outlet current waveforms; calculating Pearson's correlation
coefficients for the variables relating to residual current and
individual outlet currents waveforms from each AC power outlet;
calculating the phase having a maximum positive Pearson's
correlation when the residual current RMS is greater than a
predetermined residual current RMS threshold; and determining an
individual one of the AC power outlets having the maximum Pearson's
correlation when the residual is greater than the predetermined RC
threshold.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] This application claims the benefit of U.S. Provisional
Application No. 62/443,308, filed on Jan. 6, 2017. The entire
disclosure of the above application is incorporated herein by
reference.
FIELD
[0002] The present disclosure relates to a system and method of
identifying a path of residual current through an Intelligent Power
Strip (IPS).
BACKGROUND
[0003] The statements in this section merely provide background
information related to the present disclosure and may not
constitute prior art.
[0004] Residual currents caused by the failure of insulation can
constitute a significant risk to safety in electrical systems.
Using an appropriate protective concept it is possible to detect
residual currents, discover/eliminate insulation faults quickly,
and therefore ensure the availability of the system.
[0005] The acronym "RCM" stands for "Residual Current Monitoring"
and means the monitoring of residual currents in electrical
systems. While the current sensors in the electrical systems
referred to herein are AC current sensors, if different current
sensor types or configurations are used, DC components could also
be accurately measured using the same monitoring method. The
residual current is calculated as the non-zero sum of the currents
flowing through all current-carrying conductors, apart from the
protective earth (PE), which feed into the electrical system.
Residual currents are typically the result of insulation faults,
residual currents or electromagnetic compatibility (EMC) filter
residual currents, for example. While RCD devices (residual current
circuit breakers) switch off the power supply in the event of a
certain residual current being exceeded, RCM measuring devices
indicate the actual value, record the long-term development and
report when the measured value exceeds a critical value. This
information can also be used in order to switch off the power
supply via external switching devices (contactors, relays). Through
the use of residual current measuring (RCM) devices, it is possible
to detect and report residual currents in a timely manner. This
makes it possible to initiate counter measures within a
sufficiently short time so that it is not necessary to switch the
system off. This facilitates the implementation of measures in the
event of slowly deteriorating insulation values or steadily rising
residual currents--caused for example by aging insulation--before
the system is switched off.
[0006] A power strip, often referred to as a "power distribution
unit" ("PDU"), is typically used in a data center environment, with
one or more units installed in racks arranged in rows, to power
Internet Technology equipment (ITE). A single PDU may provide power
to dozens of devices per rack via outlet sockets, and an
"intelligent" rack PDU employed in a data center can measure and
control the loads. A powered device, e.g., server or network
switch, has one or more internal switched mode power supplies that
on occasion may fail prematurely for various reasons, e.g.,
exposure to excessively high-temperature which can degrade
electronic component lifetime. Also, the integrity of the TNS
systems (Terra Neutral Separate) may become inadvertently
disconnected or fail. These failures may develop suddenly or
gradually over time. The failure mode may result from compromised
or complete breakdown of component insulation spacings, resulting
in a lower impedance conduction path between line voltages of the
device's power supply to protective earth ground. While the
residual current that flows through the protective earth ground is
not of sufficient magnitude to trip the unit's branch overcurrent
protection device, only 30 mA can become a safety hazard to anyone
touching the chassis. While it is critical to identify and provide
an alert when a condition of excessive ground residual current
develops, it is also important that the offending device can be
quickly isolated and removed from the power distribution to
maintain high-availability of other systems components.
[0007] One particular known technique for measuring residual
current involves using a sensitive current transformer to detect
and/or measure the residual current or current that does not flow
back on the return path through an intelligent power strip.
Although this method is able to detect that a residual current
condition exists, it is unable to identify which one of a plurality
of voltages of a multiphase voltage supply is sourcing the residual
current condition, as well as which specific AC outlet of the PDU
is associated with the residual current condition.
SUMMARY
[0008] This section provides a general summary of the disclosure,
and is not a comprehensive disclosure of its full scope or all of
its features.
[0009] In one aspect the present disclosure relates to a method for
detecting a residual current flowing through an intelligent power
strip having at least one alternating current (AC) power outlet.
The method may comprise obtaining current sensing information and
performing an analog-to-digital (ADC) conversion of the current
sensing information to obtain ADC samples. The method may further
include managing a time series collection of the ADC samples as
residual and outlet current waveform samples, and then calculating
residual current (RC) RMS values for the obtained residual current
samples. The method may further include calculating phase current
from the aggregation of the outlet current waveform samples having
the same phase, and then calculating Pearson's correlation
coefficients for variables relating to residual current and the
phase outlet current waveforms. The method may further include
calculating Pearson's correlation coefficients for the variables
relating to residual current and individual outlet currents
waveforms from each AC power outlet, and calculating the phase
having a maximum positive Pearson's correlation when the residual
current RMS is greater than a predetermined residual current RMS
threshold. Finally the method may include determining an individual
one of the AC power outlets having the maximum Pearson's
correlation when the residual is greater than the predetermined RC
threshold.
[0010] Further areas of applicability will become apparent from the
description provided herein. The description and specific examples
in this summary are intended for purposes of illustration only and
are not intended to limit the scope of the present disclosure.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] The drawing described herein is for illustration purposes
only and is not intended to limit the scope of the present
disclosure in any way. In the drawing figures:
[0012] FIG. 1 is a high level block diagram of one embodiment of an
apparatus in accordance with one embodiment of the present
disclosure;
[0013] FIG. 2 is a high level block diagram showing in greater
detail various components that may be incorporated into the
microcontroller shown in FIG. 1;
[0014] FIG. 3 is a high-level flowchart summarizing operations
performed by the apparatus of FIG. 1 in detecting a residual
current condition, as well as a specific voltage phase associated
with the residual current condition. If the apparatus of FIG. 1
includes dedicated current metering circuitry to support
measurement of current at individual AC outlets, a residual current
condition at one or more AC outlets can also be detected;
[0015] FIG. 4A is a detailed flowchart illustrating operations
performed by the apparatus of FIG. 1 in detecting a residual
current condition, as well as a specific voltage phase; and
[0016] FIG. 4B shows additional operations that may be performed by
the apparatus of FIG. 1 if the apparatus includes dedicated current
metering circuitry to support measurement of current at individual
AC outlets.
DETAILED DESCRIPTION
[0017] The following description is merely exemplary in nature and
is not intended to limit the present disclosure, application, or
uses. It should be understood that throughout the drawings,
corresponding reference numerals indicate like or corresponding
parts and features.
[0018] Referring to FIG. 1, one embodiment of an apparatus 10 is
shown which forms an intelligent equipment rack power distribution
unit able to monitor for residual current associated with any one
or more of a plurality of rack mounted assets being powered by the
apparatus 10. For convenience, the apparatus 10 will be referred to
throughout simply as "PDU 10".
[0019] The PDU 10 may incorporate a residual current monitor
current transformer 11, hereinafter simply "RCMCT 11", which is
operatively coupled to three-phase AC power from some upstream
device (e.g., AC mains source, UPS, etc.) and which senses a
residual current condition affecting any one or more of a plurality
of AC outlets 10a of the PDU. The operation of the RCMCT 11 and its
connections to other components of the PDU 10 will be described in
greater detail in the following paragraphs.
[0020] The PDU 10 in this example may also incorporate a
communications module RPC2 12 (hereinafter simply "RPC 12"), which
may be a hot-swappable web card which is installed in the PDU 10,
and in this example may be the RPC2.TM. Network Interface Card
available from Vertiv Co., assignee of the present disclosure. The
RPC 12 may include a microcontroller 14 and preferably also a
non-volatile (NV) memory 16. The RPC 12 may also include a
plurality of ports including, but not limited to, a LAN Ethernet
port 18, an Expansion/Management port 20, a port 22 for coupling to
a display module (e.g., "BDM" or "Basic Display Module" available
from the assignee of the present disclosure), one or more 1-wire
sensor ports 24, an RS-232 port 26 and a USB port 28.
[0021] The PDU 10 further may include a Rack PDU Controller (RPDUC)
30 having a microcontroller 32 and a non-volatile memory 34, and
one or more branch receptacle controllers (BRC) 36. The RPDUC 30
receives current information from the RCMCT 11. Each BRC 36 may
have a complex programmable logic device (CPLD) 38 having a voltage
and current sensing subsystem 38a which senses of a loss of AC
input power, a plurality of bistable relays 40, and an open circuit
breaker (OCB) detection subsystem 42 which senses for an open
circuit breaker condition. The RPDUC 30 is in bidirectional
communication with each of the BRCs 36 via a bus 44. The RPC 12 is
in bidirectional communication with the RPDUC 30 via a bus 46. A
reset switch 48, which is easily accessible by a user via a
faceplate of the PDU 10, is provided for enabling the user to
initiate a hard reset of the PDU.
[0022] FIG. 1 also shows a plurality of branch circuit breakers
(CB) 50 that each receives AC input power directly from the
current-carrying conductors passing through the RCMCT 11. By
"branch" circuit breaker it is meant that each one of the CBs 50 is
typically associated with one specific BRC 36. The OCB detection
subsystem 42 monitors the CBs 50 to detect when any one or more
have been tripped to an open condition. And as explained above,
each BRC 36 includes a plurality of bistable relays 40, which in
one specific embodiment comprise eight (8) bistable relays.
However, it will be appreciated that a greater or lesser number of
bistable relays 40 could be provided per branch. Mechanical
bistable relays have coils and mechanical contacts. They can be
single coil or dual coil relays. Also, more than one CB 50 may
exist for each BRC 36. For example, each BRC 36 can have its
bistable relays arranged in two sub banks, with a separate CB 50
associated with each sub-bank. As used herein, each sub-bank of a
BRC 36 is a branch of the BRC.
[0023] FIG. 1 also shows the plurality of AC power outlets 10a. In
this example, each outlet is single-phase in line-neutral or
line-line configuration, with its associated phase established at
manufacturing time and saved in NV memory 16. Each outlet has a
first associated optical element 10a1 and a second optical element
10a2. Optical elements 10a1 may each be an LED having a first
color, for example green, that indicates a status of the specific
bistable relay 40 associated with its specific AC outlet 10a. The
second group of optical elements 10a2 may also be, for example,
LEDs having a different color, for example red, for providing
additional information to the user. Each one of the green LEDs 10a1
may indicate, for example, that the bistable relay 40 associated
with that specific AC outlet 10a is closed, and an extinguished
green LED 10a1 would therefore indicate that the associated
bistable relay is open. Each AC power outlet 10a may be used to
power an associated rack mounted asset (e.g., server, network
switch, etc.)
[0024] The RPDUC 30 is shown in greater detail in FIG. 2. The
microcontroller 32 of the RPDUC 30 may include a static random
access memory 32a (hereinafter simply "SRAM 32a") and a
non-volatile read only memory or non-volatile random access memory
32b (hereinafter simply "NV memory 32b"). The SRAM 32a may contain
a plurality of circular buffers 32a1 to hold current waveform data
samples, while the NV memory 32b may be used to store gain
coefficients 32b1 and ADC skew coefficients 32b2. The use of the
circular buffers 32a1, the gain coefficients 32b1 and the ADC
coefficients 32b2 will be discussed further in the following
paragraphs. And while the SRAM 32a and NV memory 32b in this
example are shown as part of the microcontroller 32, it will be
appreciated that they could be implemented as independent memory
devices just as well.
[0025] The RPDUC 30 includes a voltage sensing subsystem 52 and a
current sensing subsystem 54. The subsystems 52 and 54 perform real
time RMS voltage measurements and RMS current measurements,
respectively, and thus monitor the power input from the AC power
source. The RCMCT 11 is also in communication with the current
sensing subsystem 54. Three phase LEDs 55a-55c may be incorporated
and/or operatively coupled to the RPDUC 30 that indicate the
voltage and current conditions of each phase of the AC input,
including the residual current condition. The monitored power
information may be shared with the RPC 12 via bus 44. As noted
above, the voltage and current sensing subsystem 38a of each BRC 36
also monitors for a loss of AC input power, so in this regard there
is redundancy of this feature in the RPDUC 30 and the BRCs 36. The
current sensing subsystem 54 of the RPDUC 30 receives an input
current signal from each of the branch BRCs (collectively labeled
for simplicity in FIG. 2 with number 36) which it uses to perform
its current sensing function. Each branch BRC 36 also includes a
plurality of current transformers (CTs) 56 for independently
measuring a current being drawn by the AC outlets 10a associated
with each branch of bistable relays 40. The signals from each
branch CT 56 are input to the current sensing subsystem 54 for
analysis.
[0026] The RPC 12 shown in FIG. 1 manages, monitors and reports
information about PDU 10 energy metering and power distribution
status obtained from the RPDUC 30 to networked software clients.
The RPDUC 30 provides support for the energy metering measurements
and calculations, control management, and communications interfaces
to the RPC 12, as described above. The RPDUC 30 communicates with
each BRC 36 and, except upon power loss, controls the bistable
relays 40 of each BRC by sending command messages to each BRC to
independently control each one of its associated bistable relays
40.
[0027] The BRC 36, and more particularly its CPLD 38, directly
controls its bistable relays 40. The BRC 36 also senses individual
LED outlet operational status, and loss of an AC input power signal
via line frequency monitoring performed by the voltage and current
detection subsystem 38a, as well as using the OCB subsystem 42 to
detect for an open circuit breaker condition. The bistable relays
40 of each BRC 36 in this example require a nominal 16 msec pulse
to their coils to change states, that is, to open or close their
contacts. A reference herein to a bistable relay being "open" means
that its contacts are open and power is off or interrupted at the
outlet 10a to which the bistable relay switches power. As used
herein, "power up", "power down", "power failure", and "power
cycle" refer to specific conditions of input AC line voltage, which
is the AC power provided to the outlets 10a through the bistable
relays 40 of each BRC 36. The term "Configured state", when used in
connection with the bistable relays 40, means the state that a
given bistable relay is configured to be in (i.e., open or closed)
when power is on. For the purpose of the present disclosure, it may
be understood that the term "configured state" means that the
bistable relays 40 will, after the PDU 10 is powered up, have
closed contacts in order to switch power on at the AC outlets
10a."
[0028] The RPC 12 commands the RPDUC 30 via a SMBus (I2C)
communication bus, bus 46 in FIG. 1 in this example, which in turn,
commands the BRC 36 via a SPI communication bus, which is bus 44 in
this example, to configure the relay state of each bistable relay
40. The RPDUC 30 is capable of autonomous behavior without RPC 12
commands. The one or more BRCs 36 are each capable of autonomous
behavior without RPDUC 30 commands.
[0029] The PDU 10 and its method of operation significantly extend
the ability of traditional systems and methods for measuring
residual current by incorporating statistical analysis. The
statistical analysis is used to infer which voltage phase of the AC
input voltage, and which one of the AC outlets 10a (FIG. 1) of the
PDU 10, is affected by the path of a residual current. In this
manner it is possible to quickly detect which one of a plurality of
rack assets connected to the AC outlets 10a of the PDU 10 is the
faulty device giving rise to the residual current flow so that the
faulty device can be quickly disconnected from the PDU. This
enables the remaining, properly functioning, devices operating on
the PDU 10 to continue in their operation without the need to
independently power down each of the properly functioning devices
coupled to the PDU in order to identify the offending device.
[0030] Referring to FIG. 3, a flowchart 300 is shown providing a
high-level summary of the operations that may be performed by the
PDU 10 in carrying out the detection of a residual current
associated with any AC outlet 10a of the PDU 10. FIG. 3 also
provides a list of the discrete operations that may be performed to
carry out each of the high-level operations, to be discussed below
with reference to FIG. 4. It should be noted that operations
illustrated in the flowchart 300 may be performed by the
microcontroller 32 of the RPDUC 30 of the PDU 10, but for
convenience it will be stated hereinafter that an operation may be
performed by the PDU 10 or by the microcontroller 32.
[0031] As an overview of the various operations shown in FIG. 3, it
will be understood that the PDU 10 detects the voltage phase in
which the residual current path is flowing. If the PDU 10 supports
individual current measurements at the outlets, then the PDU 10
also detects the AC outlet 10a through which the residual current
path is flowing. The PDU 10 accomplishes this by using statistical
time series analysis along with a Pearson's correlation coefficient
calculation to measure the linear dependence between the discretely
sampled residual current waveform for each phase and the
aggregation of each AC outlet's 10a discretely sampled current
waveforms calculated per phase, in turn. The per-phase aggregation
may result from current measurements by a single sensor dedicated
to a set of same-phased outlets or by a plurality of sensors per
each outlet. The Pearson's correlation coefficient is the
covariance of the compared values divided by the product of their
standard deviations. A residual current as low as 1 mA can be
accurately measured/detected along with its associated voltage
phase source. From this information it can be reliably determined
which AC outlet 10a the residual current is flowing out of. The
presence of a residual current is constantly measured and monitored
for in real-time. When a residual current threshold level is
exceeded, an audible alarm may be sounded, an event notification
may be delivered, and a visual indicator of the associated phase
may be flashed on the phase LED 55a-55c and/or on the LED
corresponding to an outlet of the intelligent power strip, if the
PDU 10 is equipped with dedicated current metering circuitry per
outlet, until the residual current path is eliminated or the alarm
threshold is increased.
[0032] Referring further to FIG. 3, the methodology employed by the
PDU 10 may begin at operation 302, with analog-to-digital
conversion (ADC) and time series collection of residual and outlet
current waveform samples. This is accomplished by means of discrete
operation 106, to be discussed below. At operation 304, the root
mean square (RMS) value of the residual current may be calculated,
by means of discrete operations 108, 110, 112, 114, 116, 118, 120,
140, 142, 144, 146, 148, 150, and 184. At operation 306, the
currents of individual outlets and aggregated outlet (phase)
currents may be calculated. This may be accomplished by means of
discrete operations 122, 124, 126, 128, 130, 132, 134, 136, 140,
and 184. At operation 308, discrete operations 180, 182, and 152
may be performed in order to calculate Pearson's correlation
coefficients for residual and phase currents. At operation 310,
Pearson's correlation coefficient may be calculated for RC and
individual outlet currents by means of discrete operations 154,
182, and 156. It should be noted that operation 310 supports both
in-phase and quadrature evaluation and incremental phase shift up
to 15 degrees. At operation 312, the phase having the maximum
Pearson's correlation is calculated when the residual current
exceeds the threshold; this is accomplished by discrete operation
158. If the PDU 10 is equipped with dedicated current metering
circuitry for each outlet 10a, operation 314 may be performed to
determine the outlet having the maximum Pearson's correlation when
the residual current exceeds the threshold. This is accomplished by
the sequence of discrete operations 160, 162, 163, 164, 166, 168,
170, 172, 174, 176, and 178.
[0033] Referring to FIG. 4A, a flowchart 100 is shown illustrating
various operations that may be performed by the PDU 10 in carrying
out the detection of a residual current associated with any AC
outlet 10a of the PDU 10. It should be noted that operations
illustrated in the flowchart 100 may be performed by the
microcontroller 32 of the RPDUC 30 of the PDU 10, but for
convenience it will be stated hereinafter that an operation may be
performed by the PDU 10 or by the microcontroller 32.
[0034] Referring further to FIG. 4A, the methodology employed by
the PDU 10 may begin at operation 102 where phase analysis
variables .alpha. and .phi. are initialized, and at operation 104
the ADC skews K.sub.r and K.sub.o may be pre-calculated. The ADC
skews K.sub.r (for each residual current ADC channel) and K.sub.o
(for each outlet current channel) are pre-calculated time delays,
representing the delta time between successive samples captured one
after another; if each sample duration is approximately n
microseconds, with m outlet currents the last sample is captured
after (m-1)*n .mu.sec. The skew adjustment uses a linear
extrapolation to estimate an extrapolated normalized current
waveform sample, as if all m outlet currents were sampled
simultaneously by microcontroller 32. At operation 106 a time
series of sampled current waveforms received over an AC power line
cycle may be collected.
[0035] At operation 108 the starting index n of the s.sub.r
circular buffer 32a1 for RCM current waveform samples is
initialized in accordance with Equation 1 below.
n.rarw..PHI..sub..alpha.=.alpha.*90+.phi.
N.rarw.n+s.sub.rsize Equation 1
[0036] At operation 110 a test may then be conducted on the s.sub.r
buffer index (n<N) to determine when the time series has been
completely processed. If the test produces a "Yes" answer, then at
operation 112 the residual current sample is filtered by the
exponentially weighted moving average and copied into buffer
i.sub.r in accordance with Equation 2 below. The value of
.beta..sub.1 (in this example, 0.05) is dependent upon sample rate
(in this example, the controller samples 64 times per line cycle)
and provides acceptable smoothing characteristics for 1 ma
precision while remaining responsive to changing conditions.
s.sub.r[n].rarw.s.sub.r[n]+.beta..sub.1*(s.sub.r[n]-s.sub.r[n-1])
i.sub.r'=s.sub.r[n], i.sub.r''=s.sub.r[n-1] Equation 2
[0037] At operation 114, corrections to remove electronic and
signal offsets and apply calibrated scalars determined at
manufacturing time are then calculated for I.sub.r, ADC skew, mean
and gain, as indicated by Equation 3 below.
i.sub.r=G.sub.r*(i.sub.r'-K.sub.r/(i.sub.r'-i.sub.r'')-i.sub.r.mu.)
Equation 3
[0038] At operation 116, the residual current sample
minimum/maximum peaks may be determined, as indicated by Equation 4
below. These may be saved in order to later determine the half-wave
symmetry of the current waveform (i.e., full-wave or half-wave
characteristic).
i.sub.r .rarw.i.sub.r >i.sub.r'
i.sub.r .rarw.i.sub.r <i.sub.r' Equation 4
[0039] At operation 118, the microcontroller 32 may then calculate
the residual current sample integral in accordance with Equation 5
below. Because a zero-valued integral is expected over a line cycle
of a half-wave symmetrical periodic waveform, operation 118 enables
precise correction of creeping offset errors.
i.sub.r.SIGMA.=.SIGMA.i.sub.r Equation 5
[0040] At operation 120 the residual current sample may be
corrected with an offset depending upon its detected full-wave or
half-wave characteristic, in accordance with Equation 6 below. If
full-wave, no gross offset adjustment occurs at operation 120. If
half-wave, then a gross offset adjustment is made to reposition the
flat baseline of the residual current waveform at the mathematical
zero offset positon so that a true RMS calculation can be made.
I.sub.r.rarw.i.sub.r+i.sub.r.DELTA. Equation 6
[0041] At operation 122, the microcontroller 32 may initialize the
starting index (m=0) of i.sub.o buffers for outlet current waveform
samples, where M=number of AC outlets (i.e., number of AC outlets
10a in FIG. 1).
[0042] At operation 124 the microcontroller 32 may then test the
i.sub.o buffer index (m<M) to check if another AC outlet current
sample needs to be processed. If this test produces a "Yes" answer,
then operation 126 is performed, where the AC outlet current sample
is filtered and copied into the buffer i.sub.o in accordance with
Equation 7 below.
s.sub.o[n].rarw.s.sub.o[n]+.beta..sub.1*(s.sub.o[n]-s.sub.o[n-1])
i.sub.o'=s.sub.o[n], i.sub.o''=s.sub.o[n-1] Equation 7
[0043] At operation 128, the microcontroller 32 may then correct
for i.sub.o[m] ADC skew, mean and gain in accordance with Equation
8 below.
I.sub.o[m]=G.sub.o*(i.sub.o'-K.sub.o[m]/(i.sub.o'-i.sub.o'')-i.sub.o.mu.-
[m]) Equation 8
[0044] At operation 130 the microcontroller 32 may then calculate
phase currents (I.sub.L1, I.sub.L2, and I.sub.L3 in this example)
for each phase of the AC input (phases L1, L2, and L3 in this
example) by aggregating each of the M outlet current waveform
samples. This operation is performed in accordance with Equation 9
as shown below.
I.sub.L1=.SIGMA.i.sub.o[m].A-inverted.m.di-elect cons.L1
I.sub.L2=.SIGMA.i.sub.o[m].A-inverted.m.di-elect cons.L2
I.sub.L3=.SIGMA.i.sub.o[m].A-inverted.m.di-elect cons.L3 Equation
9
[0045] At operation 132, the microcontroller 32 may then calculate
the outlet current sample integral in accordance with Equation 10
below. Because a zero-valued integral is expected over a line cycle
of a half-wave symmetrical, periodic waveform, this operation
enables precise correction of creeping offset errors due to small
measurement imprecision.
i.sub.o.SIGMA.[m]=.SIGMA.i.sub.o[m] Equation 10
[0046] At operation 134 the microcontroller 32 may then calculate
the Pearson's terms for the outlet current sample using the
corrected residual current values from operation 120. This
operation is performed in accordance with Equation 11 below:
.SIGMA.I.sub.r*I.sub.o[m], .SIGMA.i.sub.o.sup.2[m] Equation 11
[0047] At operation 136 the outlet index is incremented
(m.rarw.m+1), and operations from 124 may then be re-performed.
[0048] If the i.sub.o buffer index test at operation 124 produces a
"No" answer, this means that an output current sample has been
collected for each one of the AC outlets 10a, and the
microcontroller 32 may then execute operation 180, in which the
microcontroller calculates the Pearson's terms for the phase
current samples, in accordance with Equation 12 below.
.SIGMA.I.sub.r*I.sub.L1, .SIGMA.I.sub.L1.sup.2
.SIGMA.I.sub.r*I.sub.L2, .SIGMA.I.sub.L2.sup.2
.SIGMA.I.sub.r*I.sub.L3, .SIGMA.I.sub.L3.sup.2 Equation 12
[0049] At operation 182 the microcontroller 32 may then calculate
the Pearson's term for the residual current sample
(.SIGMA.I.sub.r.sup.2). The sample index (n.rarw.n+1) may then be
incremented at operation 184, and operation 110 may be repeated,
and operations from 110 may then be re-performed.
[0050] If the s.sub.r buffer index test at operation 110 (n<N)
produces a "No" answer, then at operation 138 the time series has
been completely processed and the microcontroller 32 may calculate
the RMS residual current in accordance with Equation 13 below.
I.sub.rRMS= (.SIGMA.I.sub.r.sup.2/(N-n)) Equation 13
[0051] The microcontroller 32 may then calculate the running mean
of residual and outlet current samples from weighted sample
integrals calculated in operations 118 and 132, as indicated at
operation 140, in accordance with Equation 14 below.
i.sub.r.mu.+=.beta..sub.2*i.sub.r.rho.
I.sub.o.mu.[m]+=.beta..sub.2*i.sub.o.SIGMA.[m] Equation 14
[0052] The microcontroller 32 may then initialize the half/full
wave offset (i.sub.r.DELTA.=0) as indicated at operation 142, and
then perform a test of the positive half-wave residual current at
operation 144 by magnitude comparison of half-wave symmetry, in
accordance with Equation 15 below.
(|i.sub.r -i.sub.r.mu.|)>.beta..sub.3*(|i.sub.r -i.sub.r.mu.|)
Equation 15
[0053] If the test at operation 144 produces a "Yes" answer, then
at operation 146 the microcontroller 32 calculates the positive
offset in accordance with Equation 16 below.
i.sub.r.DELTA.=+(|i.sub.r -i.sub.r.mu.|) Equation 16
[0054] If the test at operation 144 produces a "No" answer or if
operation 146 has been performed, then at operation 148 the
microcontroller 32 tests the negative half-wave residual current in
accordance with Equation 17.
(|i.sub.r |-i.sub.r.mu.)>.beta..sub.3*(|i.sub.r -i.sub.r.mu.|)
Equation 17
[0055] If the test at operation 148 produces a "Yes" answer, then
at operation 150 the microcontroller 32 calculates the negative
offset in accordance with Equation 18 below.
i.sub.r.DELTA.=-(|i.sub.r -i.sub.r.mu.|) Equation 18
[0056] If either operation 148 produces a "No" answer or if
operation 150 has been performed, then at operation 152 the
microcontroller 32 calculates the Pearson's correlation coefficient
for the phase currents in accordance with Equation 19 below.
r.sub.L1=.SIGMA.I.sub.r*I.sub.L1/(.SIGMA.I.sub.L1.sup.2.sub.*.SIGMA.I.su-
b.r.sup.2).A-inverted..PHI..sub.0
r.sub.L2=.SIGMA.I.sub.r*I.sub.L2/(.SIGMA.I.sub.L2.sup.2.sub.*.SIGMA.I.su-
b.r.sup.2).A-inverted..PHI..sub.0
r.sub.L3=.SIGMA.I.sub.r*I.sub.L3/(.SIGMA.I.sub.L3.sup.2.sub.*.SIGMA.I.su-
b.r.sup.2).A-inverted..PHI..sub.0 Equation 19
[0057] At operation 154 the microcontroller 32 calculates the
Pearson's correlation coefficients for the outlet currents in
accordance with Equation 20 below.
r.sub.o.PHI.0.SIGMA.I.sub.r*I.sub.o/(.SIGMA.I.sub.o.sup.2.sub.*.SIGMA.I.-
sub.r.sup.2).A-inverted..PHI..sub.0
r.sub.o.PHI.1.SIGMA.I.sub.r*I.sub.o/(.SIGMA.I.sub.o.sup.2.sub.*.SIGMA.I.-
sub.r.sup.2).A-inverted..PHI..sub.1 wquation 20
[0058] At operation 156 the microcontroller 32 calculates the RMS
Pearson's correlation coefficient for outlet currents, in
accordance with Equation 21 below.
r.sub.oRMS= ((r.sub.o.PHI.0).sup.2+(r.sub.o.PHI.1).sup.2) Equation
21
[0059] At operation 158 the microcontroller 32 then sorts
correlated phase currents (e.g., r.sub.L
=r.sub.L1>r.sub.L2>r.sub.L3) in order to determine the
largest positive max correlation value (r.sub.Lmax).
[0060] With reference to FIG. 4B, it will be understood that the
sequence of operations from 160 through 178 is performed only if
the PDU 10 includes dedicated current metering circuitry to support
measurement of current at individual AC outlets. If this is the
case, then in FIG. 4B at operation 160, the microcontroller 32 then
sorts correlated outlet currents (e.g., r.sub.o
=r.sub.oRMS(1)>r.sub.oRMS(2) . . . >r.sub.oRMS(m)) in order
to determine the largest positive max correlation value
(r.sub.omax).
[0061] At operation 162 the microcontroller 32 then performs a test
of the residual current threshold (I.sub.rRMS>I.sub.rthres). If
this test produces a "Yes" answer, indicating the residual current
threshold has been exceeded, then at operation 163 the value of the
phase variable .alpha. is tested (.alpha.=1) for the presence of a
quarter cycle or quadrature phase shift. If this test produces a
"Yes" answer, then at then at operation 164 the microcontroller 32
tests the maximum phase shift angle (.phi.=15.degree.) to determine
if the phase shift angle has reached 15.degree.. This is done in
order to evaluate the max correlation over a small range, in case a
resistive fault has a small parallel stray capacitance which would
produce a corresponding incremental phase shift of the residual
current waveform with respect to the outlet current waveform. A
small power factor difference between outlet loads would result in
a larger, more discriminating max correlation and therefore even
more reliable outlet detection. If this test produces a "No"
answer, then at operation 166 the microcontroller 32 tests to
determine if the maximum outlet correlation (r.sub.o >r.sub.o ')
has been exceeded. If the test at operation 166 produces a "Yes"
answer, then at operation 168 the microcontroller 32 saves the new
max outlet correlation and its phase shift (.phi.'.rarw..phi. and
r.sub.o '.rarw.r.sub.o ). After operation 168, or if the test at
operation 166 produces a "No" answer, then at operation 170 the
microcontroller 32 increments the phase shift
(.phi..rarw..phi.+1).
[0062] If the residual current threshold test at operation 162
produces a "No" answer, then at operation 172 the microcontroller
32 resets the phase shift max correlation (.phi.=0, .phi.'=0, and
r.sub.o '=0). Once either of operations 170 or 172 is performed, or
if the test at operation 163 produces a "No" answer, then at
operation 174 the microcontroller 32 may alternate quarter-cycle or
quadrature phase shifts (.alpha..rarw.(.alpha.+1) mod 1). The
outlet current waveforms may have non-linear sinusoidal
characteristics and/or be phase shifted for reactive loads,
producing a non-unity power factor. Thus, considering the in-phase
and quadrature calculations in testing for max correlation produces
a larger discriminating max correlation result. After operation 174
is performed, then operations 108 and 110 may be repeated.
[0063] If the test for max phase shift angle at operation 164
produces a "Yes" answer, then at operation 176 the microcontroller
32 sets the phase shift (.phi..rarw..phi.'), and then at operation
178 reports the RMS residual current (I.sub.rRMS), the outlet
correlation (r.sub.o ), and the max outlet correlation (r.sub.o ).
Operations 108 and 110 may then be repeated.
[0064] The PDU 10 and its method of operation as described herein
do not require costly circuitry dedicated per each phase and/or
outlet to directly measure residual current flow. A particularly
important and useful property of the PDU 10 and its method of
operation is that it is invariant to scale or magnitude of the
compared values. A set of correlation values are calculated between
-1 and +1, where a value closest to +1 means the linear dependence
and phase matching is the highest, and therefore, the most likely
path of the residual current flow. A residual current as low as 1
mA can be accurately measured and its associated voltage phase
source and flow through outlet can be reliably determined.
[0065] The various embodiments of the present disclosure provide a
significant advantage over prior art methods that can merely
measure for a residual current condition. The Vertiv MPH2 rack PDU
controller firmware, for example, may be used to support the
methodology described herein.
[0066] The various embodiments and methodology of the present
disclosure presents a lower cost solution when compared to directly
measuring differential currents at each phase and/or outlet. The
various embodiments and methodology of the present disclosure offer
highly sensitive detection of small phase differences of 0.5% power
factor between compared current waveforms. The methodology of the
present disclosure automatically discriminates the most likely
residual current path external to the unit, so the powered device
causing the fault can be quickly removed to minimize down times and
reduce the associated costs, rather than being required to manually
identify the residual current path by a trial and error method.
[0067] The various embodiments of the PDU 10 and its method of
operation further facilitate a preventative maintenance program
through the ability to obtain additional indirect information about
the health of insulation system(s) associated with electrical
cabling being used, and thus may help prevent unanticipated
downtime of all types of electrically powered devices. The PDU 10
and its method of operation, by its ability to quickly detect
residual currents and identify specific AC outlets that such
currents are associated with, may also provide a measure of fire
protection for a facility in which the present disclosure is
used.
[0068] While various embodiments have been described, those skilled
in the art will recognize modifications or variations which might
be made without departing from the present disclosure. The examples
illustrate the various embodiments and are not intended to limit
the present disclosure.
[0069] Therefore, the description and claims should be interpreted
liberally with only such limitation as is necessary in view of the
pertinent prior art.
* * * * *