The present invention is generally directed to a system and method for confirming the detection of a fire in a monitored region. More particularly, the present invention is directed to a fire detection system including an operationally redundant flame sensor and logic for discriminating between a fire event and a false fire event in a monitored region.
Optical fire detection systems including multiple flame sensors are known in the art. Exemplary systems are described in U.S. Pat. Nos. 6,518,574, 5,373,159, 5,311,167, 5,995,008 and 5,497,003. The flame sensors in such systems are typically equipped with a radiation detector and a unique optical filter that ranges from the ultraviolet to the infrared to allow for the measurement of the spectral content of objects within the flame sensor's field of view (FOV). By judiciously choosing the type of radiation detector, e.g., a Geiger-Mueller, a silicon, a pyroelectric, etc., in combination with the appropriately-specified optical filter for each radiation detector and electronically combining the output signals from the flame sensors, a flame can be discriminated from other innocuous sources. In this manner, based on the emissive characteristics of a flame and the anticipated false fire alarm sources, e.g., a radiant heater, cigarette, cigar, etc., within a monitored region a fire detection system can be developed by selecting the appropriate combination of radiation detectors and optical filters so that the anticipated false alarm sources does not cause a false alarm. In fire detection systems of this type, a fire alarm condition is identified and reported by the system when the sensed source of radiation appears to be spectrally similar to a flame as defined by the system designer and determined by the designer's choice of radiation detectors, optical filters and electronic combination of the resulting signals from the radiation detectors.
A shortcoming of optical fire detection systems of this type is manifested when a spatially small source of radiation is brought in close proximity to the flame sensors. That is because there is an inherent spatial disparity between the multiple flame sensors. This spatial disparity often results from the use of the discrete radiation detectors and can be directly measurable as a physical distance. Alternatively, this spatial disparity can result from the use of refractive, diffractive or reflective optical elements.
In particular, the radiation detector of each flame sensor has its own field of view that may not significantly overlap with that of an adjacent radiation detector until an object is several inches away from the radiation detector. If the spatially small radiation source is brought closer than the common field range of the radiation detectors, i.e., the range at which FOV of the radiation detectors overlap, a significant chance exists that one detector will observe more of the radiation source than any other radiation detector. As a result, the radiation detector that observed more of the radiation will have the chance to collect more radiation from the radiation source depending on the spectral characteristics of the radiation source and the optical filter associated with that particular radiation detector. Consequently, the electronic output from the flame sensor including that particular radiation detector could be skewed relative to the other flame sensors. Once received and analyzed, the information transmitted in the electronic output of that flame sensor could cause the fire detection system to trigger a false alarm.
The present invention is directed to a system for confirming the detection of a fire using a fire detection system having a plurality of flame sensors each equipped with a radiation detector and an optical filter having a spectral transmission characteristic in which at least one optical filter is redundant to at least one other optical filter. The present invention is further directed to a method for testing for the condition in which a spatially small source of radiation is in close proximity to a flame detector so that the multiple radiation sensors of the detector each view different spatial extents of the object so that a false alarm is avoided. As such, the present invention is particularly suited for detecting fires where low false alarms rates are required and the distance and size of the fire varies over a wide range.
According to one aspect of the invention there is disclosed a system for discriminating between a fire event and a false fire event. The system includes a first radiation detecting structure configured for transmitting a first signal and a second radiation detecting structure being operationally redundant to the first radiation detecting structure and configured for transmitting a second signal. A computer-based processor is provided for receiving and analyzing the first signal and at least one other signal for producing a first output, and comparing the first output to a predetermined fire condition for determining whether the first output indicates a fire. The computer-based process is further configured for receiving and analyzing the second signal and the at least one other signal for producing a second output, and comparing the first output to the second output. In the event the first output and the second output satisfy a predetermined criteria for similarity or the presence of fire, a fire alarm command signal is transmitted to a fire extinguishing system for extinguishing the fire. However, if the first and second output are not sufficiently similar or they do not meet the predetermined fire-presence criteria, the system will not transmit the fire alarm command signal, even if the first output indicates the presence of a fire event.
According to another aspect of the invention, there is disclosed a method for discriminating between a fire event and a false fire event in a monitored region. The method includes positioning a plurality of flame sensors within the monitored region, wherein the plurality of flame sensors includes at least a first radiation sensor and a second radiation sensor that is operationally redundant to the first radiation sensor. Upon detection by the plurality of radiation sensors of a potential fire event, the plurality of flame sensors transmit signals to a computer based processor. The processor calculates a first output and a second output based upon the signals. The first output is calculated using a first signal transmitted by the first sensor absent a second signal transmitted by the second sensor. The second output is calculated using the second signal absent the first signal. In the event the first output indicates a fire event, the first output and the second output are compared to one another for similarity. If the first and second output are not sufficiently similar, the first output is ignored and no fire alarm command is transmitted to a fire extinguishing system. On the other hand, if the first output indicates a fire event and the first and second outputs are sufficiently similar, the fire alarm command is sent to the fire extinguishing system, and the fire is extinguished.
According to yet another aspect of the invention, there is disclosed a method of making a system for discriminating between a fire event and a false fire event. The method includes operatively coupling a plurality of radiation sensors to a computer based processor, and configuring a first radiation sensor of the plurality of radiation sensors to be operationally redundant to a second radiation sensor of the plurality of radiation sensors. The method further includes configuring the computer based processor for receiving and analyzing signals generated by the plurality of radiation sensors upon detection thereby of a potential fire event, calculating a first output using a first signal transmitted by the first sensor absent a second signal transmitted by the second sensor, and calculating a second output using the second signal absent the first signal. The processor is further configured for transmitting a fire alaan command signal to a fire extinguishing system when the first output and the second output satisfy a predetermined criteria for similarity or a predetermined fire-presence criteria.
A process and system for detecting sparks, flames or fire in accordance with a preferred embodiment of the present invention is described herein. It should be noted that the terms “fire sensor,” “flame sensor” and “radiation sensor” “ are used interchangeably in the present text and refer generally to any sensor for detecting sparks, flames, or fires, including explosive type fires or fireballs and other dangerous heat-energy phenomena.
A problem addressed by the present invention is that fire detection systems often produce inconsistent results for fires occurring at different points in the fields of view of the radiation detectors of the flame sensors of the system. This problem arises due to the interference filters employed with the radiation detectors to transmit radiation in the desired spectral bands. The passbands of the interference filters vary with the angle at which the radiation from a fire is incident on the filter. As a result, the amount of radiation sensed is dependent on the angle of incidence, and, in consequence, a particular flame sensor may not be as effective at detecting a fire when the fire is positioned off-axis from the radiation detector of the flame sensor. Thus, optical flame detection systems utilizing multiple radiation sensors including ultraviolet, visible and infrared radiation detectors, each equipped with unique optical filters for measuring the spectral signature of the objects in the field of view, work well at distances where the individual fields of view overlap. However, at close range, the fields of view do not overlap and one radiation detector may see more of the object than another.
To illustrate this phenomenon, at
To solve this problem, the present invention relies upon the addition of an operationally redundant flame sensor to the bank of sensors so that if a fire is detected without including the operationally redundant radiation sensor in the calculation, the algorithm can switch to the operationally redundant sensor to check for confirmation of a fire. Due to the spatial separation of the operationally redundant sensor and the mimicked sensor, and if the object is small and close, a different result will be obtained with the operationally redundant sensor being used in the calculation compared to the primary sensor that is associated with or mimicked by the operationally redundant sensor. Herein, by “operationally redundant sensor,” “operationally redundant flame sensor” and “operationally redundant radiation sensor” it is meant a sensor that operates substantially similar to another sensor within the flame detection system, either as an exact copy or through manipulation of the sensor material, sensor temperature, sensor wavelength filter, sensor preamplifier, sampling mechanism (if so equipped), and/or the software algorithm (if so equipped) so that it could be used as an effective replacement of the other sensor, i.e., the mimicked sensor. Thus, the operationally redundant sensor can be identical in function and structure to the mimicked sensor or it can have a different detector material and a different filter so long as it is substantially similar in performance to the mimicked sensor. For example, many detector materials overlap when considering their spectral response so that a silicon photodetector—a visible spectrum sensor—equipped with a unique optical filter, and a thermopile detector—an infrared spectrum sensor—equipped with its own unique optical filter could be configured through preamplifiers, calibration and software gains to perform substantially similar to one another.
Referring to
Optical sensors 101, 103, 105 and 107 each include a respective radiation detector 119 which can be selected, for example, from a Geiger-Mueller radiation detector, a silicon radiation, a pyroelectric radiation detector, a thermopile detector, a lead sulfide detector, a lead selenide detector, an indium antimonide detector, etc. Based on the emissive characteristics of a flame, the type of radiation detector 119 and the anticipated false fire alarm sources, an appropriately-specified optical filter 121 is combined with each radiation detector 119. Thus, for example, depending on the radiation detector type, each radiation detector 119 of sensors 101, 103, 105 and 107 can combined with an optical filter 121 selected from an ultraviolet band spectra filter, a visible band spectra filter, a near band infrared spectra filter, a mid band infrared spectra filter, a fax band infrared spectra filter, a water band spectra filter or a carbon dioxide band spectra filter. Preferably, sensors 101, 103, 105 are configured to detect radiation in the ultraviolet, visible and infrared portions of the electromagnetic spectrum, respectively. Sensor 107 is the operationally redundant sensor.
Referring to
To illustrate the spatial disparity of operationally redundant sensor 107 and mimicked sensor 101 relative to sensors 103 and 105, there is depicted at
Specifically, for example, in
The detection algorithm executed by the computer program of the present invention is substantially the same as the detection algorithm in current fire detection systems with the exception that when a flame is detected, the algorithm of flame detection apparatus 100 performs calculations twice, once including only the signals of sensors 101, 103 and 105 and once more including only the signals of sensors 103, 105 and 107. More particularly, referring to
In an alternative embodiment, rather than compare the first and second outputs to a predetermined fire-presence criteria, the first output of the algorithm is compared to the second output of the algorithm. In this instance, the second output of the algorithm must be within a predetermined percentage, e.g., 5%, of the first output for an alarm to be reported to the fire extinguishing system. Otherwise, no instructions are sent to the extinguishing system. This allows for the fact that some algorithms have a range over which the algorithm output is defined as a fire.
A fire detection system having an operationally redundant flame sensor is described where the redundant flame sensor is structurally different from but substantially similar in performance to the flame sensor it mimics. In particular, the fire detection system includes three optical flame sensors. One of these sensors is chosen to be mimicked by a fourth optical flame sensor. In theory, any one of the three flame sensors could be chosen to be mimicked. However, it is preferred that the flame sensor that, in general, has the highest signal to noise ratio is mimicked. This flame sensor can be mimicked using various approaches that are functionally different and then implementing some form of compensation to make the operationally redundant flame sensor operate in a substantially similar fashion to the flame sensor chosen for mimicry.
In this manner, a Geiger-Mueller sensor and a UV-enhanced Silicon sensor, or a Lead-Selenide sensor and a thermopile sensor could be made operationally redundant with the use of appropriate filters and/or electronic circuits and/or software algorithms that correct for any operational difference. Although the particular performance of the two flame sensors would be somewhat different in terms of their detectivity (D*), signal to noise ratio, and noise equivalent power, the two would operate over the same wavelength and give nearly the same output in the presence of a flame when used with the corrective filters, circuits, and/or algorithms.
Having given an example of two operationally redundant flame sensors that are functionally different, examples of how the flame sensors could be used to reject a false alarm are provided. In the first method, one operationally redundant flame sensor is considered to be the primary flame sensor while the other is considered to be the secondary sensor. Assuming multiple sensors, the flame-presence criteria are calculated without using the secondary operationally redundant flame sensor. If the criteria are satisfied, the criteria are calculated a second time without using the primary operationally redundant flame sensor, substituting the secondary flame sensor for the primary flame sensor. If the flame-presence criteria are confirmed in both cases, a fire alarm is announced.
In the second method, the calculations for the flame-presence criteria are performed using the primary operationally redundant flame sensor. Rather than go through the same calculations a second time, the primary and secondary operationally redundant flame sensors are simply compared to each other. A second flame-presence criteria is computed, which may be a simple ratio between the primary and secondary operationally redundant flame sensors, and if the second flame-presence criteria is satisfied subsequent to the first flame-presence criteria then a fire is announced. In both methods, any corrective filters, circuit, and/or algorithms are assumed to be in place so that the exact method of correction is not important.
As will be apparent to one skilled in the art, various modifications can be made within the scope of the aforesaid description. Such modifications being within the ability of one skilled in the art form a part of the present invention and are embraced by the claims below.
Number | Name | Date | Kind |
---|---|---|---|
3255352 | Johnston | Jun 1966 | A |
3741309 | McCulloch | Jun 1973 | A |
4097851 | Klein | Jun 1978 | A |
4151522 | Yamauchi | Apr 1979 | A |
4647785 | Morita | Mar 1987 | A |
4691196 | Kern et al. | Sep 1987 | A |
4697172 | Kimura | Sep 1987 | A |
4701624 | Kern et al. | Oct 1987 | A |
4749987 | Ishii | Jun 1988 | A |
4764758 | Skala | Aug 1988 | A |
4803469 | Matsushita | Feb 1989 | A |
4866420 | Meyer, Jr. | Sep 1989 | A |
4871999 | Ishii et al. | Oct 1989 | A |
4963856 | Yukitomo | Oct 1990 | A |
4965453 | Hoschette et al. | Oct 1990 | A |
5153563 | Goto et al. | Oct 1992 | A |
5237512 | Davidson | Aug 1993 | A |
5311167 | Plimpton et al. | May 1994 | A |
5373159 | Goldenberg et al. | Dec 1994 | A |
5497003 | Baliga et al. | Mar 1996 | A |
5552763 | Kirby | Sep 1996 | A |
5592147 | Wong | Jan 1997 | A |
5594417 | Morita | Jan 1997 | A |
5612537 | Maynard et al. | Mar 1997 | A |
5625342 | Hall et al. | Apr 1997 | A |
5691703 | Roby et al. | Nov 1997 | A |
5691704 | Wong | Nov 1997 | A |
5696503 | Nasburg | Dec 1997 | A |
5767776 | Wong | Jun 1998 | A |
5798700 | Wong | Aug 1998 | A |
5804825 | Schuler | Sep 1998 | A |
5838242 | Marsden | Nov 1998 | A |
5850182 | Schuler | Dec 1998 | A |
5936531 | Powers | Aug 1999 | A |
5949235 | Castleman et al. | Sep 1999 | A |
5966077 | Wong | Oct 1999 | A |
5995008 | King et al. | Nov 1999 | A |
6031223 | Morishita | Feb 2000 | A |
6046452 | Castleman et al. | Apr 2000 | A |
6057549 | Castleman | May 2000 | A |
6064064 | Castleman | May 2000 | A |
6071114 | Cusack et al. | Jun 2000 | A |
6078050 | Castleman | Jun 2000 | A |
6135760 | Cusack et al. | Oct 2000 | A |
6147598 | Murphy et al. | Nov 2000 | A |
6150659 | Baliga et al. | Nov 2000 | A |
6153881 | Castleman | Nov 2000 | A |
6166647 | Wong | Dec 2000 | A |
6239435 | Castleman | May 2001 | B1 |
6373393 | Matsukuma et al. | Apr 2002 | B1 |
6448888 | Horner et al. | Sep 2002 | B1 |
6518574 | Castleman | Feb 2003 | B1 |
6677590 | Nakauchi et al. | Jan 2004 | B2 |
6727511 | Cusick et al. | Apr 2004 | B2 |
6856252 | Pfefferseder et al. | Feb 2005 | B2 |
6927394 | Parham et al. | Aug 2005 | B2 |
7009497 | Nicoletti et al. | Mar 2006 | B2 |
7119697 | King | Oct 2006 | B2 |
7319228 | Zhevelev et al. | Jan 2008 | B2 |
7327247 | Tice | Feb 2008 | B2 |
7408453 | Breed | Aug 2008 | B2 |
7421321 | Breed et al. | Sep 2008 | B2 |
7504633 | Zhevelev et al. | Mar 2009 | B2 |
7542522 | Ross et al. | Jun 2009 | B2 |
7551096 | Tice | Jun 2009 | B2 |
7573032 | Zhevelev et al. | Aug 2009 | B2 |
7580782 | Breed et al. | Aug 2009 | B2 |
7603894 | Breed | Oct 2009 | B2 |
7638770 | Serero et al. | Dec 2009 | B2 |
7642924 | Andres et al. | Jan 2010 | B2 |
7663502 | Breed | Feb 2010 | B2 |
7684029 | Tan et al. | Mar 2010 | B2 |
7705310 | Zhevelev et al. | Apr 2010 | B2 |
7760080 | Breed et al. | Jul 2010 | B2 |
7786864 | Shostak et al. | Aug 2010 | B1 |
7786877 | Hou | Aug 2010 | B2 |
7880594 | Breed et al. | Feb 2011 | B2 |
7889096 | Breed | Feb 2011 | B2 |
7920102 | Breed | Apr 2011 | B2 |
7969296 | Stell | Jun 2011 | B1 |
7991187 | Hou | Aug 2011 | B2 |
8077046 | Wong | Dec 2011 | B1 |
8138478 | Zhevelev et al. | Mar 2012 | B2 |
8169311 | Breed | May 2012 | B1 |
20020011570 | Castleman | Jan 2002 | A1 |
20020139942 | Cusick et al. | Oct 2002 | A1 |
20020174367 | Kimmel et al. | Nov 2002 | A1 |
20030009270 | Breed | Jan 2003 | A1 |
20030174056 | Harshaw | Sep 2003 | A1 |
20030214583 | Sadok | Nov 2003 | A1 |
20040130442 | Breed et al. | Jul 2004 | A1 |
20040183667 | Nicoletti et al. | Sep 2004 | A1 |
20040216895 | Boyce et al. | Nov 2004 | A1 |
20050012626 | Owrutsky et al. | Jan 2005 | A1 |
20050125117 | Breed | Jun 2005 | A1 |
20050195086 | King | Sep 2005 | A1 |
20060176169 | Doolin et al. | Aug 2006 | A1 |
20060244581 | Breed et al. | Nov 2006 | A1 |
20070000711 | Yamaguchi et al. | Jan 2007 | A1 |
20070018106 | Zhevelev et al. | Jan 2007 | A1 |
20070029486 | Zhevelev et al. | Feb 2007 | A1 |
20070057781 | Breed | Mar 2007 | A1 |
20070126561 | Breed | Jun 2007 | A1 |
20070139216 | Breed | Jun 2007 | A1 |
20070145277 | Zhevelev et al. | Jun 2007 | A1 |
20070152156 | Zhevelev et al. | Jul 2007 | A1 |
20070205881 | Breed | Sep 2007 | A1 |
20080030352 | Shaw | Feb 2008 | A1 |
20080036593 | Rose-Pehrsson et al. | Feb 2008 | A1 |
20080042410 | Breed et al. | Feb 2008 | A1 |
20080061984 | Breed et al. | Mar 2008 | A1 |
20080129475 | Breed et al. | Jun 2008 | A1 |
20080230701 | Serero et al. | Sep 2008 | A1 |
20090014654 | Zhevelev et al. | Jan 2009 | A1 |
20090146063 | Zhevelev et al. | Jun 2009 | A1 |
20090315722 | Hou | Dec 2009 | A1 |
20100013619 | Zimmer | Jan 2010 | A1 |
20100085199 | Gonzales | Apr 2010 | A1 |
20100207754 | Shostak et al. | Aug 2010 | A1 |
20100283626 | Breed | Nov 2010 | A1 |
20110018726 | Gonzales | Jan 2011 | A1 |
20110304728 | Owrutsky et al. | Dec 2011 | A1 |
20120001760 | Harchanko | Jan 2012 | A1 |
Number | Date | Country |
---|---|---|
588753 | Mar 1994 | EP |
10221163 | Aug 1998 | JP |
2004005754 | Jan 2004 | JP |
Number | Date | Country | |
---|---|---|---|
20120001760 A1 | Jan 2012 | US |