JPH05166078A - Fire alarm - Google Patents

Abstract

PURPOSE:To change the weighting value to be applied for a function value obtained from the processing by discriminating how the fire information as the result of deduction obtained from the processing rule meets the situation and correcting the information if it does not meet. CONSTITUTION:The system consists of a fire receiver RE and N-number of analog fire sensors DE1 to DEN which are connected with the fire receiver RE through a transmission line L1. A sensor SI1 sensing the number of ventilation is shown through a transmission line L2 and a sensor SI2 sensing the number of person through a transmission line L3. The fire receiver RE is provided with microprocessors MPU 1, ROMs 11 and 12, etc. An update means updates the function value stored in a function value storage means for each rule based on the function value stored in a function value storage means and the desired value decided by a decision means. The processing means performs the weighting for the function value which is obtained using the updated weighting value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is based on detection information relating to physical quantities of fire phenomena such as smoke, heat, gas, odor, and / or environmental information such as room size, number of people, ambient temperature, etc. The present invention relates to a fire alarm device for obtaining at least one fire information such as.

[0002]

BACKGROUND OF THE INVENTION There are various methods for detecting a fire. For example, considering the simplest method of making a fire decision based on the sensor level, that is, the information detected by the fire detector, a fire signal is output when the sensor level exceeds a certain level. .. In this case, the fire signal output from the fire detector is uniquely determined depending on whether or not it exceeds a predetermined level, and it cannot be said that the various environmental conditions are sufficiently taken into consideration. In addition to the detection information from the fire detector, it is also considered to collect environmental information and to make a comprehensive fire judgment from the detection information and the environmental information, but ambiguous environmental information In consideration of the above, it has not been possible to obtain a sufficiently reliable fire signal, and from a human sense, there are many cases in which it is not always possible to conclude that a fire is present even if the fire signal is on.

In order to solve such a problem, Japanese Patent Laid-Open No. 2-195495 filed by the applicant of the present invention discloses a method of collecting information including environmental information, which is more reliable than the conventional method. A fire alarm device is disclosed which is capable of making a more reliable fire determination by processing in step 1. That is, in the fire alarm device, various collected information relating to a fire phenomenon and processing information from the collected information are obtained, and a function for the fire information is defined for each of the obtained collected information and processing information. At the same time, at least one rule of processing to be performed using the function is defined, and the obtained information is calculated based on the rule of each processing and each corresponding function used in the rule of each processing. Fire information is obtained by performing processing, obtaining a function value for each rule of each processing, and obtaining the center of gravity of the obtained function value. The fire information thus obtained is compared with, for example, a reference value or the like to determine whether or not a fire has occurred.

In such a fire alarm device, since the obtained function values are effectively comprehensively judged, highly accurate fire information can be obtained, and therefore good fire monitoring can be performed. However, since a large number of rules are used in inference using rules for such processing, not all of the rules are all working effectively because of, for example, changes over time or changes in the environment. In addition, even if the rule of the process is working effectively, that is, even if the function value obtained from the rule of the process greatly contributes to the total judgment value, the total judgment value itself does not correspond to the actual situation. It is possible that it has become something that cannot be done. Therefore, if you can examine the data of the fire detector or environment sensor used for past inference and correct the inference result by inputting the desired inference value in each case, in order to improve the reliability of inference Very advantageous.

[0005]

SUMMARY OF THE INVENTION Therefore, the object of the present invention is to such an extent that the rules of processing currently used in a fire alarm device as disclosed in the above-mentioned publication contribute to fire information as an inference result. Accordingly, the function value obtained from the processing rule is weighted, and the fire information as the inference result obtained from the currently used processing rule corresponds to the actual situation. It is possible to change the weighting value attached to the function value obtained from the rule of each processing so that it is possible to judge whether it is correct, and if it is not correct, it can be corrected. is there.

Therefore, according to the present invention, in order to obtain fire information based on various information relating to a fire phenomenon, various collected information relating to the fire phenomenon and processing information obtained from the gathered information are obtained. An information acquisition unit, and a definition unit that defines a function for the fire information for each piece of information obtained by the information acquisition unit and also defines at least one processing rule to be performed using the function. A weighting value storage means for storing a weighting value for each rule of each process defined in the defining means, and a rule of each process and each corresponding function used in the rule of each process. , Processing the information obtained by the information acquisition means to obtain a function value for each rule of each processing,
A weighting function value is obtained by weighting the obtained function value with the corresponding weighting value stored in the weighting value storage means, and the weighting function value is comprehensively processed. Processing means for obtaining the fire information, function value storage means for storing a function value for each rule of each processing, and a desired value of the fire information to be obtained from the function value stored in the function value storage means Based on the function value stored in the function value storage means, and the desired value determined by the determination means, in the weighted value storage means for each rule of each processing. There is provided a fire alarm device comprising: an updating unit that updates the stored weight value. Preferably, the determining means is
Display means for displaying the function value stored in the function value storage means when requested, and inputting the desired value of the fire information based on the function value displayed on the display means. And an input means that enables it. Further, the updating means stores the weighted value for each rule of the processing based on the ratio between the function value stored in the function value storage means and the desired value determined by the determining means. Updating the weighting value stored in the means.

[0007]

The information relating to the fire phenomenon obtained by the information acquisition means includes not only the physical quantity detection information based on the fire phenomenon but also various environmental information such as the size of the room and the ambient temperature that affect the detection information. Also, so-called processing information such as the temporal change amount of these information and the integrated value are included.

For example, the definition means, which may be a storage means, defines a function of acquired information vs. fire information for each information obtained by the information acquisition means by a method such as an expression or a table, and processes the information. When performing the, which acquisition information versus fire information function (one or more functions)
At least one (usually more than one) regarding whether to use
The rules for processing are also defined.

The weight value storage means stores the weight value to be given for each rule of each process defined by the definition means.

The processing means processes the information obtained by the information acquisition means on the basis of the rule of each process and the corresponding function used in the rule of each process, and the function value for each rule of each process. To get Next, the obtained function value is weighted with the corresponding weight value stored in the weight value storage means. Finally, the processing means comprehensively processes the weighted function values obtained for each processing rule to obtain fire information such as fire accuracy.

A function value storage means for storing the function value obtained for each rule of each process when a fire occurs is also provided for later use.

The desired value of the fire information such as the fire accuracy to be obtained from the function value stored in the function value storage means can be determined by the determination means at any time.
As an example, the determination means displays the function value stored in the function value storage means when requested, whereby the operator or the like inputs the desired value of the fire information based on the displayed content. And it is possible to determine the input desired value as new fire information.

The updating means stores the weighting values stored in the weighting value storage means for each rule of each process based on the function value stored in the function value storage means and the desired value determined by the determining means. Update the value. As a result, the processing means weights the obtained function value using the updated weighting value from the next time.

[0014]

EXAMPLE An example of the present invention will be described below.
FIG. 1 shows that a sensor level of an analog physical quantity based on a fire phenomenon detected by each fire detector is sent to a receiving means such as a fire receiver RE or a repeater, and the receiving means is based on the collected sensor level. FIG. 9 is a block circuit diagram when the present invention is applied to a so-called analog fire alarm device that makes a fire decision by using the method. Of course, the present invention is also applicable to an on / off type fire alarm device in which each fire detector makes a fire decision and sends only the result to the receiving means.

In FIG. 1, RE is a fire receiver, DE ₁
～DE _N is, for example, a fire detector of the N analog which is connected to the fire receiver RE through a transmission line L _1, such as a pair of power and signal lines, one No.1 fire detector that The internal circuit is shown in detail only for DE ₁ .

The fire receiver RE also has a transmission line L
_The ventilation rate sensor is shown connected via 2 and the occupancy sensor is connected via a transmission line L ₃ . The ventilation frequency sensor and the number of people sensor are arranged, for example, in each room, and are provided for each fire detector, or one for some fire detectors. Whether the fire detector is related to the ventilation frequency sensor or the number of people sensor can be known from the correspondence table or the like. In FIG. 1, only the ventilation frequency sensor SI ₁ and the people sensor SI ₂ associated with the _first fire detector DE ₁ are shown.

In the fire receiver RE, the MPU 1 is a microprocessor, the ROM 11 is a program storage area for storing programs related to the operation of the present invention described later, and the ROM 12 is a storage area for individual rules, ROM 1
3 is the definition function of the individual rule, that is, the sensor level
A storage area for the definition function that stores various definition functions such as a definition function for the SLV, a definition function for the integral value, and a definition function for the time, the ROM 14 is a storage area for the weight value at the time of initializing the rule, and the RAM 11 is It is a storage area for the sensor level, including the area for each fire detector to store the sensor level collected for each fire detector, and multiple storage areas for each fire detector to obtain the difference value described later. Multiple sensor levels collected over time are stored for each fire detector. RAM 12 is a storage area for the integrated value, RAM 13 is a storage area for the number of rules to be used, RAM 14 is a storage area for the total definition function value, RAM 15 is a working area, and RAM 16 is a storage area for the rule function value. area, RAM17, the storage area for heavy bid of rules, RAM18, the storage area for the total value at the time of the fire, RAM19, the storage area for the desired overall value, SW ₁
Is a data display switch, SW ₂ is a learning switch, D
P is a display such as a CRT, OP is an operation unit, CL is a clock, TRX11 is a fire receiver RE, a fire detector DE ₁
Connecting ～DE _N, the signal transmitting and receiving section including a straight-parallel converter and parallel-serial converter, and the like, TRX12, the signal transmitting and receiving unit for connecting the ventilation rate sensor described above, TRX1
Reference numeral 3 is a signal transmission / reception unit for connecting the above-mentioned number sensor, and IF11 to IF18 are interfaces.

Further, in the fire detector DE ₁ , the MPU
2 is a microprocessor, ROM 21 is a program storage area, ROM 22 is a self-address storage area,
The RAM 21 is a work area, and FS is a fire phenomenon detecting means for detecting a physical quantity such as heat, smoke, gas or odor based on the fire phenomenon.
It has a sample and hold circuit, an analog / digital converter, and the like. TRX21 is a signal transmitting / receiving unit similar to TRX11, and IF21 and IF22 are interfaces.

The storage area ROM 13 for defining functions in the fire receiver RE stores various defining functions in the form of expressions or tables as shown in the examples in FIGS. 2 (a) to 2 (f). 2 (a) to 2 (f), the fire accuracy as fire information (vertical axis) for various acquired information, that is, input information (horizontal axis) is shown.

In FIG. 2 (a), the definition function F ₁ (SLV) for the sensor level SLV from the sensor section FS for detecting a fire phenomenon as input information, that is, the fire accuracy is shown as a value of 0 to 1. 2 (b) shows the definition function F ₂ (t) of the fire accuracy with respect to the time t after the sensor level exceeds the predetermined level LV ₁ , and FIG. 2 (c) Shows the definition function F ₃ (△ SLV) of the fire accuracy for the sensor level difference value △ SLV, and Fig. 2 (d) shows the definition of the fire accuracy for the integrated value ΣSLV of the sensor level. Function F
₄ (ΣSLV) is shown. In Fig. 2 (e), the ventilation rate /
When time influences the fire judgment value, the definition function F ₅ (k) of the fire accuracy with respect to the ventilation rate k / hour is shown as the environmental information, and (f) of FIG. 2 shows the environmental information. For example, the definition function F of fire accuracy for the number of people p in the room
₆ (p) is shown.

Various other definition functions can be stored in the definition function storage area ROM 13, and can be taken out and used as needed.

The storage area ROM 12 for individual rules in the fire receiver RE stores the content of the rules of the processing to be performed for each fire detector and the address of the definition function used for the rules. The rule of each process is 1
When one or two or more types of acquisition information are given, the relationship with the output information to be obtained is defined. For example, as an example of a processing rule that defines the relationship with output information that should be obtained when one type of acquisition information is given, for example, rule a
There is ~ f.

Rule a: If the sensor level SLV = X, the fire accuracy should be F ₁ (X) as fire information, and it should be used as fire information by using the definition function starting from the address AD ₁ in the storage area ROM 13. The fire accuracy is determined. Rule b: If the time t after the sensor level SLV exceeds a predetermined level LV ₁ t = T, the fire accuracy should be F ₂ (T), and the address AD ₂ in the storage area ROM 13
The fire accuracy as fire information is determined using a definition function starting from. Rule c: Difference of sensor level SLV over a fixed time ΔS
If LV = Y, then the fire probability should be F ₃ (Y), and the fire probability as fire information is determined using the definition function starting from the address AD ₃ in the storage area ROM 13. Rule d: If the integrated value after exceeding the predetermined level LV ₁ of the sensor level SLV is ΣSLV, the fire accuracy should be F ₄ (M), and the address AD in the storage area ROM 13
The fire accuracy as fire information is determined using the definition function starting from ₄ . Rule e: Number of ventilations k in the room where the fire detector is installed
If (/ hour) = K, then the fire probability should be F ₅ (K), and the fire probability is determined as fire information using the definition function starting from the address AD ₅ in the storage area ROM 13. Rule f: Number of people in the room where the fire detector is installed p = P
If so, the fire probability should be F ₆ (P), and the fire probability as the fire information is determined using the definition function starting from the address AD ₆ in the storage area ROM 13. Etc.

In the processing rule described above, one or two or more are defined for each fire detector, and the storage area R
It is stored in each fire detector area in the OM 12. For example, for rule ₁ fire detector DE ₁ , the above rule a,
Assuming that the rules described in b, d, and e are used, the rules a, b, d, and e are stored in the area for the first fire detector DE ₁ in the storage area ROM 12, and the storage area ROM 11 Based on the rule, the program to be described later stores the output information F ₁ (SLV) for each rule using the definition function of FIG. 2 stored in the storage area ROM 13.
F ₂ (T), F ₄ (ΣSLV), and F ₅ (K) are obtained. Then, the output information for each rule is weighted using the corresponding weight value stored in the storage area RAM 17, and the center of gravity of the weighted output information is obtained. As an operation for obtaining this center of gravity, in the present embodiment, the value obtained by summing the definition function values obtained for each rule is divided by the sum of weighted values, that is, the average value between the definition functions is obtained. ..

[0025]

[Equation 1] F = {ω ₁ · F ₁ (SLV) + ω ₂ · F ₂ (T) + ω ₃ · F ₄ (ΣSLV) + ω ₄ · F ₅ (K)} / Tr (Tr = ω ₁ + ω ₂ + ω ₃ + ω ₄ )

The average value F of the definition function thus obtained represents fire information, that is, the fire probability in this embodiment, and the fire probability is compared with a reference value or the like to determine whether or not a fire has occurred. Judgment is made.

After the fire information such as the fire accuracy is obtained in this way, the present invention further displays the obtained fire information and the output information for each rule when necessary, and the operator or the like It is possible to input desired fire information while considering the displayed content and the actual situation, and change the weighting value to be attached to each rule based on the inputted desired fire information so as to match the actual situation. I am able to do it.
In this embodiment, such display and change of the weight value are performed after it is determined that a fire has occurred and there is a first fire display.

The operation of FIG. 1 will be described below with reference to the flow charts of FIGS.

The fire receiver RE initially stores the weighting value at the time of initialization for each rule of the fire detectors Nos. 1 to N stored in the storage area ROM 14 at the time of initial setting in the storage area RAM 17. The data is stored (step 100), and thereafter, data is collected in order from the _{1st to} Nth fire detectors DE _{1 to} DE _N and signal processing is performed. The signal processing for the _first fire detector DE ₁ will be described below. After sending the data collection command to No. 1 fire detector DE ₁ ,
When the sensor level SLV ₁ is read from the fire detector DE ₁ (step 106), the sensor level SLV ₁ is compared with the predetermined level LV ₁ (step 108) and the sensor level SLV ₁ is determined to be the predetermined level. If it is smaller than the level LV ₁ (N in step 108), no further signal processing operation is performed for the _first fire detector DE ₁ and the sensor level
After the variable T ₁ for counting the time when SLV ₁ is equal to or higher than the predetermined level LV ₁ is cleared (step 110), the signal processing operation for the next fire detector DE ₂ is performed (N in step 166). , And step 104).

The sensor level SLV ₁ is the predetermined level LV ₁
If it is above (Y in step 108), the sensor
Level SLV ₁ is the memory area RAM11 for sensor level
(Step 114) and the sensor level
The variable T ₁ for counting the time when SLV ₁ is equal to or higher than the predetermined level LV ₁ is incremented by 1 (step 112), and then the signal processing operation for the _first fire detector DE ₁ is continued. ..

First, a collecting and / or computing operation is performed to obtain information used to perform a signal processing operation. In the case of the present embodiment, for the sake of explanation, the above-mentioned calculation of the variable T ₁ is performed as the information for the signal processing operation (step 112).
In addition, the difference value ΔSLV of the sensor level SLV ₁ (step 116) and the integrated value ΣSLV after the sensor level SLV exceeds a predetermined level LV ₁ (step 11)
8) is calculated, and time Time is read from the clock CL via the interface IF13 (step 120), and the ventilation frequency sensor associated with the _first fire detector DE ₁ is also transmitted via the signal transmitting / receiving unit TRX12 and the interface IF14. The ventilation frequency K is read from SI ₁ (step 122), and the number sensor SI ₂ to 1 is sent via the signal transmitting / receiving unit TRX13 and the interface IF15.
The number P of persons in the room associated with the fire detector DE ₁ is collected (step 124).

Here, the difference value ΔSLV is, for example, the sensor level collected this time and the sensor level collected previously among the sensor levels stored in plural in the sensor level storage area RAM 11. It is calculated by dividing the difference between and by the time difference between the previous time and this time.

Further, the calculation of the integral value ΣSLV is performed every time the sensor level SLV ₁ of the predetermined level LV ₁ or higher is collected from the _first fire detector DE ₁ in question, the integral value for the integral value is calculated up to the previous time. The predetermined value LV ₁ of the sensor level SLV ₁ is added to the integrated value ΣSLV stored in the storage area RAM 12.
It is performed by adding the above values (SLV ₁ −LV ₁ ) and the storage area RAM for the integrated value is obtained by the addition result.
The integrated value up to the previous time stored in 12 is updated.
That is, the contents ΣSLV = (RAM12) of the storage area RAM 12 for the integral value up to the last time is updated with a _{(RAM12) + SLV 1 -LV 1} .

When the above various information has been collected and / or calculated, first, the storage area ROM1 for individual rules
The information regarding the processing rule required for the _first fire detector DE ₁ is read from 2 and stored in the storage area RAM 13 (step 126).

Specifically, in the area for the _first fire detector DE ₁ in the storage area ROM 12, rules a, b, d,
The detailed information on e, the address of the definition function used in each rule in the storage area ROM 13 for the definition function, and the number of rules R = 4 are stored, and these are read out to store the number of rules storage area. It is stored in the RAM 13 (step 126). The storage location of each definition function in the storage area ROM 13 can be known from the address in the storage area ROM 13 stored in the storage area ROM 12.

Next, the four data stored in the storage area RAM 13
The following processing is performed in order for each rule. First r
= 1 rule, that is, the process for the rule a will be described. From the storage area ROM12 to the definition function storage area ROM13 shown in FIG.
The head address AD _{1 of the} area containing the definition function of (a) is read (step 132). Next, the value of the input information used for rule a, that is, the storage area RAM in step 114.
The latest sensor level SLV ₁ stored in 11 is added to the start address AD _1, reads the contents of the AD ₁ + SLV ₁ address region that contains the defined function (a) of FIG. 2, the definition function value The data is stored in the normal area (see FIG. 7) of the storage area RAM 16 (step 134). The content of the address AD ₁ + SLV _{1 in} this area corresponds to the definition function value for the sensor level SLV _1, that is, the fire accuracy F ₁ (SLV ₁ ).

A weighting value determined as will be described later is given to each rule and used according to the degree of contribution to the fire judgment value. The weighted value of such a rule is stored in the storage area RAM17 for each fire detector (see FIG. 8). Therefore, from the area for the _first fire detector DE ₁ in the storage area RAM17, heavy bid omega ₁ of r = 1 rule that rule a is read out is applied to the _{F 1 (SLV 1), ω} 1 × F 1 (SLV 1) is obtained (step 136).

Ω ₁ × F ₁ (SLV ₁ ) thus obtained
Is cumulatively stored in the storage area RAM 14 for the total definition function value. Since the case of r = 1 has been described, the storage area RAM 14 is stored as ω ₁ × F
₁ (SLV ₁ ) only. Further, the weight value ω ₁ is stored as a total Tr of weight values for later use (step 1
38).

Similarly, the processing for the next rule b of r = 2 will be described (step 130).
From the OM 12 in the storage area ROM 13 for the definition function,
The start address AD _{2 of the} area containing the definition function of FIG. 2B corresponding to the rule b is read (step 13
2). Then, by adding the value of the input information used for the rule b, or sensor level SLV ₁ is from exceeding a predetermined level LV ₁ time T ₁ (which is determined in step 112) the leading address AD _2, The contents of the address AD ₂ + T _{1 of the} area containing the definition function of FIG. 2B, that is, the fire probability F ₂ (T ₁ ) is read (step 134), and the fire probability F is read.
₂ (T ₁ ) the weighted value ω read from the storage area RAM 17
₂ is given to obtain ω ₂ × F ₂ (T ₁ ) (step 136), and first, the fire probability ω ₁ × F ₁ (SLV ₁ ) stored in the storage area RAM 14 for the total definition function value. To ω ₁ F
₁ (SLV ₁ ) + ω ₂ F ₂ (T ₁ ) is obtained and the total weighted value T
Obtain r = ω ₁ + ω ₂ (step 138).

Thereafter, the same processing is performed for the rule d of r = 3 and the rule e of r = 4, and step 11
8 and the fire probability F ₄ (ΣSLV) based on the integrated value ΣSLV and the ventilation rate K determined in step 122, respectively.
F ₅ (K) is obtained (step 134), and the fire probabilities are weighted values ω ₃ and ω ₄ read from the storage area RAM 17.
The corresponding weighting is performed (step 136),
Then, the total definition function value is added to the storage area RAM 14 (step 138).

When the processing for all the usage rules a, b, d, and e is completed in this way (step 140),
Fire accuracy added value stored in storage area RAM14

[0042]

[Formula 2] ω ₁ F ₁ (SLV ₁ ) ＋ ω ₂ F ₂ (T ₁ ) ＋ ω ₃ F ₄ (ΣSLV) ＋ ω ₄ F ₅ (K)

Is read (step 142), the added value is divided by the total weight Tr of the weighted values to obtain the fire probability Total (step 144), and the divided fire probability value is displayed on the display DP. (Step 146) Further, the value is compared with an appropriate reference value F (step 148), and if it is equal to or greater than the reference value F, an appropriate fire action such as displaying a fire is taken (step 150).

When it is determined that the fire accuracy Total is equal to or greater than the reference value F, it is further determined whether or not the determination is the first one (step 152). (Y in step 152), the fire accuracy Total is transferred to the storage area RAM 18 (see FIG. 9) in order to store it until a data display request or a learning request (weighting value conversion request) is issued later. Step 154), the function value MF of the fire accuracy stored in the normal location of the storage area RAM 16 obtained by the rules a, b, d, and e of each processing is used as the fire location function value of the storage area RAM 16 (see FIG. 7)) (steps 156-1)
64).

When calculating the goodness of fit of the function values obtained by the rules of each processing, it is determined whether or not the fire was displayed for the first time. Therefore, the function value obtained based on the data collected from the fire detector after that is different from the function value that was initially determined to be a fire and is a value that can be reliably determined to be a fire in progress. Even if the suitability of each processing rule for a function value different from the first fire occurrence time is calculated, the judgment of the fire occurrence time, that is, the suitability of each processing rule for early fire detection This is not the case.

With this, the signal processing operation for the _first fire detector DE ₁ is completed, and the same processing operation is performed for the fire detectors subsequent to the _second fire detector DE ₂ ( N in step 166 and step 104). In this way, the processing of the fire detector is continued in order, and when the processing of the last Nth fire detector DE _N is completed (Y in step 166), the storage area RAM 16 and RA are stored.
The display of the data stored in M18 and the rewriting operation of the weight value stored in the storage area RAM 17 are performed.

First of all, in the above processing operation, none of the fire detectors No. 1 to N are in the fire operation. Therefore, when no fire display is made, the first one is not changed. The same process is continued by returning to the fire detector DE ₁ (N in step 168, N in step 170, step 176, N in step 178 and step 102).

If any one of the fire detectors fires and displays a fire, and the fire display is not yet cleared (Y in step 168), this means that the fire is still calming. This is because the recovery operation has not been performed because the fire situation is continuing and the steps for displaying the data and rewriting the weight value are not performed. Fire detector D
Returning to E ₁ , the signal processing operation for fire monitoring is performed.

If any one of the fire detectors fires and the fire is displayed, and the fire subsides,
A fire operation return switch (not shown) is operated to perform the return operation of the fire alarm device. When the return operation is performed in this way, the fire display on the display DP is also cleared (N in step 168), so the data display switch SW ₁ (Y in step 170) and the learning switch (weighting conversion switch) SW ₂ If is operated (Y in step 178), it is possible to display data and rewrite the weighting value. The data display switch SW ₁ and the learning switch (weighting value conversion switch) SW ₂ may not be separately provided, and the data display command and the learning command (weighting value rewriting command) may be input from the operation unit OP. ..

When the data display switch SW ₁ is operated (Y in step 170), the function values MF of the four rules a, b, d and e for the _first fire detector DE ₁ are stored. It is read from the area of the area RAM 16 at the time of the fire (step 172), and those function values and storage area R
The TOTAL value for the No. 1 fire detector DE ₁ stored in the AM 18 is displayed on the cleared display DP (step 174). In this embodiment, only the function values for the four rules in the _first fire detector DE ₁ are stored in the location of the fire in the storage area RAM 16, and only the total value of the fire detector DE _{1 in} the storage area RAM 18 is stored.
However, if there is a fire operation in another fire detector as well, the function value of each rule and the Total value in the fire detector that has operated are also stored in the storage area RAM16. Fire location and RAM1
Of course, since it is stored in No. 8, it is displayed together as the function value and the Total value for the other fire detectors.

If the learning switch, that is, the weight conversion switch SW ₂ is not operated after the display operation of the total value and data is performed (N in step 178), the process returns to the _first fire detector DE _1. When the signal processing operation for fire monitoring is performed, but the learning switch SW ₂ is operated (Y in step 178), the storage area R
An operation for rewriting the weight value stored in the AM 17 is performed.

First, while observing the function value and the total value (that is, fire accuracy) displayed on the display DP,
The operator inputs the desired Total value from the operation unit OP for each fire detector for which it is desired to change the weight value.
When it is desired to change the weighted value, there are cases where the displayed Total value does not match the actual fire situation. In that case, enter the desired Total value that matches the actual fire situation. It will be.

The desired total value input in this manner is stored in the storage area RAM 19 together with the number M of fire detectors whose weighting value is desired to be changed (step 180).
Then, for each of the M fire detectors, the weighted value of the R processing rules is obtained. That is, in the case of the present embodiment, since only the _first fire detector DE ₁ is supposed to have fired, M = 1 and R = 4. And below, the desired To for the _first fire detector DE ₁
An operation is performed from the tal value to obtain respective weighted values for the four processing rules a, b, d, and e.

Storage area RAM 16 for each processing rule
The function value MF is read from the location at the time of the fire (step 19
0), desired Total for the function value MF of the rule of each processing
The value ratio A = (desired Total value) / (function value) is taken (step 192), and the weight value ω of each rule is read from the storage area RAM 17 (step 194).

In this case, in this embodiment, when the ratio A is 1.2 or more (Y in step 196), the weight value B is set to 1.2 × ω (step 198), and the ratio A is 1. If it is less than 2 and is 0.8 or more (Y of step 200), the weighting value B is set to A × ω (step 202), and if the ratio A is less than 0.8, the weighting value B is set. It is assumed to be 0.8 × ω. The storage area RAM 17 shown in FIG.
The weighting value of the fire detector area in the table is updated with the weighting value B that has just been set (step 208). When all R weight values of the fire detector are rewritten (Y in step 210), the setting value of the weight value for the next fire sensor stored in the storage area RAM 19 is changed. Then, when the weight values for all M fire detectors stored in the storage area RAM 19 are rewritten (Y in step 212), the normal fire monitoring operation from the first fire detector is returned to.

In the next processing operation in steps 136 and 138, the weighting value thus changed in steps 178 to 212 is used.

In the above embodiment, the rules a to f are shown as the processing rules, and the definition functions shown in FIGS. 2 (a) to 2 (f) are shown. Therefore, it will be easily understood that the rules of these processes and the contents of the defining functions can be appropriately changed according to the environment in which they are used.

[0058]

As described above, according to the present invention, in order to obtain highly reliable fire information in consideration of a wide range of acquired information, fire information such as fire accuracy is obtained for each information obtained by the information acquisition means. A function is defined, a rule of processing suitable for an environmental condition to be performed using the function is appropriately selected and defined in advance, and the acquired information is processed for each defined rule of processing. In a fire alarm device that obtains a function value and comprehensively judges the obtained function value, how effectively the rules of each process used for inference work on the fire information as the inference result. According to the rule of each process, the function value obtained at the time of fire or the like is stored in the function value storage means, and when necessary, the total value is calculated based on the stored function value. Judgment value By inputting a value, the weight value attached to each processing rule can be changed, so the fire information as the inference result obtained from the currently used processing rule is actually If it does not correspond to the situation of 1), it can be corrected, so that the reliability of inference using the rule of each processing can be greatly improved.

[Brief description of drawings]

FIG. 1 is a block circuit diagram showing a fire alarm device to which an embodiment of the present invention is applied.

FIG. 2 is a diagram showing an example of a definition function that can be used in an embodiment of the present invention.

3 is a flow chart for explaining the operation of the fire alarm device of FIG. 1 on the fire receiver side.

4 is a flow chart for explaining the operation of the fire alarm device of FIG. 1 on the fire receiver side.

5 is a flow chart for explaining the operation of the fire alarm device of FIG. 1 on the fire receiver side.

6 is a flowchart for explaining the operation of the fire alarm device of FIG. 1 on the fire receiver side.

FIG. 7 is a diagram showing in detail a storage area RAM 16 for rule function values.

FIG. 8 is a diagram showing in detail a storage area RAM 17 for rule weighting values.

FIG. 9 is a diagram showing in detail a storage area RAM 18 for a Total value during a fire.

[Explanation of symbols]

RE Fire receiver MPU1 Microprocessor ROM11 Program storage area ROM12 Individual rule storage area ROM13 Definition function storage area ROM14 Initial rule weight value storage area RAM11 Sensor level storage area RAM12 Integral value Storage area for RAM RAM13 Storage area for the number of rules used RAM14 Storage area for total definition function values RAM15 Working area RAM16 Storage area for rule function values RAM17 Storage area for rule weight values RAM18 Total in case of fire Storage area for value RAM19 Storage area for desired Total value SW ₁ Data display switch SW ₂ Learning switch SI ₁ Ventilation rate sensor SI ₂ People sensor CL Clock DP Display OP Operation part DE _{1 to} DE _N Fire detector FS Fire phenomenon Detection means

Claims (3)

[Claims] 1. In order to obtain fire information based on various information relating to a fire phenomenon, various collected information relating to the fire phenomenon, and information acquisition means for obtaining processing information from the collected information, and the information. Defining a function for the fire information for each piece of information obtained by the obtaining means, and defining at least one processing rule to be performed using the function, and defining means for defining the function. The weighting value storage means for storing a weighting value for each rule of each processing, and the information acquisition means based on the rule of each processing and each corresponding function used in the rule of each processing. By processing the obtained information, the function value for each rule of each processing is obtained, and the obtained function value is weighted with the corresponding weight value stored in the weight value storage means. A weighting function value, processing means for comprehensively processing the weighting function value to obtain the fire information, function value storing means for storing a function value for each rule of the processing, and the function value Determining means for determining a desired value of the fire information to be obtained from the function value stored in the storage means, the function value stored in the function value storage means, and the determination value determined by the determining means Based on the desired value
A fire alarm device, comprising: update means for updating the weight value stored in the weight value storage means for each rule of each process. 2. The determination means, when requested, displays the function value stored in the function value storage means, and the fire value based on the function value displayed on the display means. The fire alarm device according to claim 1, further comprising: an input unit capable of inputting the desired value of the information. 3. The updating means is configured to perform the processing for each rule of the processing based on a ratio between the function value stored in the function value storage means and the desired value determined by the determining means. The fire alarm device according to claim 1 or 2, wherein the weight value stored in the weight value storage means is updated.