CN111784994B - Fire detection method and device - Google Patents
Fire detection method and device Download PDFInfo
- Publication number
- CN111784994B CN111784994B CN202010673849.0A CN202010673849A CN111784994B CN 111784994 B CN111784994 B CN 111784994B CN 202010673849 A CN202010673849 A CN 202010673849A CN 111784994 B CN111784994 B CN 111784994B
- Authority
- CN
- China
- Prior art keywords
- fire
- node
- influence
- probability
- determining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B29/00—Checking or monitoring of signalling or alarm systems; Prevention or correction of operating errors, e.g. preventing unauthorised operation
- G08B29/18—Prevention or correction of operating errors
- G08B29/185—Signal analysis techniques for reducing or preventing false alarms or for enhancing the reliability of the system
- G08B29/188—Data fusion; cooperative systems, e.g. voting among different detectors
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B17/00—Fire alarms; Alarms responsive to explosion
Abstract
The invention discloses a fire detection method and a fire detection device. Wherein, the method comprises the following steps: determining a plurality of fire influence factors of each node in an engine room; determining a fire influence weight value of each fire influence factor in a plurality of fire influence factors of each node; determining an evaluation value of the fire occurrence probability of each node in the cabin according to the fire influence weighted value of each fire influence factor; determining the actual probability of fire in the cabin according to the evaluation value of each node on the probability of fire in the cabin; and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin. By the technical scheme, the fire disaster can be accurately detected, and whether the fire disaster alarm is sent or not can be accurately judged, so that the timeliness and the accuracy of the fire disaster alarm are ensured.
Description
Technical Field
The invention relates to the technical field of fire detection, in particular to a fire detection method and a fire detection device.
Background
At present, when a fire disaster is detected, measurement data acquired by a smoke detection module, a CO gas detection module and a temperature detection module are fused by adopting a gray fuzzy neural network information fusion algorithm or measurement data acquired by different sensors or modules are fused by adopting a fuzzy logic control system with feedback (namely a fuzzy logic fusion algorithm) and then whether the fire disaster occurs is judged, but the methods have low detection accuracy, are easy to misjudge the fire disaster, have poor detection timeliness and consume too long detection time.
Disclosure of Invention
The invention provides a fire detection method and a fire detection device. The technical scheme is as follows:
according to a first aspect of embodiments of the present invention, there is provided a fire detection method including:
determining a plurality of fire influence factors of each node in an engine room;
determining a fire influence weight value of each fire influence factor in the plurality of fire influence factors of each node;
determining an evaluation value of each node on the probability of the fire in the cabin according to the fire influence weight value of each fire influence factor;
determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin;
and judging whether to send out a fire alarm or not according to the actual probability of the fire in the cabin.
In one embodiment, the determining a fire impact weight value for each of the plurality of fire impact factors for each node includes:
determining a judgment matrix of each fire influence factor;
carrying out consistency check on the judgment matrix of each fire influence factor;
and calculating the fire influence weighted value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test.
In one embodiment, when there are a plurality of judgment matrices for the fire influencing factors, the calculating the fire influencing weight value for each fire influencing factor according to the judgment matrices for the fire influencing factors passing the consistency check includes:
calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrices of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
In one embodiment, determining the evaluation value of the node for the probability of the fire occurring in the cabin according to the fire influence weight value of each fire influence factor comprises:
determining a fire influence weight region of each fire influence factor according to the fire influence weight value of each fire influence factor;
determining a target influence function of each fire influence factor and a constraint condition of the target influence function according to the fire influence weight interval of each fire influence factor;
calculating an optimal fire influence weight value of each fire influence factor of each node according to the target influence function of each fire influence factor and the constraint condition of the target influence function;
and determining the evaluation value of each node on the probability of the fire in the cabin according to the optimal fire influence weight value of each fire influence factor of each node and the actual factor value of each fire influence factor of each node.
In one embodiment, the determining the actual fire occurrence probability in the cabin according to the evaluation value of each node on the fire occurrence probability in the cabin comprises:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
and determining the actual probability of fire in the cabin according to the augmented support matrix.
In one embodiment, the determining the probability of the actual fire in the cabin according to the augmented support matrix comprises:
determining an evaluation credibility coefficient of an evaluation value of the probability of fire occurrence in the cabin by each node according to the augmentation support matrix;
and determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin and the evaluation confidence coefficient of each node on the evaluation value of the probability of the fire in the cabin.
In one embodiment, the determining whether to alarm the fire based on the actual probability of the fire in the cabin includes:
calculating a probability threshold value of the fire in the cabin according to the fire occurrence critical factor value of each fire influence factor of each node;
and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin and the probability threshold value of fire in the cabin.
According to a second aspect of embodiments of the present invention, there is provided a fire detection apparatus including:
the first determining module is used for determining a plurality of fire influence factors of each node in the cabin;
the second determining module is used for determining a fire influence weight value of each fire influence factor in the plurality of fire influence factors of each node;
the third determining module is used for determining the evaluation value of the fire occurrence probability of each node in the cabin according to the fire influence weight value of each fire influence factor;
the fourth determining module is used for determining the actual probability of fire in the cabin according to the evaluation value of each node on the probability of fire in the cabin;
and the judging module is used for judging whether to send out fire alarm or not according to the actual probability of fire in the cabin.
In one embodiment, the second determining module comprises:
the determining submodule is used for determining a judgment matrix of each fire influence factor;
the detection submodule is used for carrying out consistency detection on the judgment matrix of each fire influence factor;
and the calculating submodule is used for calculating the fire influence weight value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test.
In one embodiment, the calculation submodule is specifically configured to:
when a plurality of judgment matrixes of the fire influence factors exist, calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrixes of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
In one embodiment, the third determining module comprises:
the first determining submodule is used for determining a fire influence weight region of each fire influence factor according to the fire influence weight value of each fire influence factor;
the second determining submodule is used for determining a target influence function of each fire influence factor and a constraint condition of the target influence function according to the fire influence weight interval of each fire influence factor;
the calculating submodule is used for calculating the optimal fire influence weight value of each fire influence factor of each node according to the target influence function of each fire influence factor and the constraint condition of the target influence function;
and the third determining submodule is used for determining the evaluation value of the fire occurrence probability of each node to the cabin according to the optimal fire influence weight value of each fire influence factor of each node and the actual factor value of each fire influence factor of each node.
In one embodiment, the fourth determining module is specifically configured to:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
and determining the actual probability of fire in the cabin according to the augmented support matrix.
In an embodiment, the fourth determining module is further specifically configured to:
determining an evaluation credibility coefficient of an evaluation value of the probability of fire occurrence in the cabin by each node according to the augmentation support matrix;
and determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin and the evaluation confidence coefficient of each node on the evaluation value of the probability of the fire in the cabin.
In one embodiment, the determining module is specifically configured to:
calculating a probability threshold value of the fire in the cabin according to the fire occurrence critical factor value of each fire influence factor of each node;
and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin and the probability threshold value of fire in the cabin.
The technical scheme of the invention can realize the following technical effects:
the evaluation value of the fire occurrence probability of each node in the cabin can be preliminarily calculated by calculating the fire influence weighted value of each fire influence factor of each node, and the evaluation value is possibly inaccurate and needs to be adjusted, so that the probability of actual fire occurrence in the cabin can be calculated by using the evaluation value, the fire is accurately detected, whether fire alarm is given or not is accurately judged, and timeliness and accuracy of the fire alarm are ensured.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
Fig. 1 is a schematic deployment diagram of a fire detection sensor node according to the present invention.
Fig. 2 is a fire probability error curve according to the present invention.
Fig. 3 is a comparison graph of detection time of a fire detection method according to the present invention compared to two fire detection algorithms in the related art.
FIG. 4 is a graph showing the relationship between the number k of decision matrices and the false alarm rate of fire detection according to the present invention.
Fig. 5 is a comparison graph of detection accuracy of a fire detection method provided by the present invention compared with two fire detection algorithms in the related art.
Fig. 6 is a flowchart of a fire detection method according to the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.
In order to solve the above technical problem, an embodiment of the present invention provides a fire detection method, which may be used in a fire detection program, system or device, as shown in fig. 6, and includes steps S101 to S105:
in step S101, a plurality of fire influencing factors of each node in the cabin are determined; fire influencing factors include, but are not limited to: temperature, smoke concentration, CO concentration, and infrared light intensity. The nacelle may be an aircraft cabin, an aviation cabin, etc., and each node, i.e., each sensor node, may, of course, have one or more sensors of different types for detecting actual factor values of the factors affecting the fire.
In step S102, determining a fire influence weight value of each of the plurality of fire influencing factors of each node; and the fire influence weight value is used for representing the weight of each fire influence factor in the fire. In addition, in order to ensure the accuracy of fire detection, the same fire influencing factors are adopted by all nodes.
In step S103, determining an evaluation value of the node for the probability of occurrence of the fire in the cabin according to the fire influence weight value of each fire influence factor; the evaluation value is used for representing the size of the probability of fire of each node, the larger the evaluation value of a certain node is, the larger the probability of fire of the node is, and the smaller the probability of fire of the node is.
In step S104, determining the actual fire probability in the cabin according to the evaluation value of each node on the fire probability in the cabin;
in order to ensure the accuracy of fire alarm, the evaluation values of the fire occurrence probability of each node need to be fused, and meanwhile, in order to avoid the influence on the accuracy of the fusion result when the sensor nodes deployed in the cabin break down, the invention can fuse the data of the multiple sensor nodes based on the support matrix (namely, the evaluation values of the fire occurrence probability of each node are fused).
In step S105, whether a fire alarm is issued is determined based on the probability of an actual fire in the cabin.
The evaluation value of the fire occurrence probability of each node in the cabin can be preliminarily calculated by calculating the fire influence weighted value of each fire influence factor of each node, and the evaluation value is possibly inaccurate and needs to be adjusted, so that the probability of actual fire occurrence in the cabin can be calculated by using the evaluation value, the fire is accurately detected, whether fire alarm is given or not is accurately judged, and timeliness and accuracy of the fire alarm are ensured.
In one embodiment, the determining a fire impact weight value for each of the plurality of fire impact factors for each node includes:
determining a judgment matrix of each fire influence factor;
carrying out consistency check on the judgment matrix of each fire influence factor;
and calculating the fire influence weighted value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test.
The judgment matrix of each fire influence factor is determined, consistency check can be carried out on the judgment matrix, if the consistency check is passed, the judgment matrix is determined to be a reasonable matrix, otherwise, the judgment matrix is discarded, and after the judgment matrix is determined to be the reasonable matrix, the fire influence weight value of each fire influence factor can be calculated according to the judgment matrix of each fire influence factor passing the consistency check, so that the weight occupied by each fire influence factor in fire occurrence can be determined.
For example: the influence degree relation among the variables is quantitatively expressed by a judgment matrix, and the judgment matrix is set to be C e to Rn×nWherein n represents the number of variables.
The judgment matrix C is of the form:
matrix elements
Indicating fire condition by ith and jth variablesInfluence degree importance ratio.
Element cijThe filling rules are as follows:
if xiAnd xjOf equal importance, then take cij＝1,cji＝1；
If xiRatio xjOf slight importance, take cij＝3,cji＝1/3；
If xiRatio xjObviously important, take cij＝5,cji＝1/5；
If xiRatio xjMore important, then take cij＝7,cji＝1/7；
If xiRatio xjOf absolute importance, then take cij＝9,cji＝1/9；
If xiAnd xjIs between the above relationships, c ij2,4,6 and 8 can be taken; c. CjiValues of 1/2,1/4,1/6 and 1/8 can be obtained. Thus, element cijThe above-mentioned rule can be any integer from 1 to 9.
In one embodiment, when there are a plurality of judgment matrices for the fire influencing factors, the calculating the fire influencing weight value for each fire influencing factor according to the judgment matrices for the fire influencing factors passing the consistency check includes:
calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrices of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
The judgment matrix may be a plurality of judgment matrices, and when there are a plurality of judgment matrices, a fire influence weight value may be calculated based on each judgment matrix that passes the consistency check, and thus, when there are a plurality of judgment matrices that pass the consistency check, the calculated fire influence weight values of the fire influence factors may be a plurality, and thus, a fire influence weight interval of the fire influence factors may be formed.
In one embodiment, determining the evaluation value of the node for the probability of the fire occurring in the cabin according to the fire influence weight value of each fire influence factor comprises:
determining a fire influence weight region of each fire influence factor according to the fire influence weight value of each fire influence factor;
determining a target influence function (for solving the smallest weight deviation in the interval) of each fire influence factor and a constraint condition of the target influence function according to the fire influence weight interval of each fire influence factor (the constraint condition can be that the sum of the optimal fire influence weight values of each fire influence factor of each node is 1);
calculating an optimal fire influence weight value of each fire influence factor of each node according to the target influence function of each fire influence factor and the constraint condition of the target influence function;
and determining the evaluation value of each node on the probability of the fire in the cabin according to the optimal fire influence weight value of each fire influence factor of each node and the actual factor value of each fire influence factor of each node. The actual factor value of each fire influence factor is the actual value of each fire influence factor, for example: the actual value range of the fire influence factors, namely the temperature (unit ℃) is [0,100], the actual value range of the smoke concentration (unit ppm) is [100,1000], the actual value range of the CO concentration (unit ppm) is [10,100], the actual value range of the infrared ray intensity (unit Lux) is [100,1000], and before the evaluation value is calculated, the actual factor values of the fire influence factors can be normalized.
After determining the fire influence weight regions of the fire influence factors, determining a target influence function of each fire influence factor and a constraint condition of the target influence function so as to automatically screen out an optimal fire influence weight value from the fire influence weight regions of each fire influence factor according to the target influence function and the constraint condition, wherein the optimal fire influence weight value can give consideration to all weight values in the weight regions to the maximum extent, so that the determined evaluation value of each node on the occurrence probability of the fire in the cabin can ensure that the final evaluation value can better reflect the actual situation by utilizing the optimal fire influence weight value of each fire influence factor and the actual factor value of each fire influence factor of each node; meanwhile, the optimal weight is used for weighting fusion, so that a complex calculation process during fusion can be effectively avoided, and the reaction speed to fire is improved.
In one embodiment, the determining the actual fire occurrence probability in the cabin according to the evaluation value of each node on the fire occurrence probability in the cabin comprises:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
order to
And
respectively representing sensor nodes i*And j*Estimate of probability of fire occurrence in the cabin at time k (i)*,j*E 1,2, …, N). If it is not
And
if the difference is large, it indicates that the two sensor nodes do not support each other at the time k, and if the difference is small, it indicates that the two nodes support each other. If a node is supported by a large number of nodes simultaneously, it is considered to be a valid fire probability assessment. Otherwise, the probability value evaluated by the node is given lower weight in the fusion processWeighing;
to show
And
the mutual support degree between the fuzzy sets is defined as a decay exponential function in the fuzzy set:
the parameter α is adjustable to adjust the fusion accuracy, typically set to 0.8[16 ].
And a support function
Represents a node i*And node j*Mutual support at the k-th instant.
Establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
and determining the actual probability of fire in the cabin according to the augmented support matrix.
After establishing a support function between any two nodes of the nodes according to the evaluation value of the node on the probability of the fire in the cabin, because one support function exists between every two nodes, namely a plurality of support functions exist between all the nodes, for convenience of processing, the support function may be converted into an initial support matrix with respect to the evaluation value, in order to facilitate the self-adaptive distribution of weight coefficients (namely, the evaluation credibility coefficients of the evaluation values of the nodes) for each node, the evaluation values of the probability of fire occurrence in the cabin are fused and adjusted, so as to achieve the purpose of improving the accuracy of the probability of actual fire occurrence in the cabin, an augmented support matrix can be constructed based on the initial support matrix to improve the accuracy of the calculated probability of actual fire in the cabin, and further the fire can be accurately predicted.
In one embodiment, the determining the probability of the actual fire in the cabin according to the augmented support matrix comprises:
determining an evaluation credibility coefficient of an evaluation value of the probability of fire occurrence in the cabin by each node according to the augmentation support matrix;
and determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin and the evaluation confidence coefficient of each node on the evaluation value of the probability of the fire in the cabin.
Based on the augmented support matrix, a higher evaluation credibility coefficient can be distributed to the evaluation value of each node with high credibility, and a lower evaluation credibility coefficient can be distributed to the evaluation value of each node with low credibility, so as to avoid that the fusion result accuracy of the evaluation values of each node is adversely affected due to the fault or inaccurate detection of the sensor node deployed in the cabin, so that the fusion of the evaluation values has fault tolerance, thereby greatly improving the prediction accuracy of the actual fire probability in the cabin, specifically, after the technical scheme of the invention is experimentally compared with a grey fuzzy neural network fusion algorithm and a fuzzy logic fusion algorithm in the related technology respectively from the aspects of time required for detecting the fire and sending out an alarm signal and the false alarm rate of the system, the fire detection scheme of the invention can detect the fire within 10s, and simultaneously reduces the false alarm rate to be below 0.5 percent, greatly reduces the false alarm rate of the fire disaster and improves the accuracy and timeliness of the fire disaster detection.
In one embodiment, the determining whether to alarm the fire based on the actual probability of the fire in the cabin includes:
calculating a probability threshold value of the fire in the cabin according to the fire critical factor value of each fire influencing factor of each node (namely, the value of each fire influencing factor when the critical state of the fire is reached);
and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin and the probability threshold value of fire in the cabin.
According to the fire occurrence critical factor values of the fire influence factors of the nodes, the probability threshold value of the fire occurrence in the cabin can be accurately calculated, then the currently calculated probability of the actual fire occurrence in the cabin is compared with the probability threshold value of the fire occurrence in the cabin, and whether fire alarm is given out or not can be accurately judged.
Finally, it is clear that: the above embodiments can be freely combined by those skilled in the art according to actual needs.
The technical solution of the present invention will be further explained in detail below:
aiming at the condition that a single sensor is used in an aircraft cabin to detect a fire disaster, a multi-sensor data fusion method is provided for detecting the fire disaster. Firstly, calculating the weight of temperature, smoke concentration, CO concentration and infrared ray intensity in the fire by using an improved Analytic Hierarchy Process (AHP) on each sensor node of a Wireless Sensor Network (WSN), and evaluating the probability of the fire in a cabin by using a variable weighting fusion method; then, based on the mutual support degree of each node to fire occurrence probability evaluation data, a weight coefficient is adaptively distributed to each evaluation value, and all node evaluation values are weighted and fused to finally obtain the fire occurrence probability; and finally, comparing the probability with a threshold value to judge whether a fire disaster occurs. The fire detection algorithm is compared with a grey fuzzy neural network fusion algorithm and a fuzzy logic fusion algorithm from the aspects of time required by the system to detect a fire and send an alarm signal and the false alarm rate of the system respectively, experiments show that the fire detection algorithm can detect the fire within 10s, the false alarm rate is reduced to be below 0.5%, and the superiority of the algorithm in the aspects of timeliness and accuracy is verified. And the algorithm is proved to have certain fault-tolerant capability by setting fault sensor nodes.
Based on the research on the method, aiming at the characteristics of narrow and small airplane environment, closed airplane environment and the like, a data fusion method under the action of various factors is considered in the research of fire detection alarm problems based on WSN. In order to detect the fire accurately and timely by a detection algorithm under the condition of more fusion variables, the invention calculates the weight of each variable influencing the fire occurrence probability by using an improved AHP at each sensor node in the WSN, and evaluates the fire occurrence probability in a cabin according to multivariate weighted fusion; meanwhile, in order to avoid the influence of a fault sensor on the detection precision, the invention fuses the evaluation data of each node by using a self-adaptive weighting fusion method. Under the condition that the detection capability of each node is unknown, the reliability of each node on the fire evaluation result is calculated by constructing a support matrix, a node with high reliability is assigned with higher weight, otherwise, a lower weight is assigned, and therefore the influence of a fault sensor on the fusion result is reduced to the minimum.
The main contributions of the invention are: the method aims to solve the problems that timeliness is reduced when a detection system fuses multivariate data and a fusion result has large deviation under the condition that partial sensors have faults. Firstly, the invention considers the weight of each variable influencing the fire occurrence, designs a new multivariate weighted fusion fire evaluation algorithm based on an improved AHP variable weight calculation method, can still quickly detect the fire situation when the fusion variables are more, and simultaneously fuses more related variables, thereby greatly reducing the false alarm rate of the system; secondly, the method fuses the evaluation values of the fire occurrence probability of each sensor node by using a self-adaptive weighting fusion mode, and adaptively distributes weight coefficients to each node by constructing a support matrix, thereby avoiding the problem of fusion precision reduction caused by the measurement deviation of partial fault sensors to a certain extent.
The structure of the invention is as follows: the first section introduces a method of estimating the probability of fire occurrence at each sensor node using an improved AHP and multivariate weighted fusion algorithm; the second section introduces a method for adaptively weighting and fusing the evaluation values of the fire occurrence probability of each node by constructing an augmented support matrix. The third section is an experimental part, and verifies that the fire detection algorithm provided by the invention has timeliness, accuracy and fault tolerance. The fourth section concludes the present invention.
1. Fire evaluation algorithm based on multivariate weighting fusion
In this section, an improved AHP method is studied, and a multivariate weight calculation method is proposed to determine the weight of temperature, smoke concentration, CO concentration, and infrared intensity in the occurrence of a fire. A new multivariate weighting fusion algorithm is provided at each sensor node of the WSN, and the evaluation value of the fire occurrence probability in the cabin by each node is obtained.
1.1 improved AHP multivariate weighting Algorithm
To evaluate and judge the occurrence of fire in the cabin, relevant influencing factor variables are determined, and the weight of the variables indicates the status of the variables in the evaluation process. The magnitude of the weight of one variable will have a significant impact on the fire assessment results. Therefore, it is a notable problem to determine the variable weights scientifically and reasonably, and the present invention mainly adopts the improved AHP method to determine the variable weights [11 ]. The conventional AHP calculates the relative weight of the lowest layer to the highest layer by selecting a judgment matrix, and sorts various schemes and measures in the lowest layer according to the weight. However, the number of the selected judgment matrixes is single, so that the accuracy of the weight is not high. In order to ensure the accuracy of the weight, the invention utilizes the improved AHP to calculate the weight of each influencing factor variable in the fire occurrence process, and the optimal weight of each variable is obtained by selecting a plurality of judgment matrixes and constructing an objective function in a weight interval.
The basic process of evaluating the fire problem by the improved AHP method is as follows: 1) determining various factor variables influencing the occurrence of the fire; 2) establishing a judgment matrix of each variable; 3) carrying out consistency check on the judgment matrix; 4) solving a variable weight interval by a feature vector method; 5) and setting an objective function to obtain the optimal weight. The weight calculation steps are as follows:
firstly, setting the variable of factors influencing the fire occurrence as temperature, smoke concentration, CO concentration and infrared ray intensity, wherein the variable is xi(i ═ 1,2,3, …, n) indicates that i is a variable index. Let x1Representative of temperature, x2Representing the smoke concentration, x3Represents the CO concentration, x4Representing the intensity of infrared light;
step two, establishing a judgment matrix of each variable:
the influence degree relation among the variables is quantitatively expressed by a judgment matrix, and the judgment matrix is set to be C e to Rn×nWherein n represents the number of variables.
The judgment matrix C is of the form:
matrix elements
The importance ratio of the impact degree of the ith and jth variables on the fire condition is shown.
Element cijThe filling rules are as follows:
if xiAnd xjOf equal importance, then take cij＝1,cji＝1；
If xiRatio xjOf slight importance, take cij＝3,cji＝1/3；
If xiRatio xjObviously important, take cij＝5,cji＝1/5；
If xiRatio xjMore important, then take cij＝7,cji＝1/7；
If xiRatio xjOf absolute importance, then take cij＝9,cji＝1/9；
If xiAnd xjIs in the above-mentioned relationshipBetween systems, then c ij2,4,6 and 8 can be taken; c. CjiValues of 1/2,1/4,1/6 and 1/8 can be obtained. Thus, element cijThe above-mentioned rule can be any integer from 1 to 9.
Step three, carrying out consistency check on the judgment matrix:
setting the consistency ratio
Wherein the content of the first and second substances,
(λ is the maximum eigenvalue of C, and n is the order of the decision matrix). RI is an average random consistency index, and its value is determined by different orders of the decision matrix (as shown in table 1). When CR is less than 0.1, the corresponding judgment matrix C accords with the consistency test and is a reasonable matrix, otherwise, the judgment matrix C is discarded.
TABLE 1 RI values at different orders
| Order of |
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| |
0 | 0 | 0.52 | 0.89 | 1.12 | 1.26 | 1.36 | 1.41 |
Solving a variable weight interval by a feature vector method:
based on the judgment matrix constructed as described above, the weight of each variable is calculated by the eigenvector method, and the calculated weight is written in the form of a section, which is taken as a weight section of the variable (the section is composed of independent weight points).
And corresponding to a randomly selected reasonable judgment matrix, calculating a weight of the variable according to a characteristic vector method. The weight calculation formula is as follows:
wherein i is a variable subscript (i is 1,2,3, …, n), and k represents the number of the selected reasonable judgment matrixes (k is 1,2,3, …, m; is a positive integer);
is a weight of the i-th variable found under the k-th decision matrixThis gives k weights for each variable i.
And writing the k weights of the variables into a section, wherein the section is the weight section of the variables. For any variable i, a weight interval of the variable i can be obtained based on the formulas (2) to (4), and the interval contains k weights.
Step five, setting an objective function to obtain the optimal weight:
the optimal weight can give consideration to all weight information in the weight interval to the maximum extent, so that the final evaluation result can better reflect the actual condition; meanwhile, the optimal weight is used for weighting fusion, so that a complex calculation process during fusion can be effectively avoided, and the reaction speed to fire is improved. Let wiA weight representing the ith variable is constructed, and an objective function (namely the above objective influence function) for solving the minimum weight deviation in the interval is as follows:
the constraints of the objective function are:
the optimal weights for the ith variable based on equations (5) - (6) are recorded
1.2 variable weighting fused fire occurrence probability evaluation algorithm
Aiming at the actual value taking conditions of each variable in different environments in the cabin, the temperature (unit ℃) is set to be [0,100], the smoke concentration (unit ppm) interval is [100,1000], the CO concentration (unit ppm) interval is [10,100], and the infrared ray intensity (unit Lux) interval is [100,1000 ]. Constructing a conversion function to map the real value of the variable to [0-1] so as to avoid the influence on the calculation result due to different dimensions, wherein the conversion function is as follows:
wherein, max (x)i) Is the maximum value in the interval of variable i, min (x)i) Is the minimum value of the interval where the variable i is located.
N sensor nodes are arranged in the cabin, the evaluation value of each node on the fire occurrence probability in the cabin is p, and the normalized variable value is recorded as
The variable optimum weight is
The variable weighting result is the fire occurrence probability, and the calculation formula is as follows:
wherein
Is node i*And (5) evaluating the fire occurrence probability.
2. Fusion algorithm based on node data support degree
During fire detection, if the probability of fire occurrence in the cabin is greater than the threshold probability, a fire alarm signal needs to be sent out. In order to ensure the accuracy of fire alarm, the fire occurrence probability values evaluated by all nodes need to be fused. Meanwhile, in order to avoid the influence on the accuracy of a fusion result when a sensor node deployed in a cabin breaks down, the section provides a multi-sensor node data fusion method based on a support matrix. The method objectively reflects the support degree between node evaluation data without knowing the evaluation capability of each node on the fire occurrence probability. By constructing an amplification matrix, the weight coefficient of the fire probability evaluation value of each node in the fusion process is adaptively adjusted, so that the fusion effect is optimal. The method is characterized in that a large amount of data can be fused on line, and meanwhile, in the adaptive weight coefficient distribution process, the fire disaster assessment value with high reliability is distributed with higher weight, otherwise, the fire disaster assessment value is distributed with lower weight, so that the algorithm has certain fault-tolerant capability.
2.1 construct support matrix
N sensor nodes (each node consists of temperature, smoke concentration, CO concentration and infrared ray intensity sensors) are arranged to measure environment variables, and the probability of fire occurrence in the cabin at the moment can be obtained at each node through variable weighting fusion calculation, so that
And
respectively representing sensor nodes i*And j*Estimate of probability of fire occurrence in the cabin at time k (i)*,j*E 1,2, …, N). If it is not
And
if the difference is large, it indicates that the two sensor nodes do not support each other at the time k, and if the difference is small, it indicates that the two nodes support each other. If a node is supported by a large number of nodes simultaneously, it is considered to be a valid fire probability assessment. Otherwise, the probability value evaluated by the node will be given a lower weight in the fusion process.
To show
And
the mutual support degree between the fuzzy sets is defined as a decay exponential function in the fuzzy set:
the parameter α is adjustable to adjust the fusion accuracy, typically set to 0.8[16 ].
Support function
Represents a node i*And node j*The mutual support degree at the k-th time can be generally expressed in a matrix form:
in obtaining a support matrix
Thereafter, a mutual support relationship between the nodes may be determined. For the
I th of (1)*The columns of the image data are,
the larger, at node i*The reliability of the obtained fire probability evaluation value is higher, otherwise, the reliability is lower.
2.2 construct augmentation support matrix
In order to integrate all the evaluation values at each sampling time, an extended support matrix is defined, which increases the dimension of the support matrix by one row and one column. The purpose of constructing the new dimension support degree matrix is to measure the mutual support degree between all current evaluation values and previous evaluation values, so as to adaptively allocate weight coefficients for each node.
The specific steps for constructing the augmentation matrix are as follows:
when k is 1, the average of the previous N evaluation values is used
As an initial fire probability estimate
And (3) calculating newly added rows and columns of the k-time augmentation support degree matrix:
The augmented support matrix at time k may be defined as:
2.3 weighted fusion node evaluation
Is provided with
To represent
Fusion weight coefficient (w)N+1(k) Is the (k-1) time fusion result
The weight coefficient of (d),
satisfies the following conditions:
in the augmented support matrix
In the middle, by
I th of (1)*Integrating the columns to obtain
The reliability of (a) is high, and therefore,
is to
Is calculated. Set a set of vectors
Each element is a pair
The result of the integration is carried out,
comprises the following steps:
wherein i*,j*＝1,2,…,N+1。
According to equation (12), equation (14) becomes:
wherein W ═ W1(k),w2(k),…wN+1(k)]T，A＝[a1(k),a2(k),…aN+1(k)]T。
Due to the fact that
Is a non-negative symmetric matrix formed by the Flobenius-Pelon theorem,
having a maximum eigenvalue lambda*(λ*＞0)。
Therefore, the temperature of the molten metal is controlled,
calculating lambda*The corresponding positive eigenvector a, has:
W＝λ*A (17)
the proportional relationship between the current available weight coefficients and the feature vectors is:
calculating a weight coefficient of each node according to equation (13):
wherein i*,j*＝1,2,…,N+1。
Then, the final fused expression is obtained:
wherein
Represent willAnd comparing the probability of fire occurrence in the cabin after the evaluation values of all the sensor nodes are fused with the fire threshold probability, and if the probability is greater than the threshold probability, sending a fire alarm signal by the system.
The fire occurrence threshold is set to: the temperature was 55 ℃, the smoke concentration was 700ppm, the CO concentration was 20ppm, and the infrared ray intensity was 760 Lux. The variables are normalized and then are input into a formula (8), and the obtained result is the fire threshold probability.
When the probability of fire occurrence in the cabin is calculated by using the variable weighted fusion and node fire probability evaluation data weighted fusion method, the steps from one to five in the first section are offline calculation processes, and the purpose of obtaining the variable x is to obtaini(i ═ 1,2,3,4) weights. When the fire condition in the cabin is detected on line, firstly, environment data detected by a sensor on each node is normalized by using a formula (7), and then, the fire occurrence probability is evaluated by using a formula (8); then based on the mutual support of the node data, the probability value evaluated for each node is subjected to self-adaptive weight coefficient distribution, and the probability value of actual fire occurrence in the cabin is calculated after weighted fusion
Finally, will
And comparing the probability with a threshold value to judge whether a fire alarm signal is sent out.
3. Simulation experiment verification
50 sensor nodes are randomly and uniformly deployed in a space, each node consists of a temperature sensor, a smoke concentration sensor, a CO concentration sensor and an infrared ray intensity sensor, and a node deployment schematic diagram is shown in figure 1. The environmental parameter collected by each node is
( i *1,2, …,100), the variable units x1 is deg.c, x2 is ppm, x3 is ppm and x4 is Lux. In the same experimental context, the variables that influence the probability of fire occurrence are first calculated: the weight of the temperature, the smoke concentration, the CO concentration and the infrared ray intensity is taken up, and then the fire disaster provided by the invention is detected on lineThe method is compared with a grey fuzzy neural network fusion algorithm and a fuzzy logic fusion algorithm in the related technology, and the superiority of an online detection method (namely the fire detection method) in the aspects of timeliness, accuracy and fault tolerance of fire detection is verified.
3.1 variable weight calculation
The weights of the variables are found below according to the first section with respect to the relevant introduction to the calculation of the weights of the variables. Before calculating the weights of the variables, the assumption is first made that the variable x isiThe importance relationship of the influence degree of (i ═ 1,2,3,4) on the fire occurrence is ranked as:
assume that 1: x is the number of1＞x2＞x3＞x4；
Assume 2: x is the number of1＞x2＞x4＞x3；
Assume that 3: x is the number of1＞x3＞x2＞x4；
Assume 4: x is the number of1＞x3＞x4＞x2；
Assume that 5: x is the number of1＞x4＞x2＞x3；
Assume 6: x is the number of1＞x4＞x3＞x2；
Assume that 24: x is the number of4＞x3＞x2＞x1
In 24 possible cases, the 24 hypotheses are A1, A2, … and A24.
Table 2 calculation results of variable weight interval on assumption 13
Based on the relationship of the influence degrees between the variables in the above assumptions, this experiment calculated the weight interval of each variable when k is 5 (where B1 to B5 are judgment matrices), and calculated the optimal weight of each variable within the weight interval using equations (5) and (6). Table 2 describes the correlation process in section 1 for weight calculation (where table 2 records the weight interval found under assumption 13), and table 3 lists the corresponding output results:
TABLE 3 output results of optimal weights of variables under all assumptions
Five sets of environmental parameters X1 ═ 38,500,18,188 were selected for this experiment]；X2＝[89,580,30,200]；X3＝[39,450,70,229]；X4＝[51,750,38,199]；X5＝[39,480,19,705]. The probability of the possibility of fire occurrence under the assumptions of A1, A2, … and A24 is obtained by using the online detection method of the invention under five groups of parameters. When the input parameters are the same, the fire occurrence probability calculated under the assumptions of A1, A2, … and A24 is compared with the fire occurrence probability calculated by a fuzzy fusion algorithm based on an 'IF THEN' statement (experiments on automobiles prove that the method is feasible in fire detection), and curves shown in FIG. 2 represent error curves of the fire occurrence probabilities obtained by comparing A1, A2, … and A24 with the method provided by the invention, and the error is the minimum under the assumption condition of A13 according to simulation results. It follows that the variable xiThe influence degree relation of (i ═ 1,2,3,4) on the fire occurrence probability is x3＞x1＞x2＞x4The weight of each variable is [0.1474,0.1439,0.6492,0.0595 ]]。
3.2 analysis of timeliness and accuracy of on-line detection algorithm
From the point of view of the detection of a fire and the time required for the alarm signal. Environmental variable parameter detected by 50 nodes in space
And respectively substituting the three fire detection algorithms into the three fire detection algorithms, carrying out 100 independent repeated experiments, and recording respective detection time. Wherein x1In the interval (55,100)]Internal value, x2In the interval (700,1000)]Internal value, x3In the interval (20,100)]Internal value, x4In the interval (760,1000)]An internal value. It can be seen from the simulation curve in fig. 3 that, under the condition of the same input variables, the online detection algorithm of the present invention can complete fire detection and alarm within 10s, while the gray fuzzy neural network fusion algorithm in the related art completes fire detection and alarm within 23s, and the fuzzy logic fusion algorithm in the related art completes fire detection and alarm within 20 s. Experiments show that when the input variables are the same, the online detection algorithm provided by the invention is superior to other two algorithms in the aspect of fire detection timeliness.
The core of the online detection algorithm is to calculate the weight occupied by each variable, and a set of weights [0.6492,0.1474,0.1439,0.0595] in the above experiment is obtained when the number of judgment matrixes is 5. When k is 5, the false fire alarm rate under the background of the experiment is 3%, and the relationship between the number k selected by the analysis and judgment matrix and the false fire detection alarm rate is used as an initial condition. The simulation curve in fig. 4 shows that the more reasonable judgment matrixes are selected, the more accurate the obtained variable weight is, and meanwhile, the more accurate the evaluation result of each node on the fire occurrence probability is. When k is larger than 20, the false alarm rate is reduced to be below 0.5% by the online detection algorithm, and the fire detection accuracy is effectively improved.
3.3 Fault tolerance analysis of Online detection Algorithm
On the basis of 50 sensor nodes, when the fault node is gradually increased, the deviation between the detection value and the real value of the fire occurrence probability is observed. The environmental temperature was set to 38 ℃, the smoke concentration was 650ppm, the CO concentration was 14ppm, the infrared ray intensity was 700lux and assumed to be constant, and the fire occurrence probability detected when all nodes were operating normally was assumed to be the true value, as shown in table 4.
TABLE 4 true fire probability values corresponding to three detection and calculation methods
| Fire detection algorithm | Probability of fire |
| Grey fuzzy neural network fusion method | 30.56％ |
| Fuzzy logic fusion method | 29.48％ |
| In-line detection method of the present invention | 28.62％ |
The Square Error (SE) is used as a standard for evaluating the detection accuracy of the three algorithms, and the SE is defined as:
wherein j is a detection algorithm index (j ═ 1,2, 3); p is a radical ofjRepresenting the true value of the fire occurrence probability under the j algorithm;
representing the estimated value of the fire occurrence probability after fusing 50 nodes under the j algorithm.
As can be seen from the simulation curve of fig. 5, when the number of fault nodes reaches 10, SE of the fuzzy neural network information fusion algorithm and the fuzzy logic control system with feedback in the related art will increase rapidly, and the algorithm detection accuracy will decrease, which cannot meet the requirements of the actual application system. The online detection algorithm provided by the invention is used for self-adaptively distributing weight coefficients based on mutual support degree among data when node data are fused. Therefore, when the failed node increases, the increase of the SE value thereof is gentle, and a large deviation occurs after the failed node reaches 30. Therefore, in an actual fire detection system, the fault tolerance of the online detection algorithm is better.
4 conclusion
The fire online detection algorithm of the invention firstly utilizes the improved AHP to calculate the weight of each variable at each node of the WSN, and provides a new variable weighting fusion algorithm to evaluate the fire occurrence probability, so that the system can still timely and accurately detect the fire when more input variables exist. Experiments show that the algorithm can complete the detection and alarm of the fire within 10s, compared with other fire detection algorithms, the time for detecting and alarming the fire is greatly shortened, the fire spreading is effectively avoided, and the subsequent fire extinguishing work is more smooth to a certain extent. When the WSN is used for collecting environmental data, the problem that detection precision is damaged and even a system is paralyzed due to sensor faults exists. By adaptively distributing weight coefficients to each node, the influence of deviation data acquired by a fault sensor on a fusion result is reduced, so that the online detection algorithm has strong fault-tolerant capability.
Corresponding to the fire detection method provided by the embodiment of the invention, the embodiment of the invention also provides a fire detection device, which comprises:
the first determining module is used for determining a plurality of fire influence factors of each node in the cabin;
the second determining module is used for determining a fire influence weight value of each fire influence factor in the plurality of fire influence factors of each node;
the third determining module is used for determining the evaluation value of the fire occurrence probability of each node in the cabin according to the fire influence weight value of each fire influence factor;
the fourth determining module is used for determining the actual probability of fire in the cabin according to the evaluation value of each node on the probability of fire in the cabin;
and the judging module is used for judging whether to send out fire alarm or not according to the actual probability of fire in the cabin.
In one embodiment, the second determining module comprises:
the determining submodule is used for determining a judgment matrix of each fire influence factor;
the detection submodule is used for carrying out consistency detection on the judgment matrix of each fire influence factor;
and the calculating submodule is used for calculating the fire influence weight value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test.
In one embodiment, the calculation submodule is specifically configured to:
when a plurality of judgment matrixes of the fire influence factors exist, calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrixes of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
In one embodiment, the third determining module comprises:
the first determining submodule is used for determining a fire influence weight region of each fire influence factor according to the fire influence weight value of each fire influence factor;
the second determining submodule is used for determining a target influence function of each fire influence factor and a constraint condition of the target influence function according to the fire influence weight interval of each fire influence factor;
the calculating submodule is used for calculating the optimal fire influence weight value of each fire influence factor of each node according to the target influence function of each fire influence factor and the constraint condition of the target influence function;
and the third determining submodule is used for determining the evaluation value of the fire occurrence probability of each node to the cabin according to the optimal fire influence weight value of each fire influence factor of each node and the actual factor value of each fire influence factor of each node.
In one embodiment, the fourth determining module is specifically configured to:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
and determining the actual probability of fire in the cabin according to the augmented support matrix.
In an embodiment, the fourth determining module is further specifically configured to:
determining an evaluation credibility coefficient of an evaluation value of the probability of fire occurrence in the cabin by each node according to the augmentation support matrix;
and determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin and the evaluation confidence coefficient of each node on the evaluation value of the probability of the fire in the cabin.
In one embodiment, the determining module is specifically configured to:
calculating a probability threshold value of the fire in the cabin according to the fire occurrence critical factor value of each fire influence factor of each node;
and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin and the probability threshold value of fire in the cabin.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims (6)
1. A method of fire detection, comprising the steps of:
determining a plurality of fire influence factors of each node in an engine room;
determining a fire influence weight value of each fire influence factor in the plurality of fire influence factors of each node;
determining an evaluation value of each node on the probability of the fire in the cabin according to the fire influence weight value of each fire influence factor;
determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin;
judging whether to send out a fire alarm or not according to the actual probability of the fire in the cabin;
determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin, wherein the determining comprises the following steps:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
determining the probability of actual fire in the cabin according to the augmentation support matrix;
determining an evaluation value of each node for the probability of the fire in the cabin according to the fire influence weight value of each fire influence factor, wherein the evaluation value comprises the following steps:
determining a fire influence weight region of each fire influence factor according to the fire influence weight value of each fire influence factor;
determining a target influence function of each fire influence factor and a constraint condition of the target influence function according to the fire influence weight interval of each fire influence factor;
calculating an optimal fire influence weight value of each fire influence factor of each node according to the target influence function of each fire influence factor and the constraint condition of the target influence function;
and determining the evaluation value of each node on the probability of the fire in the cabin according to the optimal fire influence weight value of each fire influence factor of each node and the actual factor value of each fire influence factor of each node.
2. The method of claim 1,
the determining a fire influence weight value of each fire influence factor of the plurality of fire influence factors of each node includes:
determining a judgment matrix of each fire influence factor;
carrying out consistency check on the judgment matrix of each fire influence factor;
and calculating the fire influence weighted value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test.
3. The method of claim 2,
when a plurality of judgment matrixes of the fire influencing factors exist, calculating the fire influencing weighted value of each fire influencing factor according to the judgment matrixes of the fire influencing factors passing the consistency test, wherein the method comprises the following steps:
calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrices of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
4. The method of claim 1,
determining the probability of the actual fire in the cabin according to the augmented support matrix, wherein the determining comprises the following steps:
determining an evaluation credibility coefficient of an evaluation value of the probability of fire occurrence in the cabin by each node according to the augmentation support matrix;
and determining the actual probability of the fire in the cabin according to the evaluation value of each node on the probability of the fire in the cabin and the evaluation confidence coefficient of each node on the evaluation value of the probability of the fire in the cabin.
5. The method according to any one of claims 1 to 4,
the judging whether to send out fire alarm according to the actual probability of fire in the cabin comprises the following steps:
calculating a probability threshold value of the fire in the cabin according to the fire occurrence critical factor value of each fire influence factor of each node;
and judging whether to send out fire alarm or not according to the actual probability of fire in the cabin and the probability threshold value of fire in the cabin.
6. A fire detection device, the device comprising:
the first determining module is used for determining a plurality of fire influence factors of each node in the cabin;
the second determining module is used for determining a fire influence weight value of each fire influence factor in the plurality of fire influence factors of each node;
the third determining module is used for determining the evaluation value of the fire occurrence probability of each node in the cabin according to the fire influence weight value of each fire influence factor;
the fourth determining module is used for determining the actual probability of fire in the cabin according to the evaluation value of each node on the probability of fire in the cabin;
the judging module is used for judging whether to send out a fire alarm or not according to the actual probability of the fire in the cabin;
the fourth determining module is specifically configured to:
establishing a support function about the evaluation value between any two nodes in each node according to the evaluation value of each node on the probability of fire occurrence in the cabin;
establishing an initial support matrix between the nodes about the evaluation values according to the support function between any two nodes about the evaluation values;
constructing an augmented support matrix for the initial support matrix among the evaluation values of each node; wherein the augmented support matrix is one row and one column more than the initial support matrix;
determining the probability of actual fire in the cabin according to the augmentation support matrix;
the second determining module includes:
the determining submodule is used for determining a judgment matrix of each fire influence factor;
the detection submodule is used for carrying out consistency detection on the judgment matrix of each fire influence factor;
the calculation submodule is used for calculating the fire influence weight value of each fire influence factor according to the judgment matrix of each fire influence factor passing the consistency test;
the calculation submodule is specifically configured to:
when a plurality of judgment matrixes of the fire influence factors exist, calculating a preset number of fire influence weighted values of the fire influence factors according to each judgment matrix of the fire influence factors passing consistency check, wherein the preset number is equal to the number of the judgment matrixes of the fire influence factors passing consistency check; the preset number is a positive integer greater than or equal to 2;
and determining the fire influence weight interval of each fire influence factor according to the preset number of fire influence weight values of each fire influence factor.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010673849.0A CN111784994B (en) | 2020-07-14 | 2020-07-14 | Fire detection method and device |
| PCT/CN2021/102072 WO2022012295A1 (en) | 2020-07-14 | 2021-06-24 | Fire detection method and apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010673849.0A CN111784994B (en) | 2020-07-14 | 2020-07-14 | Fire detection method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111784994A CN111784994A (en) | 2020-10-16 |
| CN111784994B true CN111784994B (en) | 2021-11-30 |
Family
ID=72767144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010673849.0A Active CN111784994B (en) | 2020-07-14 | 2020-07-14 | Fire detection method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN111784994B (en) |
| WO (1) | WO2022012295A1 (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111784994B (en) * | 2020-07-14 | 2021-11-30 | 中国民航大学 | Fire detection method and device |
| CN114387755A (en) * | 2021-12-13 | 2022-04-22 | 煤炭科学技术研究院有限公司 | Mine smoke detection method, device, processor and system |
| CN115498764B (en) * | 2022-09-15 | 2023-05-19 | 东营金丰正阳科技发展有限公司 | Oil well control cabinet with electric power control system |
| CN117612319A (en) * | 2024-01-24 | 2024-02-27 | 上海意静信息科技有限公司 | Alarm information grading early warning method and system based on sensor and picture |
| CN117679694B (en) * | 2024-02-04 | 2024-04-09 | 东营昆宇电源科技有限公司 | Fire-fighting pipeline control system for energy storage battery compartment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5673020A (en) * | 1994-03-30 | 1997-09-30 | Nohmi Bosai Ltd. | Early stage fire detecting apparatus |
| CN108305426A (en) * | 2017-07-25 | 2018-07-20 | 四川雷克斯智慧科技股份有限公司 | Fire scene intelligent analysis system |
| CN109686036A (en) * | 2019-01-09 | 2019-04-26 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire monitoring method, device and edge calculations device |
| CN109978005A (en) * | 2019-02-25 | 2019-07-05 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire alarm method, device, storage medium and terminal device |
| CN110011976A (en) * | 2019-03-07 | 2019-07-12 | 中国科学院大学 | A kind of network attack damage capability quantitative estimation method and system |
| CN111126701A (en) * | 2019-12-25 | 2020-05-08 | 兰州交通大学 | Forest fire danger early warning method based on GIS and meteorological monitoring network |
| CN111243214A (en) * | 2020-01-19 | 2020-06-05 | 佛山科学技术学院 | Industrial machine room fire monitoring system and method based on machine learning |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2491542A4 (en) * | 2009-10-20 | 2016-07-13 | Sensortran Inc | Calibrated linear fire detection using dts systems |
| JP6629524B2 (en) * | 2015-06-02 | 2020-01-15 | 能美防災株式会社 | Fire detection system |
| CN104966372B (en) * | 2015-06-09 | 2017-09-29 | 四川汇源光通信有限公司 | The forest fire intelligent identifying system and method for multi-data fusion |
| CN105678419A (en) * | 2016-01-05 | 2016-06-15 | 天津大学 | Fine grit-based forest fire hazard probability forecasting system |
| CN105741474B (en) * | 2016-04-11 | 2018-09-25 | 泉州师范学院 | A kind of fire alarm method based on multisensor |
| CN106501451B (en) * | 2016-10-21 | 2018-10-19 | 中国科学院上海高等研究院 | A kind of disposition optimization method, system and the server of gas sensor |
| CN206688057U (en) * | 2017-04-14 | 2017-12-01 | 中国民航大学 | A kind of aircraft monitoring fire extinguishing system in real time |
| CN108269379A (en) * | 2018-03-27 | 2018-07-10 | 吉林建筑大学 | Combined fire detector based on Internet technology |
| CN110197303B (en) * | 2019-05-30 | 2021-02-26 | 浙江树人学院（浙江树人大学） | Firefighter rescue scheduling method adaptive to dynamic changes of fire |
| CN110334660A (en) * | 2019-07-08 | 2019-10-15 | 天津城建大学 | A kind of forest fire monitoring method based on machine vision under the conditions of greasy weather |
| CN111784994B (en) * | 2020-07-14 | 2021-11-30 | 中国民航大学 | Fire detection method and device |
-
2020
- 2020-07-14 CN CN202010673849.0A patent/CN111784994B/en active Active
-
2021
- 2021-06-24 WO PCT/CN2021/102072 patent/WO2022012295A1/en active Application Filing
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5673020A (en) * | 1994-03-30 | 1997-09-30 | Nohmi Bosai Ltd. | Early stage fire detecting apparatus |
| CN108305426A (en) * | 2017-07-25 | 2018-07-20 | 四川雷克斯智慧科技股份有限公司 | Fire scene intelligent analysis system |
| CN109686036A (en) * | 2019-01-09 | 2019-04-26 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire monitoring method, device and edge calculations device |
| CN109978005A (en) * | 2019-02-25 | 2019-07-05 | 深圳市中电数通智慧安全科技股份有限公司 | A kind of fire alarm method, device, storage medium and terminal device |
| CN110011976A (en) * | 2019-03-07 | 2019-07-12 | 中国科学院大学 | A kind of network attack damage capability quantitative estimation method and system |
| CN111126701A (en) * | 2019-12-25 | 2020-05-08 | 兰州交通大学 | Forest fire danger early warning method based on GIS and meteorological monitoring network |
| CN111243214A (en) * | 2020-01-19 | 2020-06-05 | 佛山科学技术学院 | Industrial machine room fire monitoring system and method based on machine learning |
Non-Patent Citations (2)
| Title |
|---|
| 层次分析法权重计算方法分析及其应用研究;邓雪等;《数学的实践与认识》;20120430;第42卷(第7期);第93-100页 * |
| 应用改进支持度矩阵模型的光纤光栅高温传感系统;庞丹丹等;《光电子· 激光》;20121130;第23卷(第11期);第2045-2050页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2022012295A1 (en) | 2022-01-20 |
| CN111784994A (en) | 2020-10-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111784994B (en) | Fire detection method and device | |
| US11314242B2 (en) | Methods and systems for fault detection and identification | |
| CN110912867B (en) | Intrusion detection method, device, equipment and storage medium for industrial control system | |
| CN111539515B (en) | Complex equipment maintenance decision method based on fault prediction | |
| CN113486078B (en) | Distributed power distribution network operation monitoring method and system | |
| CN109143408B (en) | Dynamic region combined short-time rainfall forecasting method based on MLP | |
| Haroush et al. | A statistical framework for efficient out of distribution detection in deep neural networks | |
| CN110428191B (en) | Method for identifying fragile nodes of power distribution network | |
| CN111179592B (en) | Urban traffic prediction method and system based on spatio-temporal data flow fusion analysis | |
| Wang et al. | Fault detection and diagnosis for multiple faults of VAV terminals using self-adaptive model and layered random forest | |
| CN115618273B (en) | Railway track state evaluation method and system based on parallel graph convolution neural network | |
| CN112232370A (en) | Fault analysis and prediction method for engine | |
| CN113408114A (en) | Method and system for evaluating vulnerability threat degree of power monitoring system equipment | |
| CN115277354A (en) | Fault detection method for command control network management system | |
| CN112580254B (en) | Multi-dimensional analysis and early warning method for converter transformer based on multi-level decision fusion | |
| CN112437440B (en) | Malicious collusion attack resistance method based on correlation theory in wireless sensor network | |
| CN116151799A (en) | BP neural network-based distribution line multi-working-condition fault rate rapid assessment method | |
| CN113688506B (en) | Potential atmospheric pollution source identification method based on multi-dimensional data such as micro-station and the like | |
| KR102470520B1 (en) | Failure Mode and Effect Analysis(FMEA) Method | |
| CN111241482B (en) | Multi-parameter system-oriented abnormal working state detection method and system | |
| JP6587950B2 (en) | Program, apparatus, and method capable of detecting time series change point by scalar feature | |
| de Jesus et al. | Systematic failure detection and correction in environmental monitoring systems | |
| CN110826690A (en) | Equipment state identification method and system and computer readable storage medium | |
| CN113379455B (en) | Order quantity prediction method and equipment | |
| CN113486742B (en) | Fault identification method, device and system and computer readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |