Method for realizing automatic smoke detection by using powdery smoke generating body and heating wire

Abstract

The invention provides a method for realizing automatic smoke detection by using a powdery smoke generating body and a heating wire, and relates to the technical field of fire alarm, and the method comprises the steps: placing a detection device in a to-be-detected area; the detection device is used for periodically generating test smoke; controlling the detection device to generate test smoke, performing detection by using the smoke alarm to be detected to obtain a smoke concentration time sequence, and performing data cleaning and standardization processing on the smoke concentration time sequence; performing feature extraction according to the standardized data sequence to obtain smoke features; judging whether test smoke is detected or not according to the smoke characteristics, and if so, determining a to-be-detected smoke alarm in the to-be-detected area to be in a normal working state; and if not, determining the to-be-detected smoke alarm in the to-be-detected area to be in an abnormal working state, and sending an alarm signal. According to the invention, by periodically generating the test smoke and monitoring the smoke concentration in real time, the existence of the smoke can be timely and accurately detected, and the response speed of the smoke alarm is improved.

Description

Technical Field

[0001] The invention relates to the technical field of fire alarm, and in particular to a method for realizing automatic smoke detection by utilizing a powdery smoke-generating body and a heating wire. Background Art

[0002] In existing fire alarm systems, smoke detectors usually require regular manual testing with a smoke gun, which is not only labor-intensive but also inefficient. During batch testing, a large number of alarm signals are easily generated, causing trouble for management. Therefore, an automated smoke detection method is urgently needed to improve detection efficiency and reduce unnecessary alarms. Summary of the invention

[0003] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a method for automatic smoke detection using a powdered smoke-generating body and a heating wire, which can show good application prospects in the field of smoke detection and can effectively improve the reliability and efficiency of safety monitoring.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A method for realizing automatic smoke detection using a powdered smoke-generating body and a heating wire, comprising:

[0006] Placing a detection device in the area to be detected; the detection device is used to periodically generate test smoke;

[0007] Controlling the detection device to generate test smoke, and using the smoke alarm to be detected to perform detection, to obtain a smoke concentration time series;

[0008] Performing data cleaning on the smoke concentration time series to obtain a cleaned data feature sequence;

[0009] Performing standardization processing on the cleaned data feature sequence to obtain a standardized data sequence;

[0010] Perform feature extraction according to the standardized data sequence to obtain smoke features;

[0011] It is determined whether the test smoke is detected according to the smoke characteristics. If so, the smoke alarm to be detected in the area to be detected is determined to be in a normal working state; if not, the smoke alarm to be detected in the area to be detected is determined to be in an abnormal working state and an alarm signal is issued.

[0012] Preferably, the detection device comprises a powder smoke-generating body, a heating wire, a cavity structure and a heating module;

[0013] The powder smoke body and the heating wire are both arranged in the cavity structure; the heating module is connected to the heating wire; the powder smoke body is used to generate the test smoke in a heated state; the heating wire is used to heat the powder smoke body; the heating module is used to periodically energize the heating wire.

[0014] Preferably, data cleaning is performed on the smoke concentration time series to obtain a cleaned data feature sequence, including:

[0015] Grouping the smoke concentration time series according to a preset collection period to obtain multiple data groups;

[0016] Calculate the difference coefficient between the current data group and the previous data group in sequence;

[0017] Determine whether the value of the coefficient of difference is within a preset range;

[0018] If the value of the difference coefficient is not within the preset range, the corresponding data group is removed;

[0019] If the value of the difference coefficient is within the preset range, the corresponding data group is retained until all data groups are traversed to obtain the cleaned data feature sequence.

[0020] Preferably, the coefficient of difference calculation formula is:

[0021]

[0022] Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

[0023] Preferably, feature extraction is performed according to the standardized data sequence to obtain smoke features, including:

[0024] Extracting time domain features from the standardized data sequence to obtain time domain features;

[0025] Extracting frequency domain features from the standardized data sequence to obtain frequency domain features;

[0026] Extracting time-frequency domain features from the standardized data sequence to obtain time-frequency domain features;

[0027] Perform feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fusion feature;

[0028] Dividing the fused features into a training set, a validation set, and a test set, evaluating the performance of the fused data on the validation set, and obtaining an evaluation result;

[0029] Adjust the fusion strategy and feature selection according to the evaluation results to optimize the fusion effect.

[0030] Preferably, the time domain features include mean, variance, peak, slope and trend; the frequency domain features include power spectrum density and Fourier transform results; the time-frequency domain features include short-term and long-term change results.

[0031] Preferably, performing feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fused feature includes:

[0032] The time domain feature, the frequency domain feature and the time-frequency domain feature are respectively vectorized to obtain a first feature vector, a second feature vector and a third feature vector; the expression of the first feature vector A is A=(a1, a2, ..., a m ); the expression of the second eigenvector B is B=(b1,b2,…,b n ); the expression of the third eigenvector C is C=(c1,c2,…,c k ), where a m is the mth PCR detection feature, b n is the nth clinical data feature, c k is the kth environmental data feature;

[0033] The first feature vector, the second feature vector and the third feature vector are sequentially concatenated into a comprehensive long vector to obtain a representation vector F of the fused feature; the expression of the representation vector F of the fused feature is: F=(A, B, C).

[0034] Preferably, judging whether test smoke is detected according to the smoke characteristics comprises:

[0035] Constructing an initial detection network and obtaining a sample smoke feature data set; the network structure of the initial detection network is a CNN-LSTM model;

[0036] Training the initial detection network according to the sample smoke feature data set to obtain a trained smoke detection model;

[0037] The smoke characteristics are input into the smoke detection model to obtain a detection result.

[0038] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0039] The present invention provides a method for realizing automatic smoke detection by using a powdered smoke-generating body and a heating wire, comprising: placing a detection device in a detection area; the detection device is used to periodically generate test smoke; controlling the detection device to generate test smoke, and using a smoke alarm to be detected to detect, to obtain a smoke concentration time series; performing data cleaning on the smoke concentration time series to obtain a cleaned data feature sequence; performing standardization processing on the cleaned data feature sequence to obtain a standardized data sequence; performing feature extraction based on the standardized data sequence to obtain smoke features; judging whether the test smoke is detected based on the smoke features, if so, determining the smoke alarm to be detected in the detection area as a normal working state; if not, determining the smoke alarm to be detected in the detection area as an abnormal working state, and sending an alarm signal. The present invention can detect the presence of smoke in a timely and accurate manner by periodically generating test smoke and monitoring the smoke concentration in real time, thereby improving the response speed of the smoke alarm. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0041] Figure 1 A flow chart of a method provided by an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of a detection device module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] The purpose of the present invention is to provide a method for realizing automatic smoke detection by utilizing a powdered smoke-generating body and a heating wire, which shows good application prospects in the field of smoke detection and can effectively improve the reliability and efficiency of safety monitoring.

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0046] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a method for realizing automatic smoke detection using a powdered smoke-generating body and a heating wire, comprising:

[0047] Step 100: placing a detection device in the area to be detected; the detection device is used to periodically generate test smoke;

[0048] Step 200: Control the detection device to generate test smoke, and use the smoke alarm to be detected to obtain a smoke concentration time series;

[0049] Step 300: performing data cleaning on the smoke concentration time series to obtain a cleaned data feature sequence;

[0050] Step 400: performing standardization processing on the cleaned data feature sequence to obtain a standardized data sequence;

[0051] Step 500: extracting features according to the standardized data sequence to obtain smoke features;

[0052] Step 600: Determine whether test smoke is detected according to the smoke characteristics. If so, determine the smoke alarm to be detected in the area to be detected to be in a normal working state; if not, determine the smoke alarm to be detected in the area to be detected to be in an abnormal working state and send out an alarm signal.

[0053] Preferably, the detection device comprises a powder smoke-generating body, a heating wire, a cavity structure and a heating module;

[0054] The powder smoke body and the heating wire are both arranged in the cavity structure; the heating module is connected to the heating wire; the powder smoke body is used to generate the test smoke in a heated state; the heating wire is used to heat the powder smoke body; the heating module is used to periodically energize the heating wire.

[0055] Specifically, Figure 2 As shown, this embodiment uses a powdered smoke body, which is placed in a cavity structure. The cavity structure has a built-in heating wire, which is automatically energized according to a period (such as a specified period of one month, or other settable time). When smoke is detected, it is considered available and no signal is sent to the upper computer to avoid a large number of alarm signals during batch detection. If smoke cannot be detected, an alarm is given, and the alarm signal can be a dry node or a bus protocol signal. The bus protocol signal can count the detection time history. When the automatic detection function is realized, it is convenient for the superior inspection department to remotely check and understand the situation, analyze hidden dangers, guide the promotion of reliable products, reduce manpower to test with smoke guns on site, and improve efficiency.

[0056] Furthermore, the powdered smoke body itself has a service life of more than 20 years, and the powdered smoke body can meet the requirements of more than 400 smoke generation times and meet 20 years of daily inspection.

[0057] Furthermore, the powdery smoke-generating body of the present embodiment evaporates after emitting smoke, and the upper layer of powder continues to fall on the heating wire to ensure the next smoke emission. The powdery smoke-generating body is made of polymer materials. The powdery smoke-generating body only produces smoke when heated, and extinguishes itself when encountering fire to ensure safety.

[0058] Preferably, data cleaning is performed on the smoke concentration time series to obtain a cleaned data feature sequence, including:

[0059] Grouping the smoke concentration time series according to a preset collection period to obtain multiple data groups;

[0060] Calculate the difference coefficient between the current data group and the previous data group in sequence;

[0061] Determine whether the value of the coefficient of difference is within a preset range;

[0062] If the value of the difference coefficient is not within the preset range, the corresponding data group is removed;

[0063] If the value of the difference coefficient is within the preset range, the corresponding data group is retained until all data groups are traversed to obtain the cleaned data feature sequence.

[0064] Preferably, the difference coefficient calculation formula is:

[0065]

[0066] Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

[0067] Specifically, since the smoke concentration sensor may be affected by its own parameters or environmental factors, the measured value collected by the sensor at a certain moment may deviate greatly from the actual value. Therefore, the present application can screen out abnormal monitoring values ​​through the difference coefficient, thereby ensuring the accuracy of historical data.

[0068] Optionally, the standardization process for the cleaned data feature sequence in this embodiment is as follows:

[0069] The process of standardizing the cleaned data feature sequence mainly includes the following steps: First, calculate the mean and standard deviation of the cleaned data feature sequence. The mean is used to indicate the center position of the data, while the standard deviation reflects the degree of dispersion of the data. Then, for each data point in the cleaned data feature sequence, use the following formula for standardization:

[0070]

[0071] Among them, Z is the standardized data point, X is the original data point, mean is the mean of the cleaned data feature sequence, and std is the standard deviation. Through this formula, all data points will be converted into a standard normal distribution with 0 as the mean and 1 as the standard deviation, thereby eliminating the dimensional influence between different features and making the standardized data sequences comparable on the same scale. This embodiment ensures that the subsequent feature extraction and analysis can be more accurate and effective through the above steps.

[0072] Preferably, feature extraction is performed according to the standardized data sequence to obtain smoke features, including:

[0073] Extracting time domain features from the standardized data sequence to obtain time domain features;

[0074] Extracting frequency domain features from the standardized data sequence to obtain frequency domain features;

[0075] Extracting time-frequency domain features from the standardized data sequence to obtain time-frequency domain features;

[0076] Perform feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fusion feature;

[0077] Dividing the fused features into a training set, a validation set, and a test set, evaluating the performance of the fused data on the validation set, and obtaining an evaluation result;

[0078] Adjust the fusion strategy and feature selection according to the evaluation results to optimize the fusion effect.

[0079] Preferably, the time domain features include mean, variance, peak, slope and trend; the frequency domain features include power spectrum density and Fourier transform results; the time-frequency domain features include short-term and long-term change results.

[0080] Specifically, the feature extraction step of this embodiment includes:

[0081] 1. Time domain feature extraction:

[0082] Mean and variance, calculate the mean and variance of the standardized data series, reflecting the overall concentration level and volatility.

[0083] Peak features identify the maximum and minimum values ​​in the sequence and the time when they occur, capturing extreme changes in concentration.

[0084] Slope and trend, through linear regression analysis, the slope and trend of the sequence are extracted to determine the direction of concentration change.

[0085] 2. Frequency domain feature extraction:

[0086] Fourier transform, perform fast Fourier transform (FFT) on the standardized data sequence, extract the frequency components, and analyze the periodic changes.

[0087] Power Spectral Density, calculates the power spectral density and identifies the main frequency components and their amplitudes.

[0088] 3. Time-frequency domain feature extraction:

[0089] Wavelet transform,Use wavelet transform to perform multi-scale analysis on the sequence, extract features at different time scales, and capture short-term and long-term changes.

[0090] Optionally, this embodiment also includes a feature selection process, which selects important features and removes redundant features through a feature selection algorithm (such as recursive feature elimination and LASSO regression), and uses methods such as principal component analysis (PCA) to reduce the dimension of features and improve computational efficiency.

[0091] Furthermore, in the evaluation phase, the present embodiment performs a performance evaluation on the fused features by applying predefined evaluation indicators (such as accuracy, recall, F1-score, etc.) on the validation set. According to the evaluation results, the features that perform well in feature fusion and the features that contribute more to the model performance are identified. Next, based on these evaluation results, a feature selection algorithm (such as recursive feature elimination, LASSO regression, or random forest feature importance) is used to screen out the most representative features to reduce redundancy and noise. Then, the feature fusion strategy is readjusted according to the selected features, such as changing the feature weighting method or selecting a different fusion algorithm (such as simple averaging, weighted averaging, or using a machine learning model for fusion). Finally, the evaluation process is repeated to verify whether the adjusted fusion strategy and feature selection have improved the performance of the model, ensuring that each adjustment is based on data-driven feedback, thereby achieving continuous optimization of the feature fusion effect.

[0092] Preferably, performing feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fused feature includes:

[0093] The time domain feature, the frequency domain feature and the time-frequency domain feature are respectively vectorized to obtain a first feature vector, a second feature vector and a third feature vector; the expression of the first feature vector A is A=(a1, a2, ..., a m ); the expression of the second eigenvector B is B=(b1,b2,…,b n ); the expression of the third eigenvector C is C=(c1,c2,…,c k ), where a m is the mth PCR detection feature, b n is the nth clinical data feature, c k is the kth environmental data feature;

[0094] The first feature vector, the second feature vector and the third feature vector are sequentially concatenated into a comprehensive long vector to obtain a representation vector F of the fused feature; the expression of the representation vector F of the fused feature is: F=(A, B, C).

[0095] Preferably, judging whether test smoke is detected according to the smoke characteristics comprises:

[0096] Constructing an initial detection network and obtaining a sample smoke feature data set; the network structure of the initial detection network is a CNN-LSTM model;

[0097] Training the initial detection network according to the sample smoke feature data set to obtain a trained smoke detection model;

[0098] The smoke characteristics are input into the smoke detection model to obtain a detection result.

[0099] Specifically, this embodiment first constructs an initial detection network, using a combined model of a convolutional neural network (CNN) and a long short-term memory network (LSTM) to fully utilize the advantages of CNN in feature extraction and the capabilities of LSTM in time series data processing. Next, collect and prepare a sample smoke feature data set, which should contain smoke features in a variety of situations to ensure the generalization ability of the model. Then, use the data set to train the initial detection network, optimize the network parameters through the back propagation algorithm, minimize the loss function between the predicted results and the true labels, until the model converges, and obtain a trained smoke detection model. During the training process, cross-validation and early stopping strategies can be used to prevent overfitting. Finally, the new smoke features are input into the trained smoke detection model, and the model will infer based on the input features, output the detection results, and determine whether the test smoke is detected. Through this process, efficient and accurate smoke detection can be achieved.

[0100] Furthermore, the present embodiment analyzes the detection results obtained from the smoke detection model. If the result of the model output indicates that the test smoke is detected (i.e., the detection result is "positive"), the system will automatically update the status of the smoke alarm in the area to be detected to "normal working state" and record the state for subsequent monitoring and analysis. On the contrary, if the test result shows that the test smoke is not detected (i.e., the detection result is "negative"), the system will immediately update the status of the smoke alarm in the area to be detected to "abnormal working state". In this case, the system will trigger the alarm mechanism and send out an audible and visual alarm signal to remind relevant personnel to pay attention to potential safety hazards. At the same time, the system can record the occurrence time and duration of the abnormal state for subsequent troubleshooting and maintenance. Through this process, it is possible to ensure effective monitoring and timely response of smoke alarms in different working states, thereby improving overall safety.

[0101] The beneficial effects of the present invention are as follows:

[0102] (1) The present invention can detect the presence of smoke promptly and accurately by periodically generating test smoke and monitoring the smoke concentration in real time, thereby improving the response speed of the smoke alarm.

[0103] (2) The present invention can extract effective smoke features by performing data cleaning, standardization and feature extraction on the smoke concentration time series, thereby providing a scientific basis for judging the presence of smoke and reducing the false alarm rate.

[0104] (3) The present invention realizes the automation of smoke detection, reduces the need for manual intervention, improves the intelligence level of the system, and can maintain stable detection performance under different environmental conditions.

[0105] (4) The present invention can quickly determine the working status of the smoke alarm by monitoring the smoke characteristics in real time, ensuring that the device operates in a normal working state. If an abnormality occurs, an alarm signal is issued in time, thereby enhancing safety.

[0106] (5) The present invention can adapt to different types of environments and smoke characteristics. Through feature extraction and threshold setting, it can be adjusted according to actual conditions and is suitable for a variety of application scenarios.

[0107] (6) The present invention can be integrated with other security monitoring systems to form a comprehensive security protection system and improve the overall security management level.

[0108] (7) Since the present invention can automatically determine the status of the equipment, the frequency of manual inspection is reduced, thereby reducing maintenance costs and improving the efficiency of equipment use.

[0109] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0110] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for realizing automatic smoke detection using a powdered smoke-generating body and a heating wire, characterized in that: include: Placing the detection device in the area to be detected; The detection device is used to periodically generate test smoke; Controlling the detection device to generate test smoke, and using the smoke alarm to be detected to perform detection, to obtain a smoke concentration time series; Performing data cleaning on the smoke concentration time series to obtain a cleaned data feature sequence; Performing standardization processing on the cleaned data feature sequence to obtain a standardized data sequence; Perform feature extraction according to the standardized data sequence to obtain smoke features; It is determined whether the test smoke is detected according to the smoke characteristics. If so, the smoke alarm to be detected in the area to be detected is determined to be in a normal working state; if not, the smoke alarm to be detected in the area to be detected is determined to be in an abnormal working state and an alarm signal is issued.

2. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 1, characterized in that: The detection device comprises a powder smoke-generating body, a heating wire, a cavity structure and a heating module; The powder smoke body and the heating wire are both arranged in the cavity structure; the heating module is connected to the heating wire; the powder smoke body is used to generate the test smoke in a heated state; the heating wire is used to heat the powder smoke body; the heating module is used to periodically energize the heating wire.

3. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 1, characterized in that: The smoke concentration time series is cleaned to obtain a cleaned data feature sequence, including: Grouping the smoke concentration time series according to a preset collection period to obtain multiple data groups; Calculate the difference coefficient between the current data group and the previous data group in sequence; Determine whether the value of the coefficient of difference is within a preset range; If the value of the difference coefficient is not within the preset range, the corresponding data group is removed; If the value of the difference coefficient is within the preset range, the corresponding data group is retained until all data groups are traversed to obtain the cleaned data feature sequence.

4. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 3, characterized in that: The coefficient of difference calculation formula is: Among them, p X,Y is the coefficient of difference, cov(X,Y) represents the covariance between the current data set X and the previous data set Y, α X represents the mean of the current data set X, β Y Represents the mean of the previous data set Y.

5. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 1, characterized in that: Feature extraction is performed according to the standardized data sequence to obtain smoke features, including: Extracting time domain features from the standardized data sequence to obtain time domain features; Extracting frequency domain features from the standardized data sequence to obtain frequency domain features; Extracting time-frequency domain features from the standardized data sequence to obtain time-frequency domain features; Perform feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fusion feature; Dividing the fused features into a training set, a validation set, and a test set, evaluating the performance of the fused data on the validation set, and obtaining an evaluation result; Adjust the fusion strategy and feature selection according to the evaluation results to optimize the fusion effect.

6. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 5, characterized in that: The time domain characteristics include mean, variance, peak, slope and trend; The frequency domain features include: power spectrum density and Fourier transform results; the time-frequency domain features include: short-term and long-term change results.

7. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 5, characterized in that: Performing feature fusion according to the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fusion feature includes: The time domain feature, the frequency domain feature and the time-frequency domain feature are respectively vectorized to obtain a first feature vector, a second feature vector and a third feature vector; the expression of the first feature vector A is A=(a1, a2, ..., a m );The expression of the second eigenvector B is B=b1,b2,…,b n ); the expression of the third eigenvector C is C=(c1,c2,…,c k ), where a m is the mth PCR detection feature, b n is the nth clinical data feature, c k is the kth environmental data feature; The first feature vector, the second feature vector and the third feature vector are sequentially concatenated into a comprehensive long vector to obtain a representation vector F of the fused feature; the expression of the representation vector F of the fused feature is: F=(A, B, C).

8. The method for realizing automatic smoke detection using powdered smoke-generating body and heating wire according to claim 5, characterized in that: Determining whether the test smoke is detected according to the smoke characteristics includes: Constructing an initial detection network and obtaining a sample smoke feature data set; the network structure of the initial detection network is a CNN-LSTM model; Training the initial detection network according to the sample smoke feature data set to obtain a trained smoke detection model; The smoke characteristics are input into the smoke detection model to obtain a detection result.

Images