Smoke alarm water vapor anti-false alarm system based on dual-transmitter and dual-band

Abstract

The present invention belongs to the technical field of smoke alarm false alarm prevention. Specifically, it discloses a smoke alarm water vapor false alarm prevention system based on dual-transmitter and dual-band, which includes: a signal transmission module, a signal reception and analysis module, a propagation environment correction module, an optical path pollution compensation module, and a judgment and early warning module. The present invention innovatively establishes dual-band detection technology and an intelligent environmental adaptation mechanism. It enhances anti-interference capability by adopting dual-light source collaborative detection and real-time calibration of environmental parameters, and combines a light pollution self-compensation mechanism to reduce the impact of environmental factors on detection performance. At the same time, it achieves accurate identification by analyzing the differences in the physical properties of smoke and water vapor. By establishing a dynamic hierarchical early warning model, it can significantly reduce the risk of environmental misjudgment while quickly responding to real fire conditions, thereby systematically solving key technical problems such as smoke detection being susceptible to interference and water vapor confusion in complex scenarios.

Description

Technical Field

[0001] The present invention belongs to the technical field of smoke alarm false alarm prevention, and in particular relates to a smoke alarm water vapor false alarm prevention system based on dual-transmission and dual-band technology. Background Art

[0002] Smoke alarms play a critical role in our daily lives as important fire warning devices. However, traditional smoke alarms are often affected by moisture, leading to frequent false alarms and causing significant disruptions. This can occur in kitchens, where high levels of moisture are generated, or in bathrooms, where high humidity levels can cause unnecessary alarms.

[0003] Existing technologies, such as the Chinese invention patent application with application number 201910813667.6, disclose a high-sensitivity smoke alarm, system and method, which solves the problem of false smoke alarms caused by temperature changes and dust interference by setting up signal comparison technology between the smoke detection chamber and the reference chamber, combined with a layered detection mechanism of multiple layers of smoke detection chambers.

[0004] Another example of existing technology is the anti-false alarm security cloud operation service management system disclosed in the Chinese invention patent application with application number 202210148127.2. It solves the problem of false alarms of the security system caused by false triggering of a single sensor or environmental interference through multi-sensor redundant design, such as the collaborative alarm mechanism of dual smoke alarms and cloud management platforms.

[0005] The first technical solution relies on physically isolating the reference chamber and smoke detection chamber to prevent false alarms due to water vapor. This eliminates temperature interference through signal differences, but does not directly design compensation logic for the optical properties of water vapor. The second technical solution uses multi-sensor redundancy to reduce the probability of false alarms, but does not conduct in-depth analysis of the differences in the signal characteristics of water vapor and smoke, relying solely on repeated detection at the hardware level. Clearly, the following problems remain when considering the two technical solutions mentioned above: 1. Existing technologies do not monitor environmental parameters such as temperature, humidity, and air pressure in real time, making it impossible to dynamically adjust signal characteristics. This results in insufficient ability to distinguish between water vapor and smoke in complex environments.

[0006] 2. Existing technologies do not have a compensation mechanism designed for optical path pollution, and the false alarm rate may increase significantly after long-term use.

[0007] 3. Existing technologies mainly rely on signal strength or redundant detection, with a relatively single dimension, making it difficult to accurately distinguish the subtle differences between water vapor and smoke.

[0008] 4. The existing technology does not perform timing and path processing on the transmission band, which may lead to other potential influencing factors in subsequent analysis, resulting in weak signal anti-interference and resolution capabilities.

[0009] In summary, existing technologies for preventing false alarms of water vapor still have limitations such as poor environmental adaptability, sensitivity to optical contamination, and rough feature analysis. Summary of the Invention

[0010] In view of this, a dual-emission dual-band smoke alarm water vapor false alarm prevention system is proposed to solve the limitations of existing technologies in water vapor false alarm prevention, such as poor environmental adaptability, sensitivity to optical pollution, and rough feature analysis.

[0011] The objectives of the present invention can be achieved through the following technical solutions: The present invention provides a smoke alarm water vapor anti-false alarm system based on dual-transmission and dual-band, the system comprising: a signal transmission module, which transmits optical signals through two independent transmitters, synchronizes the transmission phases of the two optical signals through a timing control circuit, and then merges the two signals into the same path for propagation through an optical combiner.

[0012] The signal reception and analysis module receives optical signals through two independent photoelectric sensors, extracts the multi-dimensional characteristics of the received signal's scattering intensity ratio, time domain waveform, and frequency domain energy distribution, and establishes a mapping relationship model between signal characteristics and target object characteristics.

[0013] The propagation environment correction module monitors the temperature, humidity, air pressure and electromagnetic noise intensity in the signal propagation path in real time, and dynamically corrects the multi-dimensional characteristics of the received signal based on the monitoring information to generate the first multi-dimensional characteristics.

[0014] The optical path pollution compensation module evaluates the degree of optical pollution according to the optical signal attenuation rate, and dynamically adjusts the first multidimensional feature of the received signal to generate a second multidimensional feature.

[0015] The judgment and warning module outputs the target object characteristics through the mapping relationship model based on the second multidimensional characteristics. If the target object is water vapor, no warning is triggered. If the target object is smoke, a graded warning is executed based on the number of times the threshold exceeds the limit within a preset time.

[0016] Compared with the existing technology, the beneficial effects of the present invention are as follows: (1) The present invention enhances the signal anti-interference capability by adopting dual-band collaborative transmission and synchronous control technology, and can maintain high-precision signal analysis even in strong electromagnetic interference or complex reflection environments, which is significantly better than the environmental adaptability of traditional single-band detection systems.

[0017] (2) The present invention combines multi-dimensional analysis of scattering intensity, time domain waveform and frequency domain energy to accurately capture the differences in the physical properties of smoke and water vapor, significantly improve the ability to distinguish fine particles, and effectively distinguish the scattering characteristics of different forms of matter.

[0018] (3) The present invention significantly improves the accuracy of signal feature analysis by real-time monitoring of environmental parameters and performing dynamic multi-dimensional feature correction. This effectively eliminates the impact of environmental interference on detection results, significantly reduces the risk of misidentifying water vapor as smoke, and ensures stable detection performance under complex climatic conditions.

[0019] (4) The present invention can automatically compensate for signal distortion caused by mirror contamination by evaluating the degree of optical contamination and performing secondary dynamic correction of multi-dimensional features. Even after long-term use, it can still maintain stable detection accuracy, solving the problem that there is currently no compensation mechanism designed for optical path contamination, and significantly reducing the problem of false alarms caused by dust accumulation in the equipment.

[0020] (5) By combining the dual verification mechanism of target type identification and duration, the present invention can not only quickly respond to real fire situations but also effectively filter out transient interference. At the same time, through a graded early warning strategy, the false alarm rate is significantly reduced while ensuring sensitivity, meeting high-standard fire safety response requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative work.

[0022] Figure 1 This is a schematic diagram of the system module structure connection of the present invention.

[0023] Figure 2 It is a schematic flow chart of the overall implementation steps of the present invention.

[0024] Figure 3 This is a flowchart of a specific implementation of the signal transmission module of the present invention.

[0025] Figure 4 This is a flowchart of the specific implementation of the signal receiving and analyzing module of the present invention.

[0026] Figure 5 This is a flowchart of the specific implementation of the propagation environment correction module of the present invention.

[0027] Figure 6 This is a flowchart of the specific implementation of the light path pollution compensation module of the present invention.

[0028] Figure 7 This is a specific implementation flow chart of the judgment and early warning module of the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0030] See also Figure 1 and Figure 2 As shown, the present invention provides a smoke alarm water vapor false alarm prevention system based on dual-transmission and dual-band, which includes: a signal transmission module, a signal reception and analysis module, a propagation environment correction module, an optical path pollution compensation module and a judgment and warning module.

[0031] In the above, the signal receiving and analyzing module is connected to the signal transmitting module, the propagation environment correction module and the judgment and warning module respectively, and the optical path pollution compensation module is connected to the propagation environment correction module and the judgment and warning module respectively.

[0032] See also Figure 3 As shown, the signal transmission module transmits optical signals through two independent transmitters, synchronizes the transmission phases of the two optical signals through a timing control circuit, and then combines the two signals into the same path for propagation through an optical combiner.

[0033] It can be understood that the two independent emitters are an infrared emitter and a visible light emitter, the infrared emitter emits near-infrared band such as 850nm, and the visible light emitter emits visible light band such as 450nm.

[0034] Specifically, the synchronization of the transmission phase is performed as follows: 1) the phases of the two transmission signals are collected in real time and the phase difference between the two transmission signals is calculated.

[0035] 2) If the phase difference exceeds a preset threshold range, the phase difference is converted into a voltage or digital signal through an error signal generator and recorded as an error signal.

[0036] 3) Adjust the signal phase according to the error signal until the phase difference is within the preset threshold range.

[0037] It's important to note that the error signal generated by the circuit is input into the error amplifier for amplification. The amplified error signal is then transmitted to the integrator. The integrator integrates the error signal, accumulating its changing trend. The integrator's output signal is used to control the voltage-controlled oscillator (VCO). When the error signal is positive, the VCO's oscillation frequency increases, while when it is negative, the VCO's oscillation frequency decreases. The VCO's output signal is connected to a phase shifter. Based on the frequency changes of the VCO's output signal, the phase shifter shifts the phase of the optical signal that requires phase adjustment. This ensures that the phases of the two optical signals remain precisely synchronized, ensuring the stable operation of the entire smoke alarm water vapor false alarm prevention system.

[0038] In another specific example, the process of combining two signals into the same path for propagation through an optical combiner is as follows: the two independent transmitters are respectively recorded as transmitter 1 and transmitter 2, and transmitter 1 and transmitter 2 respectively emit signals with orthogonal polarization states, such as horizontal polarization light. and vertically polarized light After passing through the polarization beam splitter prism, the horizontally polarized light is directly transmitted and the vertically polarized light is reflected. The two merge into the same beam at the exit of the prism, and the polarization states remain orthogonal and the propagation directions are consistent.

[0039] By integrating the dual-band signals into a single path, they experience identical environmental interference and optical path contamination, providing the subsequent signal analysis module with comparable data. The receiver can accurately distinguish between smoke and water vapor by comparing the scattering intensity ratio and time / frequency domain characteristics of the dual-band signals, mitigating the impact of environmental noise on discrimination at the hardware level. Furthermore, the single path design after beam combining simplifies the system's optical path layout, improving device miniaturization and reliability, making it suitable for volume-sensitive scenarios.

[0040] The embodiment of the present invention enhances the signal anti-interference capability by adopting dual-band collaborative transmission and synchronous control technology, and can maintain high-precision signal analysis even in strong electromagnetic interference or complex reflection environments, which is significantly better than the environmental adaptability of traditional single-band detection systems.

[0041] See also Figure 4 As shown, the signal receiving and analyzing module receives optical signals through two independent photoelectric sensors, and then extracts the multi-dimensional characteristics of the scattering intensity ratio, time domain waveform and frequency domain energy distribution of the received signal, and establishes a mapping relationship model between the signal characteristics and the target object characteristics.

[0042] By combining multi-dimensional analysis of scattering intensity, time domain waveform, and frequency domain energy, the embodiments of the present invention can accurately capture the differences in the physical properties of smoke and water vapor, significantly improve the ability to distinguish fine particles, and effectively distinguish the scattering characteristics of different forms of matter.

[0043] Specifically, the specific process of extracting multidimensional features includes: converting two optical signals into digital signals, calculating the time domain integral or frequency domain energy of the two-band signals to generate a scattering intensity ratio.

[0044] The received signal is decomposed into multiple time-frequency sub-bands after wavelet transformation, and the waveform rise time, fall time and pulse width of each sub-band are extracted as time domain waveform features.

[0045] The received signal is Fourier transformed to generate a spectrum diagram, the energy proportion of the preset frequency band is calculated, and the energy proportion, spectrum peak and half-height width are used as frequency domain energy distribution characteristics.

[0046] It should be added that the specific calculation process of the scattering intensity ratio is as follows: A1. Time domain integration: For each band of signals, integrate them within a preset time window, such as 10ms, and calculate their total energy: , For the The voltage signal at a time point, represents the total energy, Indicates time, and Respectively represent the start time and end time of the preset time window.

[0047] A2. Frequency domain energy: Use FFT to extract the amplitude of the target frequency band, such as avoiding the power supply noise band, and calculate the effective energy , , and Respectively represent the minimum frequency value and maximum frequency value corresponding to the target frequency band, Indicates voltage signal.

[0048] Based on the above calculation rules, the total energy and effective energy of the two bands are calculated respectively. When the time domain integration method is used to calculate the scattering intensity, the scattering intensity ratio is the ratio of the total energy of the two signals. When the frequency domain energy method is used to calculate the scattering intensity, the scattering intensity ratio is the ratio of the effective energy of the two signals.

[0049] Exemplarily, the preset frequency band may be 1-10 kHz or 10-100 kHz.

[0050] In another specific embodiment, the specific process of establishing the mapping relationship model is as follows: A1. Build a sealed experimental chamber and configure it with a temperature and humidity sensor, an air pressure sensor, an electromagnetic interference shielding device, a gas generator and a humidifier.

[0051] A2. Set up several smoke concentration experimental groups and several humidity experimental groups.

[0052] Among them, the smoke concentration can be set to four groups of experiments: 0, 50%, 100% and 200% of the trigger concentration; the humidity experiment can be set to four groups of experiments: 20%RH, 50%RH, 80%RH and 100%RH.

[0053] A3. In the smoke concentration experiment, the time domain waveform, scattered light intensity, and voltage signals of the two photoelectric sensors were recorded at a set sampling frequency. After repeated experiments, the time domain features, frequency domain features, and scattering intensity ratio of each experiment were extracted. The smoke concentration was used as the target object label, and each feature was bound to the smoke concentration.

[0054] A4. In the humidity experiment, water vapor is used as the target object label. The humidity experiment is carried out in the same way as the smoke concentration experiment. The time domain characteristics, frequency domain characteristics, and scattering intensity ratio in each experiment are bound to water vapor.

[0055] Furthermore, the target object label binding also includes the following steps: B1. Calculating the importance of each feature through a random forest algorithm, eliminating feature values whose importance is lower than a preset threshold, and constructing a primary screening data set for each feature.

[0056] B2. Calculate the correlation between each feature and the target object label, and remove the feature values with correlation below the preset threshold from the first screening data set to obtain a second screening data set.

[0057] B3. The minimum and maximum values in the secondary screening data set are combined into a value interval, and the value interval of each feature is bound to the corresponding target object label.

[0058] It is understood that the importance of each feature is calculated using the random forest algorithm by calculating the Gini impurity reduction value, which is used as the importance. For example, the tree size of the random forest is set to 500, the maximum depth is dynamically adjusted according to the number of features, and the node splitting criterion is Gini impurity. The formula for calculating the Gini impurity reduction value is the existing calculation formula and is not further illustrated in the example.

[0059] It should be added that compared with other algorithms, random forest quantifies the importance of features through the Gini impurity reduction value of multiple decision trees. It can not only effectively capture the nonlinear correlation and interaction between complex features such as scattering intensity ratio, time domain waveform and frequency domain energy, but also has strong robustness to high-dimensional noise, avoiding overfitting interference.

[0060] In step B2, the correlation between each feature and the target object label can be calculated using methods such as the Spearman rank correlation coefficient or the mutual information method. In one specific embodiment, the mutual information method is preferred. Mutual information quantifies the statistical dependence between features and target variables and can capture arbitrary patterns. In smoke detection scenarios, features such as scattering intensity ratios and time domain waveforms often exhibit complex nonlinear relationships with target labels, such as exponential decay and threshold mutations. The mutual information method provides a more comprehensive assessment of feature discriminability.

[0061] Furthermore, the establishment of the mapping relationship model also includes embedding an interference mapping relationship model, and its specific construction process is as follows: C1. Oil mist particles and dust particles are used as interference types, and each interference experimental group is set according to the interference concentration gradient and particle size segment spacing under fixed smoke concentration and fixed humidity.

[0062] For example, the particle size of oil mist particles can be set to 0.5-2 μm, the concentration can be set to 0.5 mg / m³ and 1.0 mg / m³, the particle size of dust particles can be set to 1-5 μm, the concentration can be set to 0.5 mg / m³ and 1.0 mg / m³, and each interference experimental group can be set accordingly.

[0063] C2. Record the misjudgment rates of the smoke concentration experimental group and the humidity experimental group with and without interference. The misjudgment rate is the ratio of the number of misjudgments to the total number of experiments.

[0064] C3. After repeating the experiment several times for each interference experimental group, calculate the average of the error rate and use it as the actual error rate of the corresponding interference experimental group.

[0065] C4. Calculate the mean error rates of the smoke concentration experimental group and the humidity experimental group after repeating the experiment several times without interference, and obtain the average error rate of the two experimental groups, which will be used as the baseline error rate.

[0066] C5. Comprehensive analysis of the baseline false positive rate and the actual false positive rate yields the anti-interference scores of the smoke concentration experimental group and the humidity experimental group under various interference types.

[0067] Understandably, to verify the system's anti-interference performance in complex environments, it's necessary to simulate the multiple sources of interference that may exist in real-world scenarios. By controlling the concentration gradient of the interferents and the particle size segmentation variables, we can systematically evaluate the cross-influence of different interference intensities on the detection model, establish a quantitative relationship between interference characteristics and false positive rates, and provide data support for subsequent compensation algorithms.

[0068] It should be noted that oil mist and dust particles exhibit typical aerosol properties and are most frequently found in industrial and natural environments. Their optical properties, such as refractive index and hygroscopicity, differ significantly from those of smoke, effectively validating the system's ability to distinguish heterogeneous suspended matter while also covering the modeling requirements of typical oily and inorganic interference scenarios.

[0069] Among them, the anti-interference score can be analyzed and calculated by the following formula: , represents the anti-interference score, represents the actual misjudgment rate, represents the baseline misjudgment rate, is a natural constant, when When , this indicates that the actual misjudgment rate meets or exceeds the benchmark anti-interference capability standard. At this time, the anti-interference score can take the maximum value. In order to facilitate the subsequent quantitative comparison, the value is assigned to 1. When When , it indicates that the interference is large and the anti-interference ability is not up to standard. The anti-interference score decreases as the difference between the actual false positive rate and the benchmark false positive rate increases. In order to accurately and intuitively reflect this trend, an exponential function is used to express it. hour, The value range of is (0,1), Equivalent to , along with As the error rate increases, it gradually approaches 0. That is, when the actual error rate exceeds the benchmark, the more it exceeds, the lower the anti-interference score. The anti-interference ability weakens as the error rate exceeds the benchmark.

[0070] Furthermore, the embedded interference mapping relationship model is used to adjust the target object tag trigger threshold, and the specific adjustment process includes: D1. If the anti-interference scores of oil mist particles and dust are both higher than or equal to the corresponding set thresholds, no trigger threshold adjustment is performed.

[0071] D2. If the anti-interference score of oil mist particles or dust is lower than the corresponding set threshold, the absolute value of the difference between the anti-interference score and the corresponding set threshold is taken as the corresponding amplification ratio.

[0072] D3. If both are lower than the set threshold, a compensation amplification ratio is set, and the final amplification ratio is obtained by combining the amplification ratios of oil mist particles and dust and the compensation amplification ratio.

[0073] D4. Calculate the adjusted trigger threshold range based on the amplification ratio and the preset benchmark trigger threshold range, and adjust the value interval of each feature based on this range.

[0074] In one specific embodiment, the compensatory amplification ratio is preset based on actual scenarios, such as empirical or experimental values. The final amplification ratio can be calculated by combining the corresponding amplification ratios for oil mist particles, dust particles, and the compensatory amplification ratio. Weights for the corresponding amplification ratios for oil mist particles, dust particles, and the compensatory amplification ratio can be set, and the final amplification ratio can be obtained by weighted summation. For example, the weights for the corresponding amplification ratios for oil mist particles, dust particles, and the compensatory amplification ratio can be 0.2, 0.5, and 0.3, respectively.

[0075] In a preferred embodiment of the present invention, if the anti-interference score falls below a set threshold, it indicates that the system has a high false positive rate in a noisy environment, and the smoke concentration detection trigger threshold needs to be increased. For example, the baseline trigger threshold range is ±10% of the original value range. When the amplification ratio is 50%, the range is adjusted from ±10% of the original value range to ±15% of the original value range to reduce the risk of false alarms. To prevent extreme adjustments, the maximum threshold range is set to ±20%.

[0076] In one specific implementation, in complex industrial environments, the optical properties of interfering objects such as fog and dust can easily interfere with the target smoke, leading to increased misjudgment rates in traditional detection models. By embedding an interference mapping relationship model to adjust the target label trigger threshold, the impact of different interfering object concentrations and particle sizes on detection accuracy is quantified, thereby establishing a dynamic threshold adjustment mechanism to address the lack of adaptability of fixed thresholds in changing scenarios. This facilitates effective differentiation between real smoke and typical interfering particles, reduces false alarms or missed detections caused by environmental interference, and achieves precise decoupling of environmental interference from the target signal.

[0077] See also Figure 5 As shown, the propagation environment correction module monitors the temperature, humidity, air pressure and electromagnetic noise intensity in the signal propagation path in real time, and dynamically corrects the multi-dimensional features of the received signal to generate the first multi-dimensional features.

[0078] Specifically, the temperature is monitored in real time by a digital temperature sensor, the humidity is monitored in real time by a capacitive humidity sensor, the air pressure is obtained by an integrated air pressure sensor to obtain real-time air pressure values, and the electromagnetic noise intensity is monitored in real time by a broadband radio frequency probe.

[0079] Understandably, temperature changes affect the movement of air molecules and particles, altering scattering characteristics, signal propagation speed, and attenuation, which in turn affects the scattering intensity ratio, time-domain waveform, and frequency-domain energy distribution. Increased humidity increases water vapor content, whose scattering and absorption effects change the scattering intensity ratio, affecting signal propagation and causing distortion in the time-domain waveform. Air pressure changes air density, affecting signal scattering, propagation speed, and attenuation, impacting the scattering intensity ratio and time-domain waveform. Electromagnetic noise intensity can cause signal intensity fluctuations, superimposing noise on the time-domain waveform, causing distortion, and introducing additional energy components in the frequency domain, altering the frequency-domain energy distribution. The multi-dimensional characteristics of the received signal are dynamically corrected by integrating the temperature, humidity, air pressure, and electromagnetic noise intensity along the signal propagation path.

[0080] Specifically, the multi-dimensional features of the dynamically corrected received signal include: R1, a dynamically corrected scattering intensity ratio calculated according to the temperature and humidity through a preset temperature-humidity joint compensation formula.

[0081] R2. Calculate the dynamically corrected waveform rise time using the air pressure-electromagnetic interference joint correction formula based on the air pressure and electromagnetic interference intensity. Correct the waveform fall time and pulse width in the same way to obtain the dynamically corrected time domain waveform characteristics.

[0082] R3. Adjust the energy ratio according to the temperature through the frequency domain energy compensation model, and use the frequency drift compensation formula to dynamically compensate the spectrum peak and half-height width. Combine these three to obtain the dynamically corrected frequency domain energy distribution characteristics.

[0083] R4. Generate the first multidimensional feature by integrating the dynamically corrected scattering intensity ratio, time domain waveform characteristics and frequency domain energy distribution characteristics.

[0084] The temperature-humidity joint compensation formula in step R1 is specifically expressed as follows: , and represent the scattering intensity ratio before and after dynamic correction, is the set temperature compensation coefficient, the value is 0.002 / ℃, The humidity compensation coefficient is set to -0.005 / %RH. To set the reference calibration temperature, take the value as 25℃. To set the reference humidity, take 50%RH. and Respectively represent the temperature and humidity in the signal propagation path.

[0085] in, is the temperature influence term, if , this item is the correct one Increase. , this item is negative Decreases, reflecting the compensation relationship of the increase in scattering intensity ratio as temperature rises. is the humidity influence term, if , this item is negative Decrease. , this item is the correct one increases, reflecting the compensatory relationship of the decrease in scattering intensity as humidity increases.

[0086] It's important to explain that increased humidity can cause water vapor to condense on the surface of dust or oil mist particles, changing their size, density, and other properties. Larger particles have different scattering properties than the original particles, resulting in a decrease in the overall scattering intensity ratio. Furthermore, water vapor itself absorbs or scatters light, further reducing the scattering intensity ratio.

[0087] The specific expression of the air pressure-electromagnetic interference joint correction formula in step R2 is: , and Respectively represent the waveform rise time before and after dynamic correction, is standard atmospheric pressure, and Respectively represent the air pressure and electromagnetic noise intensity in the signal propagation path, To set the air pressure influence factor, the value is 0.001 / kPa, is the set electromagnetic interference suppression factor, which is 0.01 / dBμV / m. The maximum permissible electromagnetic noise intensity is set.

[0088] in, is the air pressure influence term, Reflects the relative change of air pressure. , this item is negative, After weighting, reduce the correction factor to make , otherwise the item is a positive increase correction factor , represents the influence term of electromagnetic noise intensity, It means first judging whether the actual electromagnetic noise intensity exceeds the maximum allowable value. If it does not exceed the value, it means that the impact on the waveform time is small and can be ignored. If it exceeds the value, the excess part will be calculated, and the more the excess part is, the more it will be affected. After weighting, the correction factor is increased so that .

[0089] It should be added that as the air pressure increases, the air density increases, the probability of light scattering increases, and the waveform rise time tends to shorten. As the air pressure decreases, the air density decreases, the scattering weakens, the light intensity rises slowly, and the waveform rise time tends to lengthen. The high intensity of electromagnetic noise will mask part of the effective signal, causing the detector to delay the response of the light signal, thereby lengthening the waveform rise time.

[0090] The specific calculation formula of the frequency domain energy compensation model in step R3 is: , and are the energy proportions before and after dynamic correction, is the set temperature attenuation factor, which can be -0.01 / °C. The frequency drift compensation formula is: , and Represent the spectrum peaks before and after adjustment, It represents the temperature drift coefficient, and the specific value can be 0.005 / ℃.

[0091] It should be noted that temperature affects the signal's energy contribution, spectral peak, and half-height width. In terms of energy contribution, high temperatures may increase signal attenuation in certain frequency bands, leading to a decrease in the energy contribution of these frequency bands to the overall signal. As for spectral peaks, temperature-induced frequency drift can cause them to change. For example, the frequency of an oscillator will vary with temperature fluctuations. Rising temperature causes the frequency peak to shift toward higher frequencies, increasing the frequency peak, while shifting toward lower frequencies decreases the frequency peak. The half-height width is also affected by temperature-induced changes in the Q value: as temperature increases, the Q value decreases, and the half-height width increases accordingly.

[0092] The embodiments of the present invention significantly improve the accuracy of signal feature analysis by real-time monitoring of environmental parameters and performing dynamic multi-dimensional feature correction. This effectively eliminates the impact of environmental interference on detection results, significantly reduces the risk of misidentifying water vapor as smoke, and ensures stable detection performance under complex climatic conditions.

[0093] See also Figure 6 As shown, the optical path pollution compensation module evaluates the optical pollution degree according to the optical signal attenuation rate, and dynamically adjusts the first multidimensional feature of the received signal to generate a second multidimensional feature.

[0094] Specifically, the optical pollution degree assessment process includes: U1, transmitting a dual-band optical signal with a known intensity, taking the intensity as the initial intensity, and recording the received signal intensities of the two bands.

[0095] U2. The optical attenuation rate is calculated by combining the initial intensity, the received signal strength, and the layout distance between the transmitter and the photoelectric sensor through the attenuation rate formula.

[0096] Among them, the optical attenuation rate can be calculated based on the Beer-Lambert law, which is an existing calculation formula and will not be shown again.

[0097] U3. If the optical attenuation rate of a certain band is less than the corresponding preset threshold, the optical pollution factor of the band is assigned a value of 0. Otherwise, the optical pollution factor is obtained by combining the optical attenuation rate and the corresponding preset threshold.

[0098] Understandably, the preset threshold is a limit value determined based on extensive experimental data, industry standards, or environmental factors, used to measure the degree of optical contamination. When the optical attenuation rate in a particular wavelength band is less than the corresponding preset threshold, it means that the attenuation of light in that wavelength band is within a relatively normal and acceptable range, without significant contamination or interference. Therefore, it can be reasonably assumed that there is no optical contamination issue in that wavelength band, and the optical contamination factor is assigned a value of 0.

[0099] In a preferred embodiment of the present invention, the optical pollution factor is obtained by combining the optical attenuation rate and the corresponding preset threshold. The optical attenuation rate difference is calculated by taking the difference between the optical attenuation rate and the corresponding preset threshold, and the optical attenuation rate difference is input as an input variable into the exponential decay function to output the optical pollution factor, such as , Represents the input variable.

[0100] It is understandable that the optical pollution factor can also be calculated through normalization or various calculation methods such as the Sigmoid function. In the optical pollution scenario, as the difference in optical attenuation rate increases, the degree of optical pollution often does not increase linearly, but changes rapidly in the early stage and tends to be stable in the later stage. Therefore, the present invention prefers an exponential decay function as the processing function.

[0101] U4. Calculate the optical pollution degree by weighted summation based on the optical pollution factors of the dual bands and the preset weights.

[0102] Furthermore, the specific adjustment process of dynamically adjusting the first multi-dimensional feature of the received signal is as follows: matching the optical pollution degree with the optical pollution degree interval corresponding to each optical pollution level to obtain the corresponding optical pollution level.

[0103] A pollution compensation algorithm that matches the current optical pollution level is extracted from a preset pollution compensation algorithm configuration table, and the second multidimensional feature is obtained after adjusting the first multidimensional feature according to the algorithm.

[0104] It should be added that the pollution compensation algorithm configuration table is specifically shown in Table 1.

[0105] Table 1 Pollution compensation algorithm configuration table

[0106]

[0107] In the above table, y, q, and g represent the scattering intensity ratio correction coefficient, the time domain delay correction coefficient, and the frequency domain energy attenuation coefficient, respectively. When scattering correction is used, only the scattering intensity ratio is corrected. When scattering + time domain correction is used, the scattering intensity ratio and time domain waveform characteristics are corrected. When scattering + time domain + frequency domain combined correction is used, the scattering intensity ratio, time domain waveform characteristics, and frequency domain energy distribution characteristics are corrected. The correction formula for scattering correction is: , It represents the scattering correction ratio after optical path pollution compensation. The time domain correction formula is demonstrated using the waveform rise time as an example. The specific correction formula is: , It represents the waveform rise time after optical path pollution compensation. The frequency domain correction formula is demonstrated using energy ratio as an example. The specific correction formula is: , Indicates the energy proportion after adjustment for optical path pollution compensation.

[0108] The embodiment of the present invention can automatically compensate for signal distortion caused by mirror contamination by evaluating the degree of optical contamination and performing secondary dynamic correction of multi-dimensional features. Even after long-term use, it can still maintain stable detection accuracy, solve the problem that there is currently no compensation mechanism designed for optical path contamination, and significantly reduce the problem of false alarms caused by dust accumulation in the equipment.

[0109] See also Figure 7 As shown, the judgment and warning module outputs the target object characteristics through a mapping relationship model based on the second multidimensional characteristics. If the target object is water vapor, no warning is triggered. If the target object is smoke, a graded warning is executed based on the number of times the threshold exceeds the limit within a preset time.

[0110] It should be added that the preset time of the early warning module is determined as follows: the duration of the smoke triggering alarm is matched from the factory settings of the smoke alarm according to the usage scenario of the smoke alarm as the preset time.

[0111] In a preferred embodiment of the present invention, assuming that the smoke alarm is used in a factory, the smoke alarm triggering time is set to 1 minute. When the threshold value is exceeded 3 times within 1 minute, the primary warning is triggered to start the sound and light alarm. When the threshold value is exceeded 5 times within 1 minute, the intermediate warning is triggered and the ventilation system is started. When the threshold value is exceeded 10 times within 1 minute, the advanced warning is triggered and the external fire protection system is linked.

[0112] By combining target type identification and duration verification, this embodiment of the present invention can rapidly respond to real fires while effectively filtering out transient interference. Furthermore, through a graded early warning strategy, it significantly reduces false alarm rates while maintaining sensitivity, meeting high-standard fire safety response requirements.

[0113] It should be noted that the above formula is a formula obtained by collecting a large amount of data and performing software simulation to obtain the latest real situation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0114] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0115] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0116] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0117] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0118] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The smoke alarm water vapor anti-false alarm system based on dual-transmitter and dual-band is characterized by: The system includes: The signal transmission module transmits optical signals through two independent transmitters. After the transmission phases of the two optical signals are synchronized by the timing control circuit, the two signals are combined into the same path through the optical combiner for propagation. The signal receiving and analysis module receives optical signals through two independent photoelectric sensors, extracts the multi-dimensional characteristics of the received signal, including the scattering intensity ratio, time domain waveform, and frequency domain energy distribution, and establishes a mapping relationship model between the signal characteristics and the target object characteristics; The propagation environment correction module monitors the temperature, humidity, air pressure and electromagnetic noise intensity in the signal propagation path in real time, and dynamically corrects the multi-dimensional characteristics of the received signal to generate the first multi-dimensional characteristics based on these characteristics; An optical path pollution compensation module evaluates the degree of optical pollution based on the optical signal attenuation rate and dynamically adjusts the first multidimensional feature of the received signal to generate a second multidimensional feature; The judgment and warning module outputs the target object characteristics through the mapping relationship model based on the second multidimensional characteristics. If the target object is water vapor, no warning is triggered. If the target object is smoke, a graded warning is executed based on the number of times the threshold exceeds the limit within a preset time.

2. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 1, characterized in that: The specific synchronization of the transmission phase is performed as follows: Collect the phase of the two transmitted signals in real time and calculate the phase difference between the two transmitted signals; If the phase difference exceeds a preset threshold range, the phase difference is converted into a voltage or digital signal by an error signal generator and recorded as an error signal; The signal phase is adjusted according to the error signal until the phase difference is within a preset threshold range.

3. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 1, characterized in that: The specific extraction process of the multidimensional features includes: Convert the two optical signals into digital signals and calculate the time domain integral or frequency domain energy of the two-band signals to generate the scattering intensity ratio; The received signal is decomposed into multiple time-frequency sub-bands after wavelet transformation, and the rise time, fall time and pulse width of each sub-band are extracted as time domain waveform features; The received signal is Fourier transformed to generate a spectrum diagram, the energy proportion of the preset frequency band is calculated, and the energy proportion, spectrum peak and half-height width are used as frequency domain energy distribution characteristics.

4. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 1, characterized in that: The specific process of establishing the mapping relationship model is as follows: Build a sealed experimental chamber and configure it with temperature and humidity sensors, air pressure sensors, electromagnetic interference shielding devices, gas generators and humidifiers; Set up several smoke concentration experimental groups and several humidity experimental groups; In the smoke concentration experiment, the time domain waveform, scattered light intensity, and voltage signals of the two photoelectric sensors were recorded at a set sampling frequency. After repeated experiments, the time domain features, frequency domain features, and scattered intensity ratio of each experiment were extracted. The smoke concentration was used as the target object label, and each feature was bound to the smoke concentration. In the humidity experiment, water vapor is used as the target object label. The humidity experiment is carried out in the same way as the smoke concentration experiment. The time domain characteristics, frequency domain characteristics and scattering intensity ratio under each experiment are bound to water vapor.

5. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 4, characterized in that: The target tag binding also includes the following steps: The importance of each feature is calculated using the random forest algorithm, and feature values with importance lower than the preset threshold are eliminated to construct a primary screening dataset for each feature; Calculate the correlation between each feature and the target object label, and remove the feature values with correlation below the preset threshold from the first screening data set to obtain the second screening data set; The minimum and maximum values in the secondary screening data set are combined into a value interval, and the value interval of each feature is bound to the corresponding target object label.

6. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 4, characterized in that: The establishment of the mapping relationship model also includes embedding the interference mapping relationship model, and the specific construction process is as follows: Oil mist particles and dust particles were used as interference types, and interference experimental groups were set up according to the interference concentration gradient and particle size segment spacing under fixed smoke concentration and fixed humidity. The misjudgment rates of the smoke concentration experimental group and the humidity experimental group were recorded separately when there was no interference and when there was interference. The misjudgment rate was the ratio of the number of misjudgments to the total number of experiments. After repeating the experiment several times for each interference experimental group, the mean of the error rate is calculated and used as the actual error rate of the corresponding interference experimental group; Calculate the mean error rates of the smoke concentration experimental group and the humidity experimental group after repeated experiments without interference, and obtain the average error rates of the two experimental groups as their baseline error rates; The comprehensive analysis of the baseline false positive rate and the actual false positive rate yielded the anti-interference scores of the smoke concentration experimental group and the humidity experimental group under various interference types.

7. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 6, characterized in that: The embedded interference mapping relationship model is used to adjust the target object tag trigger threshold, and the specific adjustment process includes: If the anti-interference scores of oil mist particles and dust are both higher than or equal to the corresponding set thresholds, no trigger threshold adjustment is performed; If the anti-interference score of oil mist particles or dust is lower than the corresponding set threshold, the absolute value of the difference between the anti-interference score and the corresponding set threshold is taken as the corresponding amplification ratio; If both are lower than the set threshold, the compensation amplification ratio is set, and the final amplification ratio is obtained by combining the amplification ratios of oil mist particles and dust with the compensation amplification ratio; The adjusted trigger threshold range is calculated based on the amplification ratio and the preset benchmark trigger threshold range, and the value interval of each feature is adjusted based on this range.

8. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 1, characterized in that: The multi-dimensional features of the dynamic correction received signal include: The dynamically corrected scattering intensity ratio is calculated based on the temperature and humidity using a preset temperature-humidity joint compensation formula; The rise time of the dynamically corrected waveform is calculated based on the air pressure and electromagnetic interference intensity using the air pressure-electromagnetic interference joint correction formula. The waveform fall time and pulse width are corrected in the same way to obtain the dynamically corrected time domain waveform characteristics. The energy ratio is adjusted according to the temperature through the frequency domain energy compensation model, and the spectrum peak and half-height width are dynamically compensated using the frequency drift compensation formula. The dynamic correction of the frequency domain energy distribution characteristics is obtained by combining these three factors. The first multidimensional feature is generated by integrating the dynamically corrected scattering intensity ratio, time domain waveform characteristics and frequency domain energy distribution characteristics.

9. The dual-transmitter, dual-band smoke alarm water vapor false alarm prevention system according to claim 1, characterized in that: The optical contamination level assessment process includes: transmitting a dual-band optical signal of known intensity, taking the intensity as the initial intensity, and recording the received signal intensities of the two bands; The optical attenuation rate is calculated by combining the initial intensity, the received signal intensity and the distance between the transmitter and the photoelectric sensor through the attenuation rate formula; If the optical attenuation rate of a certain band is less than the corresponding preset threshold, the optical pollution factor of the band is assigned to 0; otherwise, the optical pollution factor is calculated by combining the optical attenuation rate and the corresponding preset threshold. The optical pollution degree is calculated based on the dual-band optical pollution factors and preset weights.

10. The smoke alarm water vapor false alarm prevention system based on dual emission and dual band according to claim 1, characterized in that: The specific adjustment process of dynamically adjusting the first multidimensional feature of the received signal is as follows: Match the optical pollution degree with the optical pollution degree interval corresponding to each optical pollution level to obtain the corresponding optical pollution level; A pollution compensation algorithm that matches the current optical pollution level is extracted from a preset pollution compensation algorithm configuration table, and the second multidimensional feature is obtained after adjusting the first multidimensional feature according to the algorithm.

Images