Abnormal state detection method, system and storage medium of point-type explosion-proof smoke sensor

Abstract

The present application relates to the technical field of fire detection and alarm, and discloses a method, system, and storage medium for detecting abnormal conditions of point-type explosion-proof smoke sensors. The method first obtains multiple explosion-proof smoke sensor addresses, then sends polling signals to the corresponding smoke sensor units in sequence, and waits for a response signal. If no response is received, it indicates a line abnormality; if a response is received, the smoke sensor signal and detection image data are obtained at regular intervals to identify the smoke data. When the smoke sensor alarms but there is no smoke data, it indicates a pollution abnormality; when there is no smoke signal but there is smoke data, it indicates a fault abnormality. The method combines polluting signals with image analysis to achieve line, pollution, and fault abnormality detection, thereby improving the reliability and accuracy of explosion-proof smoke sensors in complex environments.

Description

Technical Field

[0001] The present application relates to the technical field of fire detection and alarm, and in particular to a method, system and storage medium for detecting abnormal conditions of a point-type explosion-proof smoke sensor. Background Art

[0002] Point-type explosion-proof smoke detectors are fire detection devices used in flammable and explosive environments. They detect changes in smoke concentration and issue warnings at the early stages of a fire. They are widely used in industrial production scenarios such as petrochemicals, coal mining, and natural gas extraction. Their operating principle is typically based on photoelectric or ion sensing technology. When smoke particles enter the sensing area, they trigger an alarm by changing the optical path or ionization balance.

[0003] In industrial production scenarios, point-type explosion-proof smoke detectors face complex and harsh working environments. Oil refining plants are subject to high temperatures, high humidity, and corrosive gases, while underground coal mines are filled with large amounts of dust and gas. These environmental factors not only place extremely high demands on equipment stability but can also easily lead to equipment failure.

[0004] Because the sensing channel must be connected to the outside world to detect smoke, foreign objects such as dust, dirt, and insects can easily enter the device housing through this channel. These foreign objects can accumulate on the sensing element, interfering with the normal optical or ion detection process, distorting the sensor data and causing false or missed smoke alarms. More seriously, long-term accumulation of foreign matter can cause internal circuit shorts and other malfunctions. Summary of the Invention

[0005] In order to detect the abnormal state of a point-type explosion-proof smoke detector in a harsh working environment, the present application provides a method, system and storage medium for detecting the abnormal state of a point-type explosion-proof smoke detector.

[0006] In a first aspect, the present application provides a method for detecting abnormal conditions of a point-type explosion-proof smoke sensor, which adopts the following technical solutions:

[0007] A method for detecting abnormal state of a point-type explosion-proof smoke sensor comprises the following steps:

[0008] Get multiple explosion-proof smoke sensor addresses;

[0009] Sending polling signals to corresponding explosion-proof smoke sensing units in sequence according to the plurality of explosion-proof smoke sensing addresses;

[0010] Waiting for a response signal returned by the current explosion-proof smoke sensor unit in response to the polling signal;

[0011] If the response signal is not received, it is prompted that the circuit of the corresponding explosion-proof smoke sensing unit is abnormal; otherwise, the smoke sensing signal of the explosion-proof smoke sensing unit is obtained at regular intervals, and detection image data is obtained by photographing the explosion-proof smoke sensing unit, and smoke data is identified in the detection image data;

[0012] If the smoke sensor signal is a smoke sensor alarm and the smoke data is not recognized, it indicates that the smoke sensor is contaminated abnormally;

[0013] If the smoke sensor signal is a no-smoke signal and the smoke data is recognized, it indicates that the smoke sensor is faulty.

[0014] By adopting the above technical solution, through polling signal detection of line continuity and cross-verification of smoke sensor signals and image analysis, accurate identification of three abnormal states can be achieved: when no response signal is received, the line abnormality is prompted to solve the problem of open circuit and short circuit caused by dust and moisture; when the smoke sensor alarm is triggered but there is no smoke in the image, it is judged as pollution abnormality to avoid false alarms caused by dust and oil; when there is no smoke signal but there is smoke in the image, the fault abnormality is identified to prevent missed reports due to sensor failure.

[0015] Optionally, the step of identifying smoke data in the detection image data further includes the following sub-steps:

[0016] Extracting hue data and saturation data from the detection image data based on the HSV color space;

[0017] Identifying temporary pixel particles whose hue data is within a preset hue range and whose saturation data is within a preset saturation range;

[0018] Obtaining the pixel position of the temporary pixel particle;

[0019] Calculating the position distance between adjacent temporary pixel particles according to the pixel positions;

[0020] If the position distance is greater than the preset interference distance, the position distance is eliminated, and the average value of the remaining position distances is calculated as the aggregated data;

[0021] If the aggregated data is smaller than the preset target data, the smoke data exists; otherwise, the smoke data does not exist.

[0022] By adopting the above technical solution, when identifying smoke data in the detection image data, hue and saturation data are extracted based on the HSV color space, which can accurately locate the color feature areas unique to smoke. After identifying temporary pixel particles that meet the preset range, the position distances of adjacent particles are calculated and interference distances are eliminated. The average of the remaining distances is used as the aggregation data to determine whether smoke data exists. This technical solution can effectively filter out interference from non-smoke particles such as dust and oil, avoiding misidentification due to environmental pollutants. At the same time, by quantifying the degree of aggregation of pixel particles, it can accurately distinguish between real smoke and scattered interference objects, improving the accuracy of smoke detection, ensuring more reliable judgment of explosion-proof smoke pollution anomalies in complex industrial environments, and further reducing the false alarm rate.

[0023] Optionally, the method further comprises the following steps:

[0024] Acquire a positioning mark based on the abnormal prompt, and identify a positioning unit corresponding to the positioning mark and having a fixed position in the detection image data;

[0025] identifying the explosion-proof smoke sensing unit in the detection image data;

[0026] Calculating a distance change between the positioning unit and the explosion-proof smoke sensing unit;

[0027] If the distance change value is outside the preset change range, an on-site abnormality warning is issued; otherwise, the target data is adjusted in direct proportion to the distance change value.

[0028] By adopting the above technical solution, by obtaining the positioning mark corresponding to the abnormal prompt, identifying the positioning unit and the explosion-proof smoke sensor unit in the detection image, calculating the distance change value between the two and comparing it with the preset range, it is possible to accurately judge whether the smoke sensor has undergone physical displacement (such as loose installation, external force collision causing position offset), and issue an on-site warning when it exceeds the range to avoid detection failure caused by installation abnormalities; at the same time, the target data of smoke detection (such as pixel aggregation threshold) is adjusted in direct proportion to the distance change value, and the imaging perspective difference caused by the position change of the smoke sensor can be dynamically adapted to avoid smoke recognition misjudgment caused by the change in the relative distance between the lens and the smoke sensor. This not only improves the accuracy of abnormal positioning, but also enhances the environmental adaptability of the detection system under complex working conditions through adaptive adjustment, ensuring the reliability and robustness of explosion-proof smoke sensor status monitoring.

[0029] Optionally, the step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps:

[0030] Calculate the number of addresses according to the explosion-proof smoke sensor addresses;

[0031] Matching the content of the response signal according to the number of addresses;

[0032] After sending the polling signal to the current explosion-proof smoke sensor unit, if the response signal returned by the current explosion-proof smoke sensor unit is received within the preset first time period or is not received after the first time period, the polling signal is sent to the next explosion-proof smoke sensor unit.

[0033] By adopting the above technical solution, efficient polling detection of multiple explosion-proof smoke detection units can be achieved, avoiding the blockage of the overall polling process due to communication anomalies of a single smoke detection unit, and improving the system detection efficiency; at the same time, by matching the number of addresses with the content of the response signal, the signal integrity can be verified, and anomalies in data transmission can be discovered in time to ensure the reliability of polling detection. In industrial scenarios, all smoke detection unit states can be quickly traversed, effectively shortening the overall fault detection time.

[0034] Optionally, the step of matching the content volume of the response signal according to the number of addresses further includes the following sub-steps:

[0035] The content volume is adjusted in direct proportion to the number of addresses N; content volume = M × (N / reference volume), where M is a preset reference content volume;

[0036] or,

[0037] The content volume is adjusted in direct proportion to the data bit length of the explosion-proof smoke sensor address; content volume=M×K, where M is the preset reference content volume and K is the growth coefficient.

[0038] By adopting this technical solution, as the number of addresses increases, the response signal needs to carry more status data corresponding to each address (such as smoke concentration and device status code), thereby proportionally increasing the content to ensure data integrity. Conversely, as the number of addresses decreases, the content is correspondingly reduced to avoid data redundancy. Alternatively, this can ensure that smoke detectors with different encoding rules can operate compatibly in the same system.

[0039] Optionally, the method further comprises the following steps:

[0040] The number of the explosion-proof smoke sensing units that have not sent the response signal within the statistical period is counted as the number of non-response;

[0041] Adjust the first duration according to the number of unanswered responses:

[0042] If the number of unanswered calls is less than or equal to the preset threshold, the first duration is adjusted according to the first formula: first duration = initial duration × (1-first adjustment coefficient × number of unanswered calls);

[0043] If the number of unanswered responses is greater than the preset threshold, the first duration is adjusted according to the second formula: first duration = initial duration × (1 + second adjustment coefficient × (number of unanswered responses - preset threshold)).

[0044] By adopting this technical solution, when a small number of smoke sensors experience communication delays, shortening the first duration can speed up polling and reduce the overall inspection cycle. When a large number of smoke sensors experience communication anomalies, extending the first duration can reduce the polling frequency and avoid excessive bus load on the PLC due to frequent signal transmission.

[0045] Optionally, the step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps:

[0046] After sending the polling signal to the current explosion-proof smoke sensor unit, directly sending the polling signal to the next explosion-proof smoke sensor unit;

[0047] waiting for the response signal while sending the polling signal, wherein the potential value of the initial data segment of the response signal is lower than the potential value of the initial data segment of the polling signal;

[0048] If the response signal is received, the polling signal sent to the next explosion-proof smoke sensing unit is suspended, and after a complete response signal is obtained, the polling signal is resent to the next explosion-proof smoke sensing unit.

[0049] By adopting the above technical solution, after sending the polling signal, the polling signal is directly sent to the next explosion-proof smoke detection unit to realize parallel polling, and the difference between the potential value of the response signal and the initial data of the polling signal is used to identify the signal. When the response signal is received, the sending of the next polling signal is paused to obtain a complete response, which effectively improves the polling efficiency, shortens the inspection cycle, enhances the signal anti-interference ability, and reduces the misjudgment rate. At the same time, it optimizes the bus load, adapts to complex industrial environments, can quickly locate faults, improve operation and maintenance efficiency, and simplify hardware design and reduce costs.

[0050] Optionally, the method further comprises the following steps:

[0051] Obtaining the latest multiple smoke sensor signals;

[0052] Identifying the number of jump pulses in the smoke sensor signal;

[0053] Calculating the drift amplitude of the zero point drift in the smoke sensor signal;

[0054] If the number of pulses is greater than a preset warning number and the drift amplitude is greater than a preset warning amplitude, it indicates that the sensor unit is malfunctioning.

[0055] By adopting the above technical solution, the latest smoke sensor signal is obtained, the number of jump pulses therein is identified and the zero-point drift amplitude is calculated. When the number of pulses exceeds the preset warning number and the drift amplitude exceeds the preset warning amplitude, it prompts that the sensor unit is malfunctioning. It can accurately capture signal disorder caused by dust accumulation, oil contamination and other pollution of the sensor unit (such as an abnormal increase in the number of pulses) and zero-point drift caused by component aging and moisture corrosion (such as baseline offset exceeds the normal range), realize early warning of potential faults of the sensor unit, avoid false alarms or missed alarms of smoke sensors due to signal distortion, improve the detection reliability of explosion-proof smoke sensors in complex industrial environments, provide a quantitative basis for equipment maintenance, and ensure the long-term stable operation of the fire detection system.

[0056] In a second aspect, the present application provides a point-type explosion-proof smoke sensor abnormal state detection system, which adopts the following technical solutions:

[0057] A system for detecting an abnormal state of a point-type explosion-proof smoke detector comprises a processor, wherein the processor executes the steps of any one of the above-mentioned methods for detecting an abnormal state of a point-type explosion-proof smoke detector.

[0058] In a third aspect, the present application provides a storage medium that adopts the following technical solution:

[0059] A storage medium stores a program, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for detecting an abnormal state of a point-type explosion-proof smoke detector.

[0060] In summary, the present application includes at least one of the following beneficial technical effects: accurate detection of abnormal states of explosion-proof smoke sensors is achieved through the integration of multi-dimensional technologies: polling signals and response mechanisms are used to detect line continuity to solve the problem of open circuits and short circuits caused by dust and moisture; pollution anomalies and fault anomalies are identified through cross-verification of smoke sensor signals and image analysis to avoid false alarms of dust and oil and missed reports of sensor element failure; based on HSV color space and pixel concentration analysis, environmental interference is filtered to improve smoke detection accuracy; with the help of positioning marks and distance change values, physical displacement of smoke sensors is warned and detection parameters are adaptively adjusted; the content of the response signal is dynamically adjusted according to the number of addresses or data bit lengths to adapt to the equipment scale and coding rules to ensure communication efficiency and integrity; the polling duration is dynamically adjusted according to the number of unanswered questions to balance detection efficiency and system load; parallel polling and potential difference mechanisms are used to shorten the inspection cycle and enhance anti-interference capabilities; early warning of sensor unit failure is achieved through jump pulse and zero drift analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a step diagram of a method for detecting abnormal conditions of a point-type explosion-proof smoke sensor.

[0062] Figure 2It is set according to the polling number and the feedback conditions. Once there is feedback, the next PLC diagram will be polled.

[0063] Figure 3 It is a PLC diagram that skips polling if there is no feedback for more than 1 second and triggers a power failure, short circuit or smoke sensor loss alarm.

[0064] Figure 4 This is a diagram of the steps for identifying smoke data in the detection image data. DETAILED DESCRIPTION

[0065] Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings.

[0066] Throughout this specification, reference to the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0067] The present application embodiment discloses a method for detecting abnormal state of a point-type explosion-proof smoke sensor, referring to Figures 1 to 3 , including the following steps:

[0068] Obtain multiple explosion-proof smoke detector addresses and create a device identification list.

[0069] According to multiple explosion-proof smoke sensor addresses, polling signals are sent to the corresponding explosion-proof smoke sensor units in turn to actively detect the connection status of the equipment.

[0070] When the smoke sensor unit receives the polling signal, it must return a response signal within a preset time. Wait for the response signal returned by the current explosion-proof smoke sensor unit in response to the polling signal.

[0071] If no response signal is received, the corresponding explosion-proof smoke sensor unit's circuit is abnormal, such as a short circuit or disconnection in the 485 communication line, or the device is offline. This effectively solves circuit failures caused by dust accumulation and moisture corrosion in industrial environments. Otherwise, the smoke sensor unit's smoke signal is regularly acquired, and the explosion-proof camera is used to capture detection image data. The image is then analyzed in the HSV color space using the OpenCV algorithm to accurately identify smoke characteristics.

[0072] When the smoke sensor signal triggers an alarm but no smoke is detected in the image, it is determined that the smoke sensor is contaminated abnormally, such as dust or oil stains adhering to the sensor and interfering with its normal operation, thus avoiding false alarms caused by non-fire factors; conversely, if the smoke sensor signal shows no smoke, but the image analysis finds real smoke, it indicates a smoke sensor failure abnormality, which can locate problems such as sensor element failure and sensing channel blockage to prevent the risk of missed alarms.

[0073] Through a triple detection mechanism consisting of polling signal detection, smoke sensor signal and image analysis cross-validation, this method systematically solves the pain points of explosion-proof smoke sensors in complex industrial environments, such as the difficulty in locating line faults and frequent false alarms and missed alarms, significantly improving the reliability and stability of the fire detection system.

[0074] Reference Figure 2 and Figure 3 ,Section: Corresponding to the functional module division of the smoke detection process, the logic of "communication polling, timed verification, abnormality counting" is split into independent sections, so that the program corresponds to the abnormal state detection method process one by one, such as polling → response → verification, which is convenient for debugging and troubleshooting.

[0075] Sections 7 and 8 focus on “RS-485 communication polling trigger”: simulating the underlying logic of “sending polling signals to smoke sensors and detecting communication status”.

[0076] Sections 9 and 10 focus on “timing verification and abnormal counting”: corresponding to the counting / alarm triggering when “smoke sensor response timeout, signal and image cross-verification abnormality”.

[0077] COM3 (RS-485) ON: The physical channel corresponding to "Sending Polling Signals to Smoke Sensors" is enabled. RS-485 is the communication bus between the smoke sensor (slave device) and the PLC (master). "COM3 ON" indicates that the polling process for the smoke sensor has been initiated.

[0078] M1318: This is the "trigger flag" for communication polling. "COM3 on" drives the rising edge of M1318, simulating the "polling signal issuance" action and triggering the subsequent "wait for smoke sensor response and timer verification" logic. For example, in Section 7, using the rising edge to trigger INC counting can count the "polling times" or "communication trigger events."

[0079] M114 (Feedback Waiting): corresponds to the enabling condition of "Waiting for Smoke Response Signal" in the abnormal state detection method.

[0080] In the smoke detection process, after the PLC sends a polling signal, it needs to "wait for the smoke detection unit to return a response." M114 is the trigger mark of this "waiting process."

[0081] When M114=ON, the T17 timer is started in section 9 to simulate the "preset response waiting time" (such as "returning the response signal within the preset time" in the method). If the timeout occurs, it is determined that "the line is abnormal / the device is offline".

[0082] T17 (timer): corresponds to "response timeout detection" and "cross-validation timing":

[0083] Timeout detection: The T17 timing duration (S2=100, the actual time must be confirmed in conjunction with the time base), corresponding to the "Preset time (timeout threshold for waiting for response)" in the method; if the timing is reached (T17 contact actuates) but no response is received, it is determined that "line abnormality / smoke sensor is offline" (for example, if the INC count is triggered in section 10, the "number of timeout abnormalities" can be counted).

[0084] Cross-validation timing: This function can also be extended to the "synchronous verification duration of smoke sensor signals and image analysis." For example, this function can be used to trigger the "camera capture and smoke sensor signal reading" function at regular intervals, implementing the "regular acquisition of smoke sensor signals and capture of detection images" logic in the method.

[0085] The PLC logic supports the smoke detection process as follows:

[0086] Polling trigger: simulate "sending polling signal to smoke sensor" through "COM3 (RS-485) communication ~ + M1318" to start the detection process;

[0087] Response timeout judgment: Use M114 to enable T17 timing, simulate the "preset time for waiting for response", and the timeout triggers the abnormal count;

[0088] Abnormal statistics and alarms: INC (increase 1) and M115 (output) in the segment correspond to the underlying counts and alarm tags of "prompt line abnormality, smoke pollution / fault abnormality" in the method. This can be expanded: after M115 is triggered, the upload system determines "offline, false alarm or missed alarm" and pops up a window / audio-visual alarm.

[0089] Reference Figure 4 In the step of identifying smoke data in the detection image data, the following sub-steps are also included:

[0090] Based on the HSV color space, the system extracts hue and saturation data from the detected image data. It identifies temporary pixel particles whose hue data falls within a preset hue range and whose saturation data falls within a preset saturation range. The detected image is converted from RGB format to the HSV color space, separating the hue (H), saturation (S), and value (V) channels. Because smoke has a unique color distribution in HSV space (e.g., H values ​​are typically concentrated between 0 and 180°, and S values ​​are below 30), the system sets precise hue and saturation thresholds to extract temporary pixel particles that match the smoke's color characteristics.

[0091] For the extracted temporary pixel particles, record their pixel coordinate positions in the image.

[0092] The Euclidean distance between adjacent pixel particles is calculated as the position distance. If the distance is greater than the preset interference distance (such as 10 pixel units), the particle is determined to be a scattered interference object (such as dust or splashing oil) and is removed; the position distances of the remaining particles are averaged to obtain the aggregation data.

[0093] The aggregated data is compared with the preset target data (such as 3 pixel units). If the aggregated data is smaller than the target data, it indicates that the pixel particles are highly aggregated, which is consistent with the smoke diffusion characteristics, and it is determined that smoke data exists; otherwise, it is determined to be non-smoke interference.

[0094] The detection image data is analyzed based on the HSV color space. By accurately setting the hue and saturation thresholds, the smoke feature area is locked and the temporary pixel particles that meet the conditions are screened out. Subsequently, the system calculates the position distance between the particles, eliminates the discrete interference distance, and uses the mean of the remaining distance as the aggregation data for quantitative evaluation. Compared with traditional detection methods, this method can effectively avoid the interference of environmental pollutants such as dust and oil on the detection results, and avoid false alarms caused by impurities being misjudged as smoke. By digitizing the degree of aggregation of pixel particles, it is possible to accurately distinguish between real smoke and scattered interference objects, greatly improving the accuracy and stability of smoke detection, ensuring that the explosion-proof smoke sensor is more reliable in judging pollution anomalies, significantly reducing the risk of false alarms, and providing strong protection for industrial safety production.

[0095] When the system triggers an abnormality (such as a line fault or contamination alarm), it retrieves the detection image data and, based on preset positioning markers (such as fluorescent markers or QR code labels installed on the walls of explosion-proof smart cabinets), uses computer vision algorithms (such as template matching and feature point detection) to identify fixed positioning units as spatial references. Simultaneously, it uses object detection algorithms (such as YOLO and SSD) to identify and locate the explosion-proof smoke sensor units in the image and obtain their pixel coordinates.

[0096] The Euclidean distance between the positioning unit and the explosion-proof smoke sensor is calculated and compared with the historical baseline distance to determine the distance change. If the change exceeds a preset range (e.g., ±5 pixel deviation), it indicates that the smoke sensor may have physically moved (e.g., loose installation, external force collision), immediately triggering an audible and visual alarm and sending an abnormality message to the monitoring center, prompting operations personnel to investigate the risk. If the change is within the normal range, the smoke detection target data is adjusted proportionally to the change (e.g., initially to 5 pixel units).

[0097] Target data = initial target data + adjustment coefficient k × distance change value. For example, in high-precision detection scenarios that are sensitive to displacement, k can be set to a small value (such as k=0.2) to make target data changes more gradually. In scenarios with high environmental interference and a certain detection tolerance, the k value can be increased (such as k=0.5) to achieve rapid adaptive adjustment of target data. When Δd>0, the smoke sensor moves away from the positioning unit, and the target data increases. When Δd<0, the smoke sensor moves closer to the positioning unit, and the target data decreases, thus ensuring that the accuracy of smoke detection is not affected by changes in the smoke sensor's position.

[0098] Through the above method, not only can timely warnings be issued for abnormal physical installation of smoke sensors, but the adaptive adjustment mechanism can also compensate for the impact of position changes on detection results, significantly improving the anti-interference ability and detection stability of explosion-proof smoke sensors in complex industrial environments.

[0099] The step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps:

[0100] The system traverses all connected explosion-proof smoke sensor addresses and counts the number of addresses N through counter instructions (such as CTU). Refreshing in real time, when smoke sensors are added or removed, the N value is automatically updated to ensure that subsequent communication parameters match the actual device scale.

[0101] The content of the response signal is matched by looking up the table according to the number of addresses. The content is the character length corresponding to the digital signal feature.

[0102] After sending the polling signal, the system starts a timer (such as T17, with an accuracy of 100ms). If a response signal is received within the preset first time duration t0 (such as t0 = 1 second), the data is immediately parsed and the device status is marked.

[0103] If no response is received after t0, the current smoke sensor communication is determined to be abnormal (such as line failure or device offline), but the subsequent process is not blocked and a polling signal is directly sent to the next address. Regardless of whether a response is received, polling is cyclically performed in the order of addresses to ensure continuous operation of the system in a multi-device environment and avoid paralysis of the entire communication due to a single device abnormality.

[0104] A digital signal monitoring mechanism, combined with a dynamic polling queue and adaptive content allocation strategy, enables parallel communication management across multiple devices. The system adjusts the response signal length in real time based on the number of addresses to ensure data frame integrity, while a timeout skipping mechanism prevents global congestion caused by single-node failures.

[0105] The step of matching the content of the response signal according to the number of addresses also includes the following sub-steps:

[0106] Method 1: Adjust the content volume in direct proportion to the number of addresses N; Content volume = M × (N / base number), where M is the preset base content volume. Set the base address number to N0 (e.g., N0 = 16) and the preset base content volume to M (e.g., M = 32 bytes, corresponding to 2 bytes of status data allocated to each address). When the actual number of addresses is N, adjust the content volume in direct proportion using the formula Content volume = M × (N / N0). For example:

[0107] If N=8, then the content size = 32×(8 / 16)=16 bytes, and only the status data of 8 addresses is transmitted;

[0108] If N=24, the content size = 32×(24 / 16)=48 bytes, which can completely carry data such as the smoke concentration of 24 addresses (such as register 40001) and the device status code (such as 40102).

[0109] The data frame length is dynamically expanded or contracted based on the number of connected smoke sensors. When the system is expanded (e.g., with the addition of new smoke sensors), the data frame size increases proportionally, preventing truncation of register values ​​(e.g., the alarm status bit 40003) due to insufficient data frame capacity. When a device is removed, the data frame size automatically decreases, eliminating the "invalid byte padding" issue associated with traditional fixed frame formats (e.g., redundant 0x00s in 485 communications), thereby maximizing bus bandwidth utilization.

[0110] Method 2: Adjust the content volume in direct proportion to the data bit length of the explosion-proof smoke sensor address; Content volume = M × K, where M is the preset baseline content volume and K is the growth coefficient. Analyze the data bit length of the address code (such as 8 bits, 16 bits, 32 bits) and set the growth coefficient K:

[0111] 8-bit address (can represent 0~255 devices) corresponds to K=1;

[0112] 16-bit address (0~65535) corresponds to K=2;

[0113] The 32-bit address (0~4294967295) corresponds to K=4.

[0114] The content size is adjusted using the formula Content Size = M × K. For example, when a 16-bit address smoke sensor is connected, if M = 16 bytes, the content size is automatically adjusted to 16 × 2 = 32 bytes to ensure that the status data encoded in the long address (such as 32-bit address + 16-bit smoke value) is fully transmitted.

[0115] When the number of connected explosion-proof smoke sensor addresses increases, the response signal needs to carry more device status data (such as smoke concentration value, device operation status code, etc.). Therefore, the system dynamically expands the content volume according to the proportion of the number of addresses to ensure the complete transmission of real-time data of each smoke sensor; if the number of addresses decreases, the content volume is reduced synchronously to eliminate the problem of invalid data filling under the fixed frame format.

[0116] The method further comprises the steps of:

[0117] At a set statistical period (e.g., every 10 minutes or after each polling cycle), the PLC uses the counter CTU to count the number of explosion-proof smoke detectors that haven't responded. During the statistical process, if no response is received when polling a specific address, the counter automatically increments by 1 and resets after the cycle ends. For example, in a chemical plant's polling system, the number of unresponsive smoke detectors is counted every 5 minutes, providing real-time information on the communication link status.

[0118] Adjust the first duration based on the number of missed responses:

[0119] If the number of unanswered calls is less than or equal to the preset threshold (e.g., the preset threshold is 2), a local communication delay is determined (e.g., a single smoke sensor is temporarily blocked by dust). The first duration is shortened using the following formula: First duration = Initial duration × (1 - First adjustment coefficient × Number of unanswered calls). For example, the initial duration is 1 second, and α is the first adjustment coefficient, such as 0.1. For example, if the number of unanswered calls is 1, the first duration = 1 × (1 - 0.1 × 1) = 0.9 seconds, increasing the polling speed by 10% and shortening the overall inspection cycle.

[0120] If the number of unanswered calls exceeds the preset threshold, a systemic communication anomaly (such as bus interference or multiple device failures) is detected. The first duration is extended using the second formula: First duration = Initial duration × (1 + Second adjustment factor × (Number of unanswered calls - Preset threshold)). β is the second adjustment factor (e.g., 0.2). For example, if the number of unanswered calls is 4 and the preset threshold is 2, the first duration = 1 × [1 + 0.2 × (4 - 2)] = 1.4 seconds. This reduces the polling frequency by 28.6%, preventing the PLC from sending signals too frequently and causing excessive bus load.

[0121] When only a small number of smoke sensors experience communication delays, the polling speed can be accelerated by shortening the first time duration, which can effectively compress the overall inspection cycle and improve system response efficiency. When a large number of smoke sensors experience communication anomalies, the first time duration can be extended to reduce the polling frequency, which can avoid PLC bus overload caused by high-frequency signal transmission and ensure stable system operation.

[0122] The step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps:

[0123] After sending a polling signal to the current explosion-proof smoke sensor, the PLC directly sends a polling signal to the next one. After the PLC sends a polling signal (such as the MODBUS query frame 0x010x030x400010x00010xCRC on the 485 bus) to the current smoke sensor, it immediately sends the polling signal to the next smoke sensor without waiting for a response, creating a "send-send-wait" pipeline operation mode. For example, in a network of 10 smoke sensors, traditional serial polling takes 10 seconds per cycle. This mechanism reduces the cycle to 1.5 seconds (only the maximum response time + signal transmission time is required), improving inspection efficiency.

[0124] While sending a polling signal, the system waits for a response signal. The initial data segment of the response signal has a lower potential than the initial data segment of the polling signal. The initial segment of the polling signal uses a high potential (e.g., 3.3V TTL), while the initial segment of the response signal is set to a low potential (e.g., 0.5V). A hardware comparator (e.g., LM393) compares the potentials in real time. This allows reliable signal differentiation even when the bus experiences ±0.3V interference fluctuations, preventing false reception due to noise triggering.

[0125] After sending the polling signal, the system starts dual-thread processing: the main thread continues to send polling signals to the next device, and the secondary thread monitors the response signal; when the secondary thread detects the low voltage at the beginning of the response signal, it immediately suspends the polling of the main thread through the interrupt mechanism to ensure that the current response signal is received completely (such as a 16-byte data frame); after the response parsing is completed, the main thread resumes polling and sending to avoid CRC check errors caused by data truncation.

[0126] Using analog signal monitoring technology and a "send-send-wait" parallel polling mechanism, the system achieves efficient communication management for multiple devices. Based on the potential difference between the polling signal and the initial data segment of the response signal, the system uses a hardware comparator to identify the signal type in real time. Upon receiving the response signal, the system dynamically pauses polling and sending to ensure data frame integrity.

[0127] The method further comprises the steps of:

[0128] Acquire the latest multiple smoke sensor signals; the system samples the analog signals (such as 0-5V voltage signals) output by the smoke sensor in real time at a 10kHz sampling rate (which satisfies the Nyquist sampling theorem and ensures the capture of μs-level pulse signals), continuously acquiring the latest 500 signal samples (approximately 50ms of data) to form a dynamic signal buffer.

[0129] Identify the number of jump pulses in the smoke sensor signal. Use a sliding window difference method (with a window width of 5 sampling points) to calculate the signal slope. When the absolute value of the slope exceeds a threshold (e.g., 0.8V / ms), it is considered a valid jump pulse. For example, in a normal, smoke-free state, the signal fluctuates smoothly, with a slope of less than 0.2V / ms. However, when dust particles enter the sensor cavity, light scattering causes a sudden change in the signal, with the slope exceeding 1.5V / ms. A CTD counter counts the number of pulses M per unit time (e.g., 1 second) in real time, creating a pulse count curve.

[0130] Calculate the drift amplitude of the smoke sensor signal's zero-point drift. Define the smoke-free signal baseline as "zero." The system collects 100 interference-free samples every 10 minutes (e.g., during early morning equipment downtime) and calculates the mean value, μ, as the current zero-point reference. Compare this to the initial calibration zero point, μ0 (e.g., 0.3V), and calculate the drift amplitude, Δμ = |μ - μ0|. A first-order low-pass filter (cutoff frequency 1Hz) is used to eliminate high-frequency noise interference to ensure accurate drift calculation.

[0131] If the number of pulses exceeds the preset warning number and the drift amplitude exceeds the preset warning amplitude, it indicates that the sensor unit is malfunctioning. The preset pulse warning number M0 = 15 pulses / second (under normal operating conditions, M≤5 pulses / second), and the drift warning amplitude Δμ0 = 0.4V (exceeding this value may cause signal distortion).

[0132] The "AND gate" logic is used to trigger the early warning: when M>M0 and Δμ>Δμ0, the sensor unit is judged to be malfunctioning to avoid single parameter misjudgment. For example, strong electromagnetic interference may only cause a pulse surge, while component aging may only cause a zero-point drift.

[0133] By collecting the latest smoke sensor signals and analyzing the number of jump pulses and zero-point drift amplitude, the system accurately determines sensor unit malfunction when either indicator exceeds the preset warning threshold. This system effectively identifies signal disturbances caused by contamination factors such as dust accumulation and oil contamination (manifested as an abnormal increase in the number of pulses), as well as zero-point drift (baseline offset outside the normal range) caused by component aging and moisture corrosion, providing early warning of potential sensor unit failures. This technology avoids false or missed smoke alarms caused by signal distortion, improves the detection reliability of explosion-proof smoke sensors in complex industrial environments, provides quantitative data support for equipment maintenance, and ensures the long-term stable operation of fire detection systems.

[0134] An embodiment of the present application further discloses a system for detecting an abnormal state of a point-type explosion-proof smoke detector, comprising a processor, wherein the processor executes the steps of any one of the above-described methods for detecting an abnormal state of a point-type explosion-proof smoke detector.

[0135] An embodiment of the present application further discloses a storage medium, wherein a program is stored in the storage medium. When the program is executed by a processor, the steps of any one of the above-mentioned methods for detecting an abnormal state of a point-type explosion-proof smoke detector are implemented.

[0136] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for detecting abnormal state of point-type explosion-proof smoke detectors, characterized in that: The steps include: Get multiple explosion-proof smoke sensor addresses; Sending polling signals to corresponding explosion-proof smoke sensing units in sequence according to the plurality of explosion-proof smoke sensing addresses; Waiting for a response signal returned by the current explosion-proof smoke sensor unit in response to the polling signal; If the response signal is not received, it indicates that the circuit of the corresponding explosion-proof smoke sensor unit is abnormal; Otherwise, regularly acquiring a smoke signal from the explosion-proof smoke sensing unit, acquiring detection image data by aiming at the explosion-proof smoke sensing unit, and identifying smoke data in the detection image data; If the smoke sensor signal is a smoke sensor alarm and the smoke data is not recognized, it indicates that the smoke sensor is contaminated abnormally; If the smoke sensor signal is a no-smoke signal and the smoke data is recognized, it indicates that the smoke sensor is faulty.

2. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 1 is characterized in that: The step of identifying smoke data in the detection image data further includes the following sub-steps: Extracting hue data and saturation data from the detection image data based on the HSV color space; Identifying temporary pixel particles whose hue data is within a preset hue range and whose saturation data is within a preset saturation range; Obtaining the pixel position of the temporary pixel particle; Calculating the position distance between adjacent temporary pixel particles according to the pixel positions; If the position distance is greater than the preset interference distance, the position distance is eliminated, and the average value of the remaining position distances is calculated as the aggregated data; If the aggregated data is smaller than the preset target data, the smoke data exists; otherwise, the smoke data does not exist.

3. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 2, characterized in that: The method further comprises the steps of: Acquire a positioning mark based on the abnormal prompt, and identify a positioning unit corresponding to the positioning mark and having a fixed position in the detection image data; identifying the explosion-proof smoke sensing unit in the detection image data; Calculating a distance change between the positioning unit and the explosion-proof smoke sensing unit; If the distance change value is outside the preset change range, an on-site abnormality warning is issued; Otherwise, the target data is adjusted in direct proportion to the distance change value.

4. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 1 is characterized in that: The step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps: Calculate the number of addresses according to the explosion-proof smoke sensor addresses; Matching the content of the response signal according to the number of addresses; After sending the polling signal to the current explosion-proof smoke sensor unit, if the response signal returned by the current explosion-proof smoke sensor unit is received within the preset first time period or is not received after the first time period, the polling signal is sent to the next explosion-proof smoke sensor unit.

5. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 4 is characterized in that: The step of matching the content of the response signal according to the number of addresses further includes the following sub-steps: The content volume is adjusted in direct proportion to the number of addresses N; content volume = M × (N / reference volume), where M is a preset reference content volume; or, The content volume is adjusted in direct proportion to the data bit length of the explosion-proof smoke sensor address; content volume=M×K, where M is the preset reference content volume and K is the growth coefficient.

6. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 4 is characterized in that: The method further comprises the steps of: The number of the explosion-proof smoke sensing units that have not sent the response signal within the statistical period is counted as the number of non-response; Adjust the first duration according to the number of unanswered responses: If the number of unanswered calls is less than or equal to the preset threshold, the first duration is adjusted according to the first formula: first duration = initial duration × (1-first adjustment coefficient × number of unanswered calls); If the number of unanswered responses is greater than the preset threshold, the first duration is adjusted according to the second formula: first duration = initial duration × (1 + second adjustment coefficient × (number of unanswered responses - preset threshold)).

7. The abnormal state detection method of point-type explosion-proof smoke detector according to claim 1 is characterized in that: The step of waiting for the explosion-proof smoke sensor unit to return a response signal in response to the polling signal further includes the following sub-steps: After sending the polling signal to the current explosion-proof smoke sensor unit, directly sending the polling signal to the next explosion-proof smoke sensor unit; waiting for the response signal while sending the polling signal, wherein the potential value of the initial data segment of the response signal is lower than the potential value of the initial data segment of the polling signal; If the response signal is received, the polling signal sent to the next explosion-proof smoke sensing unit is suspended, and after a complete response signal is obtained, the polling signal is resent to the next explosion-proof smoke sensing unit.

8. The abnormal state detection method of a point-type explosion-proof smoke detector according to claim 1, characterized in that: The method further comprises the steps of: Obtaining the latest multiple smoke sensor signals; Identifying the number of jump pulses in the smoke sensor signal; Calculating the drift amplitude of the zero point drift in the smoke sensor signal; If the number of pulses is greater than a preset warning number and the drift amplitude is greater than a preset warning amplitude, it indicates that the sensor unit is malfunctioning.

9. A point-type explosion-proof smoke sensor abnormal state detection system, characterized in that: The method comprises a processor, wherein the processor executes the steps of the abnormal state detection method of the point-type explosion-proof smoke detector according to any one of claims 1 to 8.

10. A storage medium, characterized in that: The medium stores a program, and when the program is executed by a processor, the steps of the abnormal state detection method of a point-type explosion-proof smoke detector according to any one of claims 1 to 8 are implemented.

Images