Detector explosion-proof compliance detection method and device based on internet of things, terminal and medium

Abstract

The application discloses an explosion-proof compliance detection method and device based on an Internet of Things, a terminal and a medium. The method comprises the following steps: performing deep feature coding processing on a static feature vector and a dynamic feature vector to obtain a deep feature coding vector; performing classification prediction on an explosion-proof hidden danger part of a detector based on the deep feature coding vector to obtain an explosion-proof hidden danger part prediction probability vector of the detector; determining an explosion-proof hidden danger part risk assessment result of the detector based on the explosion-proof hidden danger part prediction probability vector and dynamic feature data of the detector; and obtaining a target explosion-proof compliance detection result corresponding to the detector based on the explosion-proof hidden danger part risk assessment result and historical explosion-proof compliance detection record data of the detector. The application aims to improve the accuracy and timeliness of explosion-proof compliance detection of the detector, thereby reducing the probability of explosion hazards of the detector.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to an explosion-proof compliance testing method, device, terminal and medium for detectors based on the Internet of Things. Background Technology

[0002] In hazardous environments such as petroleum, chemical, and coal mines where explosive gases or dust exist, detectors (such as combustible gas detectors, temperature detectors, and pressure detectors) serve as core equipment for safety monitoring. Their explosion-proof performance must consistently meet national standards and industry specifications. If a detector loses its explosion-proof capability due to damage to the explosion-proof surface of its casing, aging of seals, or loose cable inlets, it not only fails to effectively perform its monitoring function but may also become an ignition source, potentially causing a catastrophic accident.

[0003] However, the shortcomings of existing explosion-proof compliance testing technologies for detectors are mainly reflected in the following aspects: Existing technologies separate the static attributes of detectors (such as explosion-proof type, material grade, electrical clearance, etc.) from their real-time operating status (such as temperature rise, vibration, internal pressure, etc.), failing to extract the deep-seated hidden danger characteristics after the interaction and integration of the two. Many hidden dangers originate precisely from the gradual failure of static structures under specific dynamic operating conditions, and isolated analysis is prone to missed detections; the coupling relationship between the occurrence probability of each hidden danger location and the actual dynamic anomaly degree is not comprehensively assessed, making it impossible to generate quantitative comprehensive hidden danger risk scores and risk levels, which is difficult to support graded disposal and predictive maintenance; relying on manual periodic inspections has limitations such as long inspection intervals, great subjective influence, and difficulty in covering operational blind spots. Summary of the Invention

[0004] The main purpose of this application is to provide an IoT-based method, device, terminal, and medium for detecting explosion-proof compliance of detectors, aiming to improve the accuracy and timeliness of detector explosion-proof compliance detection, thereby reducing the probability of explosion-proof hazards occurring in detectors.

[0005] To achieve the above objectives, this application provides an IoT-based method for detecting explosion-proof compliance of detectors, the method comprising: The static feature data and dynamic feature data of the detector are preprocessed separately to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The static feature vector and the dynamic feature vector are subjected to deep feature encoding processing to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. Based on the deep feature encoding vector, the explosion-proof hazard parts of the detector are classified and predicted to obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic characteristic data of the detector, the risk assessment result of the explosion-proof hazard location of the detector is determined. Based on the risk assessment results of the explosion-proof hazard areas of the detector and the historical explosion-proof compliance test records of the detector, the target explosion-proof compliance test results corresponding to the detector are obtained.

[0006] Specifically, the step of performing deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoding vector includes: The static feature vector and the dynamic feature vector are concatenated to obtain a concatenated feature vector. By using a pre-defined first fully connected hidden network, a first feature vector is obtained based on the concatenated feature vector; By using a pre-defined second fully connected hidden network, a deep encoding feature vector is obtained based on the first feature vector.

[0007] Specifically, the step of obtaining the first feature vector by means of a preset first fully connected hidden network and the concatenated feature vector includes: Calculate the product between the first weight matrix of the preset first fully connected hidden network and the concatenated feature vector to obtain the first intermediate feature vector; The first intermediate feature vector is added element-wise to the first bias vector of the preset first fully connected hidden network to obtain the first vector after addition; The first feature vector is obtained by performing modified linear unit activation on each scalar element of the first summed vector.

[0008] Specifically, the step of obtaining the deep encoding feature vector based on the first feature vector through a preset second fully connected hidden network includes: The second intermediate feature vector is obtained by multiplying the second weight matrix of the preset second fully connected hidden network with the first feature vector. The second intermediate feature vector is added element-wise to the second bias vector of the preset second fully connected hidden network to obtain the second vector after addition; By using a preset activation function, each element in the second summed vector is activated to obtain the deep coding feature vector.

[0009] Specifically, the step of classifying and predicting the explosion-proof hazard locations of the detector based on the deep feature encoding vector to obtain the predicted probability vector of the explosion-proof hazard locations of the detector includes: By using a pre-defined third fully connected hidden network, the feature vector of the hidden danger location is obtained based on the deep coding feature vector; The original scoring vector is obtained by multiplying the feature vector of the hidden danger location with the preset output weight matrix; The original scoring vector is added element by element to the preset output bias vector to obtain the third vector after addition; The vector obtained by the third summation is normalized to obtain the predicted probability vector of the explosion-proof hazard location of the detector. The location corresponding to the maximum predicted probability value in the predicted probability vector of the explosion-proof hazard location of the detector is the predicted explosion-proof hazard location.

[0010] Specifically, the risk assessment results for the explosion-proof hazard locations of the detector include suspected hazard locations, risk assessment levels, and comprehensive hazard risk scores; The process of determining the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location and the dynamic feature data of the detector includes: Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic feature data of the detector, the probability sensitivity threshold and the location risk weight table corresponding to the explosion-proof hazard location are obtained. The probability sensitivity threshold is used to characterize the differentiated severe consequences after the failure of each explosion-proof hazard location, and the location risk weight table is used to characterize the risk probability of each explosion-proof hazard location in the overall detector. The explosion-proof hazard locations whose predicted probability values ​​in the predicted probability vector of the detector's explosion-proof hazard locations are greater than the probability sensitivity threshold are identified as the suspected hazard locations. Based on the aforementioned risk weight table for specific locations, risk factors corresponding to locations with potential hazards are obtained. The dynamic feature values ​​in the dynamic feature data of the detector are normalized to obtain a normalized dynamic anomaly index. Calculate the product between the risk factor corresponding to each suspected hidden danger location and the normalized dynamic abnormal indicator corresponding to each suspected hidden danger location to obtain the product result corresponding to each suspected hidden danger location, and sum up the product results corresponding to each suspected hidden danger location to obtain the comprehensive hidden danger risk score; The comprehensive hazard risk score is converted into the risk assessment level, wherein the risk assessment level includes high risk, medium risk and low risk.

[0011] Specifically, the step of obtaining the target explosion-proof compliance test result for the detector based on the risk assessment results of the detector's explosion-proof hazard locations and the detector's historical explosion-proof compliance test records includes: Each historical explosion-proof compliance test record in the detector's historical explosion-proof compliance test record data is weighted to obtain a time decay weight, wherein the time decay weight is used to characterize the time distance between the historical explosion-proof compliance test record and the current time; Based on the time decay weight, the weighted compatibility score between the suspected hidden danger location and the unqualified hidden danger location in the historical explosion-proof compliance test record data of the detector is calculated; Based on the risk assessment level and the weighted consistency score, the target explosion-proof compliance test result is determined. If the risk assessment level is high risk and the weighted consistency score is greater than a first preset threshold, the target explosion-proof compliance test result is determined to be non-compliant. If the risk assessment level is medium risk and the weighted consistency score is between the first preset threshold and a second preset threshold, the target explosion-proof compliance test result is determined to be pending review. If the risk assessment level is low risk and the weighted consistency score is less than a second preset threshold, the target explosion-proof compliance test result is determined to be compliant. Wherein, the first preset threshold value is greater than the second preset threshold value, and both the first preset threshold value and the second preset threshold value are positive numbers.

[0012] To achieve the above objectives, this application also provides an IoT-based detector explosion-proof compliance testing device, the device comprising: The first unit is used to preprocess the static feature data and dynamic feature data of the detector respectively to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The second unit is used to perform deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. The third unit is used to classify and predict the explosion-proof hazard parts of the detector based on the deep feature encoding vector, and obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector of the detector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. The fourth unit is used to determine the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic feature data of the detector. The fifth unit is used to obtain the target explosion-proof compliance test result of the detector based on the risk assessment results of the explosion-proof hazard parts of the detector and the historical explosion-proof compliance test record data of the detector.

[0013] To achieve the above objectives, this application also provides a terminal, including a memory storing multiple instructions; the processor loads instructions from the memory to execute the steps in any of the methods provided in this application.

[0014] To achieve the above objectives, this application also provides a medium storing a plurality of instructions adapted for loading by a processor to execute the steps in any of the methods provided in this application.

[0015] This application provides a method, device, terminal, and medium for detecting explosion-proof compliance of detectors based on the Internet of Things. First, deep feature encoding is performed on static and dynamic feature vectors to obtain deep feature encoded vectors. Based on these deep feature encoded vectors, explosion-proof hazard locations of the detector are classified and predicted to obtain a predicted probability vector for these locations. Based on this predicted probability vector and the detector's dynamic feature data, a risk assessment result for the detector's explosion-proof hazard locations is determined. Based on this risk assessment result and the detector's historical explosion-proof compliance detection records, a target explosion-proof compliance detection result for the detector is obtained. This improves the accuracy and timeliness of detector explosion-proof compliance detection, thereby reducing the probability of explosion-proof hazards occurring. Attached Figure Description

[0016] Figure 1 A flowchart illustrating the method provided in the embodiments of this application; Figure 2 This is a schematic diagram illustrating the application process of deep feature encoding processing provided in the embodiments of this application; Figure 3 This is a schematic diagram of the device provided in the embodiments of this application; Figure 4 This is a schematic diagram of the terminal structure provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] The shortcomings of existing explosion-proof compliance testing technologies for detectors are mainly reflected in the following aspects: Existing technologies separate the static attributes of detectors (such as explosion-proof type, material grade, electrical clearance, etc.) from their real-time operating status (such as temperature rise, vibration, internal pressure, etc.), failing to extract the deep-seated hidden danger characteristics after the interaction and integration of the two. Many hidden dangers originate precisely from the gradual failure of static structures under specific dynamic operating conditions, and isolated analysis is prone to missed detections; the coupling relationship between the occurrence probability of each hidden danger location and the actual dynamic anomaly degree is not comprehensively assessed, making it impossible to generate quantitative comprehensive hidden danger risk scores and risk levels, which is difficult to support graded disposal and predictive maintenance; relying on manual periodic inspections has limitations such as long inspection intervals, great subjective influence, and difficulty in covering operational blind spots.

[0019] Therefore, this application provides an IoT-based method, device, terminal, and medium for detecting explosion-proof compliance of detectors to solve practical technical problems.

[0020] In some embodiments, the device may be integrated into an electronic device, such as a terminal or server.

[0021] In some embodiments, the server may also be implemented as a terminal.

[0022] The server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0023] The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and the server can be connected directly or indirectly through wired or wireless communication, which is not limited herein.

[0024] The following sections provide detailed descriptions of each example. It should be noted that the sequence numbers of the following embodiments are not intended to limit the preferred order of the embodiments.

[0025] This application provides an IoT-based method for detecting the explosion-proof compliance of detectors. This method can improve the accuracy and timeliness of the explosion-proof compliance detection of detectors, thereby reducing the probability of explosion-proof hazards occurring in detectors.

[0026] In some embodiments, the Internet of Things (IoT) upon which the method relies is not merely a data transporter, but also the infrastructure supporting the entire intelligent detection process. The IoT architecture may include the following four core layers: The perception layer, the "nerve endings" of the Internet of Things, is directly deployed on the detector itself and in its surrounding environment. In addition to the detector's built-in gas concentration and temperature sensing units, it also collects key dynamic characteristics related to explosion protection through integrated or additional sensors. These include: changes in the gap of the explosion-proof joint surface (via miniature displacement sensors), shell temperature rise and key point temperatures (infrared thermography or patch sensors), internal pressure fluctuations, localized heating caused by loose wiring terminals, and abnormal shell vibration. Simultaneously, it acquires static characteristic data such as the detector's explosion-proof markings, shell material, and protection level by reading the electronic nameplate or through backend data entry. The perception layer provides all the raw dynamic and static data input required by the neural network model.

[0027] Network Layer: Responsible for the secure and reliable transmission of sensed data to the platform. Considering the harshness of explosion-proof environments, a combination of wired and wireless methods is typically used. For example, intrinsically safe industrial Ethernet or fieldbuses (such as Modbus TCP, PROFIBUS) are used in fixed installation areas, while explosion-proof certified industrial wireless technologies (such as WirelessHART, a wireless network protocol designed specifically for process industries) are used in mobile inspections or areas where cabling is difficult. This layer ensures that massive amounts of raw characteristic data can be aggregated in real time and completely, forming the foundation for implementing detection methods.

[0028] Platform Layer: The "brain" of the Internet of Things (IoT) and the carrier on which the methods described run. The platform layer provides the following three main capabilities: Data access and preprocessing: Cleaning, aligning and standardizing the heterogeneous data uploaded by the sensing layer, completing the specific operations of "preprocessing the static feature data and dynamic feature data of the detector respectively" in step one of the method. Algorithm Model and Decision Deployment: The core dual-mode algorithm described above is deployed and run here. First, the preprocessed data is subjected to deep feature encoding and hazard location classification prediction through a neural network model; then, the driving logic algorithm calls the historical detection record database based on the prediction results, performs risk quantification assessment and compliance judgment logic that integrates historical records, and finally outputs the target explosion-proof compliance detection result; Historical Data Lake and Knowledge Base: Stores massive amounts of historical testing records, equipment files, maintenance logs, and domain knowledge, providing resident data support for the logic algorithm part with "detector historical explosion-proof compliance testing record data".

[0029] Application Layer: The ultimate window showcasing the value of the Internet of Things (IoT), directly facing security engineers and managers. The application layer dynamically displays the risk level (high, medium, low) of each detector, specific locations of potential hazards (e.g., "the sealing point of wiring cavity 3"), hazard recurrence alerts based on historical data analysis, and ultimately, a clear conclusion and action recommendations for "compliant," "requires review," or "non-compliant." This layer transforms the abstract output of the algorithm into actionable security management actions.

[0030] In some embodiments, the ethylene cracking unit area of ​​a petrochemical plant belongs to Zone 1 of gas explosion protection and is equipped with a large number of fixed infrared combustible gas detectors to monitor hydrocarbon leaks at critical locations such as pipeline flanges and valves. This specific embodiment uses one detector, tag number GT-1205, as an example to illustrate how the method performs online explosion-proof compliance testing.

[0031] like Figure 1 The specific process of the method can be as follows: S110. Preprocess the static feature data and dynamic feature data of the detector respectively to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data.

[0032] In some embodiments, the IoT platform acquires the raw data from the GT-1205 through two methods: Static characteristic data: The detector's factory records and design parameters can be directly read from the factory's asset management system (AMS).

[0033] Dynamic characteristic data: The operating status and environmental parameters of the GT-1205 can be collected in real time through various sensors deployed on and near the GT-1205 body.

[0034] Therefore, the specific data obtained is as follows: Static characteristics data: Explosion-proof type 'Ex d IIC T6 Gb' (numerical code), housing material '316L stainless steel' (code), protection level IP66 (code), nameplate rated electrical clearance 12mm, creepage distance 15mm, and service life 3 years (numerical 3); Dynamic characteristic data: internal temperature 45°C, ambient combustible gas concentration 0%LEL (normal), internal pressure 101.5kPa, terminal temperature rise in wiring cavity 4°C, and housing vibration amplitude 0.5mm / s.

[0035] Specifically, these two types of unstructured or heterogeneous data can be cleaned, normalized, and vectorized for encoding, ultimately generating a static feature vector (dimension m=6) and a dynamic feature vector (dimension n=5).

[0036] S120. Perform deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector.

[0037] In some embodiments, such as Figure 2 The static feature vector and the dynamic feature vector are subjected to deep feature encoding processing to obtain a deep feature encoded vector, including the steps S121 to S123 as shown below: S121. The static feature vector and the dynamic feature vector are concatenated to obtain a concatenated feature vector.

[0038] In some embodiments, the static feature vector (1×6) and the dynamic feature vector (1×5) are concatenated end to end to form a concatenated feature vector with a length of 11.

[0039] S122. A first feature vector is obtained by using a preset first fully connected hidden network and the spliced ​​feature vector.

[0040] In some embodiments, obtaining the first feature vector by means of a preset first fully connected hidden network and the concatenated feature vector includes the following steps S1221 to S1223: S1221. Calculate the product between the first weight matrix of the preset first fully connected hidden network and the concatenated feature vector to obtain the first intermediate feature vector.

[0041] In some embodiments, the product of the first weight matrix (11×64) of the preset first fully connected hidden network and the concatenated feature vector (1×11) is calculated to obtain the first intermediate feature vector with a dimension of 64.

[0042] S1222. Add the first intermediate feature vector to the first bias vector of the preset first fully connected hidden network element by element to obtain the first vector after addition.

[0043] In some embodiments, the first intermediate feature vector and the first bias vector (1×64) are added element by element to obtain the first added vector.

[0044] S1223. Perform modified linear unit activation processing on each scalar element of the first summed vector to obtain the first feature vector.

[0045] In some embodiments, the first summed vector is activated by ReLU, that is, all positive values ​​are retained and negative values ​​are set to zero, and the first feature vector (1×64) is output.

[0046] S123. Using a pre-set second fully connected hidden network, a deep coding feature vector is obtained based on the first feature vector.

[0047] In some embodiments, the step of obtaining the deep coding feature vector based on the first feature vector through a preset second fully connected hidden network includes the following steps S1231 to S1233: S1231. Calculate the product between the second weight matrix of the preset second fully connected hidden network and the first feature vector to obtain the second intermediate feature vector.

[0048] In some embodiments, the product of the second weight matrix (64×32) of the preset second fully connected hidden network and the first feature vector (1×64) is calculated to obtain a second intermediate feature vector with a dimension of 32.

[0049] S1232. Add the second intermediate feature vector to the second bias vector of the preset second fully connected hidden network element by element to obtain the second vector after addition.

[0050] In some embodiments, the second intermediate feature vector and the second bias vector (1×32) are added element by element to obtain the second added vector.

[0051] S1233. By using a preset activation function, each element in the second summed vector is activated to obtain the deep coding feature vector.

[0052] In some embodiments, each element in the second summed vector is processed by a preset activation function (which may be ReLU) to obtain the final deep coding feature vector (1×32), which is a deep coding feature vector containing highly abstract composite risk features such as "the tendency of the stainless steel shell to relax the stress on the explosion-proof surface under internal 45°C and slight vibration".

[0053] S130. Based on the deep feature encoding vector, the explosion-proof hazard parts of the detector are classified and predicted to obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector of the detector is used to characterize the prediction probability of each explosion-proof hazard part of the detector.

[0054] In some embodiments, the step of classifying and predicting the explosion-proof hazard locations of the detector based on the deep feature encoding vector to obtain the predicted probability vector of the explosion-proof hazard locations of the detector includes the following steps S131 to S134: S131. By using a preset third fully connected hidden network, the feature vector of the hidden danger location is obtained based on the deep coding feature vector.

[0055] In some embodiments, the deep encoded feature vector (1×32) is further mapped by a preset third fully connected hidden network. The product of its third weight matrix (32×16) and the input vector is calculated, and after adding the third bias vector (1×16), it is activated by ReLU to output the extracted feature vector of the hidden danger location (1×16).

[0056] S132. Calculate the product between the feature vector of the hidden danger location and the preset output weight matrix to obtain the original scoring vector.

[0057] In some embodiments, the product of the feature vector (1×16) of the hazard location and the preset output weight matrix (16×K) is calculated to obtain the original score vector of 1×K. The total number of hazard location categories K is set to 4, corresponding to: 0-explosion-proof joint surface, 1-cable entry device, 2-transparent sealing area, and 3-inside wiring cavity.

[0058] S133. The original scoring vector is added element by element to the preset output bias vector to obtain the third vector after addition.

[0059] In some embodiments, the original scoring vector is added element-wise to a preset output bias vector of length K to obtain a third vector after addition.

[0060] S134. Normalize the vector after the third summation to obtain the predicted probability vector of the explosion-proof hazard location of the detector, wherein the location corresponding to the maximum predicted probability value in the predicted probability vector of the explosion-proof hazard location of the detector is the predicted explosion-proof hazard location.

[0061] In some embodiments, the vector after the third summation is normalized using Softmax to obtain the final probability vector for predicting explosion-proof hazard locations of the detector. For example, the probability vector for predicting explosion-proof hazard locations of the detector is [0.05, 0.70, 0.15, 0.10]. Among them, the probability value of index 1 (cable entry device) is the largest at 0.70, therefore the explosion-proof hazard location predicted by the model is "cable entry device".

[0062] S140. Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic characteristic data of the detector, determine the risk assessment result of the explosion-proof hazard location of the detector.

[0063] In some embodiments, the risk assessment results of the detector's explosion-proof hazard locations include suspected hazard locations, risk assessment levels, and comprehensive hazard risk scores.

[0064] Specifically, determining the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location and the dynamic feature data of the detector includes the following steps S141 to S146: S141. Based on the predicted probability vector of the explosion-proof hazard parts of the detector and the dynamic feature data of the detector, obtain the probability sensitivity threshold and the part risk weight table corresponding to the explosion-proof hazard parts. The probability sensitivity threshold is used to characterize the differentiated severe consequences after the failure of each explosion-proof hazard part, and the part risk weight table is used to characterize the risk probability of each explosion-proof hazard part in the overall detector.

[0065] In some embodiments, corresponding parameters can be retrieved from a preset knowledge base based on the GT-1205 model and explosion-proof rating. For example, the obtained probability sensitivity thresholds are: explosion-proof surface 0.6, cable entry 0.40, transparent part 0.65, and wiring cavity 0.55. Simultaneously, a risk weight table for each part is obtained, recording the probability of an explosion caused by the failure of each part, i.e., the risk factor.

[0066] S142. Determine the explosion-proof hazard locations whose predicted probability values ​​in the predicted probability vector of the detector's explosion-proof hazard locations are greater than the probability sensitivity threshold as the suspected hazard locations.

[0067] In some embodiments, for example, the predicted probability vector [0.05, 0.70, 0.15, 0.10] is compared one by one with a threshold. Only the probability of 0.70 for "cable entry device" is greater than its threshold of 0.40. Therefore, the suspected potential hazard location is identified as "cable entry device".

[0068] S143. Based on the risk weight table of the location, obtain the risk factors corresponding to the locations with suspected hidden dangers.

[0069] In some embodiments, referring to the risk weight table, the risk factor corresponding to "cable entry device" is 0.85, indicating that the risk is extremely high and may directly damage the explosion-proof integrity.

[0070] S144. Normalize the dynamic feature values ​​in the dynamic feature data of the detector to obtain normalized dynamic anomaly index.

[0071] In some embodiments, current dynamic characteristic data is extracted. For example, "terminal temperature rise of 4°C" is a key warning parameter, i.e., a dynamic characteristic value, which is caused by loose cable entry device and increased contact resistance leading to overheating. The normal operating range of the above parameter is preset to 0-60°C. The dynamic characteristic value is normalized, 4°C / 60°C≈0.067, resulting in a normalized dynamic anomaly index of 0.067.

[0072] S145. Calculate the product between the risk factor corresponding to each suspected hidden danger location and the normalized dynamic abnormal indicator corresponding to each suspected hidden danger location to obtain the product result corresponding to each suspected hidden danger location, and sum up the product results corresponding to each suspected hidden danger location to obtain the comprehensive hidden danger risk score.

[0073] In some embodiments, continuing from the above, a single calculation is performed on the only suspected location. The risk factor (0.85) is multiplied by the normalized dynamic anomaly index (0.067), and the product is 0.057. After summing, the comprehensive hazard risk score is obtained as 0.057.

[0074] S146. The comprehensive hidden danger risk score is converted into the risk assessment level, wherein the risk assessment level includes high risk, medium risk and low risk.

[0075] In some embodiments, the system can be pre-set with the following scoring ranges: 0-0.2 for low risk, 0.2-0.6 for medium risk, and 0.6-1 for high risk. Since the score of 0.057 falls within the low-risk range, the current risk assessment level of GT-1205 is "low risk".

[0076] S150. Based on the risk assessment results of the explosion-proof hazard parts of the detector and the historical explosion-proof compliance test records of the detector, the target explosion-proof compliance test results corresponding to the detector are obtained.

[0077] In some embodiments, obtaining the target explosion-proof compliance test result of the detector based on the risk assessment result of the explosion-proof hazard location of the detector and the historical explosion-proof compliance test record data of the detector includes the following steps S151 to S153: S151. Weighting is performed on each historical explosion-proof compliance test record in the historical explosion-proof compliance test record data of the detector to obtain a time decay weight, wherein the time decay weight is used to characterize the time distance between the historical explosion-proof compliance test record and the current time.

[0078] In some embodiments, for example, three historical explosion-proof compliance test records from the past two years are retrieved from the database for the GT-1205 detector: 10 days ago: Not up to standard, the potential problem was "cable entry device (aging seal ring)", which has been repaired; 190 days ago: Passed; 370 days ago: Unqualified, the potential problem was "explosion-proof joint surface (scratched)", the entire shell has been replaced.

[0079] In some embodiments, weighting is performed using an exponential decay function with a decay coefficient of 0.005 and a time-to-day time-to-weight ratio, including: The weight of record 1 is e^(-0.005*10)≈0.95; The weight of record 2 is e^(-0.005*190)≈0.39; The weight of record 3 is e^(-0.005*370)≈0.16.

[0080] S152. Based on the time decay weight, calculate the weighted compatibility score between the suspected hidden danger location and the unqualified hidden danger location in the historical explosion-proof compliance test record data of the detector.

[0081] In some embodiments, continuing from the above, the current suspected potential hazard location is the "cable entry device". Traversing the historical records, only record 1 has the same "unqualified potential hazard location". Therefore, the weighted compliance score is the weight of record 1, with a score of 0.95.

[0082] S153. Based on the risk assessment level and the weighted consistency score, determine the target explosion-proof compliance test result. If the risk assessment level is high risk and the weighted consistency score is greater than the first preset threshold, then the target explosion-proof compliance test result is determined to be non-compliant. If the risk assessment level is medium risk and the weighted consistency score is between the first preset threshold and the second preset threshold, then the target explosion-proof compliance test result is determined to be pending review. If the risk assessment level is low risk and the weighted consistency score is less than the second preset threshold, then the target explosion-proof compliance test result is determined to be compliant. Wherein, the first preset threshold is greater than the second preset threshold, and both the first preset threshold and the second preset threshold are positive numbers.

[0083] In some embodiments, continuing from the above, the system presets a first threshold value of 0.5 and a second threshold value of 0.15.

[0084] Specifically, the current risk assessment level is "low risk," and the weighted overlap score of 0.95 is less than the first threshold of 0.5. According to the "compliance" criteria (low risk and score less than the second threshold), it's important to note that although the score of 0.95 > 0.15, which doesn't meet the stringent condition of "less than the second threshold," because the risk is low and the high overlap score only indicates a recent similar hazard, it doesn't qualify for a direct "non-compliant" or "pending review" assessment. Since the risk level is low, and the only matching historical record shows that the hazard was repaired 10 days ago, under the low-risk premise, even with a high overlap, the system's final judgment logic is: "The historical similar hazard has been marked for repair; currently, there are only very early, minor abnormal signs, which do not constitute a direct threat." Therefore, the target explosion-proof compliance test result for GT-1205 is "compliant." Simultaneously, a prompt can be automatically generated in the maintenance recommendations: "The detector has a recent history of repairing a cable entry device hazard; it is recommended to focus on the condition of its sealing components during the next routine inspection." In summary, this application provides an IoT-based method for detecting the explosion-proof compliance of detectors, which improves the accuracy and timeliness of the detection, thereby reducing the probability of explosion-proof hazards occurring in detectors.

[0085] To better implement the above methods, this application also provides an IoT-based detector explosion-proof compliance testing device. This device can be integrated into an electronic device, such as a terminal or server. The terminal can be a mobile phone, tablet, smart Bluetooth device, laptop, or personal computer; the server can be a single server or a server cluster consisting of multiple servers.

[0086] For example, in this embodiment, the method of this application embodiment will be described in detail by taking the specific integration of an IoT-based detector explosion-proof compliance testing device into a terminal as an example.

[0087] For example, such as Figure 3 As shown, the IoT-based detector explosion-proof compliance testing device 300 may include a first unit 301, a second unit 302, a third unit 303, a fourth unit 304, and a fifth unit 305. The device includes: The first unit is used to preprocess the static feature data and dynamic feature data of the detector respectively to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The second unit is used to perform deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. The third unit is used to classify and predict the explosion-proof hazard parts of the detector based on the deep feature encoding vector, and obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector of the detector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. The fourth unit is used to determine the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic feature data of the detector. The fifth unit is used to obtain the target explosion-proof compliance test result of the detector based on the risk assessment results of the explosion-proof hazard parts of the detector and the historical explosion-proof compliance test record data of the detector.

[0088] In practice, each of the above units can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units, please refer to the previous method embodiments, which will not be repeated here.

[0089] As can be seen from the above, the embodiments of this application can improve the accuracy and timeliness of explosion-proof compliance testing of detectors, thereby reducing the probability of explosion-proof hazards occurring in detectors.

[0090] This application also provides an electronic device, which can be a terminal, a server, or other similar device. The terminal can be a mobile phone, tablet computer, smart Bluetooth device, laptop computer, personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers, etc.

[0091] In some embodiments, the product processing device may also be integrated into multiple electronic devices, such as multiple servers, with multiple servers implementing the IoT-based detector explosion-proof compliance testing method of this application.

[0092] In this embodiment, the electronic device will be described in detail as a terminal, for example, such as... Figure 4 As shown, it illustrates a structural schematic diagram of the terminal 400 involved in an embodiment of this application. Specifically: The terminal 400 may include components such as a processor 401 with one or more processing cores, a memory 402 with one or more media, a power supply 403, an input module 404, and a communication module 405. Those skilled in the art will understand that... Figure 4The terminal 400 structure shown does not constitute a limitation on the terminal 400, and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein: The processor 401 is the control center of the terminal 400. It connects various parts of the terminal 400 via various interfaces and lines, and performs various functions and processes data by running or executing software programs and / or modules stored in the memory 402, and by calling data stored in the memory 402, thereby providing overall monitoring of the terminal 400. In some embodiments, the processor 401 may include one or more processing cores; in some embodiments, the processor 401 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor 401.

[0093] The memory 402 can be used to store software programs and modules. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the terminal 400, etc. In addition, the memory 402 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 with access to the memory 402.

[0094] The terminal 400 also includes a power supply 403 that supplies power to the various components. In some embodiments, the power supply 403 can be logically connected to the processor 401 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 403 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0095] The terminal 400 may also include an input module 404, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0096] The terminal 400 may also include a communication module 405. In some embodiments, the communication module 405 may include a wireless module. The terminal 400 can perform short-range wireless transmission through the wireless module of the communication module 405, thereby providing users with wireless broadband Internet access. For example, the communication module 405 can be used to help users send and receive emails, browse web pages, and access streaming media.

[0097] Although not shown, terminal 400 may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, processor 401 in terminal 400 loads the executable files corresponding to the processes of one or more applications into memory 402 according to the following instructions, and processor 401 runs the applications stored in memory 402 to realize various functions, as follows: The static feature data and dynamic feature data of the detector are preprocessed separately to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The static feature vector and the dynamic feature vector are subjected to deep feature encoding processing to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. Based on the deep feature encoding vector, the explosion-proof hazard parts of the detector are classified and predicted to obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic characteristic data of the detector, the risk assessment result of the explosion-proof hazard location of the detector is determined. Based on the risk assessment results of the explosion-proof hazard areas of the detector and the historical explosion-proof compliance test records of the detector, the target explosion-proof compliance test results corresponding to the detector are obtained.

[0098] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0099] As can be seen from the above, the embodiments of this application can improve the accuracy and timeliness of explosion-proof compliance testing of detectors, thereby reducing the probability of explosion-proof hazards occurring in detectors.

[0100] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be accomplished by instructions, or by instructions controlling related hardware. These instructions can be stored in a medium and loaded and executed by a processor.

[0101] To this end, embodiments of this application provide a medium storing multiple instructions that can be loaded by a processor to execute steps in any of the IoT-based detector explosion-proof compliance testing methods provided in embodiments of this application. For example, the instructions can execute the following steps: The static feature data and dynamic feature data of the detector are preprocessed separately to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The static feature vector and the dynamic feature vector are subjected to deep feature encoding processing to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. Based on the deep feature encoding vector, the explosion-proof hazard parts of the detector are classified and predicted to obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic characteristic data of the detector, the risk assessment result of the explosion-proof hazard location of the detector is determined. Based on the risk assessment results of the explosion-proof hazard areas of the detector and the historical explosion-proof compliance test records of the detector, the target explosion-proof compliance test results corresponding to the detector are obtained.

[0102] The medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0103] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a medium. A processor of a computer device reads the computer instructions from the medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations of the above embodiments.

[0104] Since the instructions stored in the medium can execute the steps in any of the IoT-based detector explosion-proof compliance testing methods provided in the embodiments of this application, the beneficial effects that any of the IoT-based detector explosion-proof compliance testing methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0105] The above provides a detailed description of the IoT-based explosion-proof compliance testing method, device, terminal, and medium for detectors. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for detecting explosion-proof compliance of detectors based on the Internet of Things, characterized in that, The method includes: The static feature data and dynamic feature data of the detector are preprocessed separately to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The static feature vector and the dynamic feature vector are subjected to deep feature encoding processing to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. Based on the deep feature encoding vector, the explosion-proof hazard parts of the detector are classified and predicted to obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic characteristic data of the detector, the risk assessment result of the explosion-proof hazard location of the detector is determined. Based on the risk assessment results of the explosion-proof hazard areas of the detector and the historical explosion-proof compliance test records of the detector, the target explosion-proof compliance test results corresponding to the detector are obtained.

2. The method as described in claim 1, characterized in that, The step of performing deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoded vector includes: The static feature vector and the dynamic feature vector are concatenated to obtain a concatenated feature vector. By using a pre-defined first fully connected hidden network, a first feature vector is obtained based on the concatenated feature vector; By using a pre-defined second fully connected hidden network, a deep encoding feature vector is obtained based on the first feature vector.

3. The method as described in claim 2, characterized in that, The step of obtaining the first feature vector through a preset first fully connected hidden network based on the concatenated feature vector includes: Calculate the product between the first weight matrix of the preset first fully connected hidden network and the concatenated feature vector to obtain the first intermediate feature vector; The first intermediate feature vector is added element-wise to the first bias vector of the preset first fully connected hidden network to obtain the first vector after addition; The first feature vector is obtained by performing modified linear unit activation on each scalar element of the first summed vector.

4. The method as described in claim 2, characterized in that, The process of obtaining a deep encoding feature vector by means of a pre-defined second fully connected hidden network based on the first feature vector includes: The second intermediate feature vector is obtained by multiplying the second weight matrix of the preset second fully connected hidden network with the first feature vector. The second intermediate feature vector is added element-wise to the second bias vector of the preset second fully connected hidden network to obtain the second vector after addition; By using a preset activation function, each element in the second summed vector is activated to obtain the deep coding feature vector.

5. The method as described in claim 1, characterized in that, The process of classifying and predicting explosion-proof hazard locations of the detector based on the deep feature encoding vector to obtain a predicted probability vector for explosion-proof hazard locations of the detector includes: By using a pre-defined third fully connected hidden network, the feature vector of the hidden danger location is obtained based on the deep coding feature vector; The original scoring vector is obtained by multiplying the feature vector of the hidden danger location with the preset output weight matrix; The original scoring vector is added element by element to the preset output bias vector to obtain the third vector after addition; The vector obtained by the third summation is normalized to obtain the predicted probability vector of the explosion-proof hazard location of the detector. The location corresponding to the maximum predicted probability value in the predicted probability vector of the explosion-proof hazard location of the detector is the predicted explosion-proof hazard location.

6. The method as described in claim 1, characterized in that, The risk assessment results for the explosion-proof potential hazards of the detector include the suspected hazard location, the risk assessment level, and the comprehensive hazard risk score; The process of determining the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location and the dynamic feature data of the detector includes: Based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic feature data of the detector, the probability sensitivity threshold and the location risk weight table corresponding to the explosion-proof hazard location are obtained. The probability sensitivity threshold is used to characterize the differentiated severe consequences after the failure of each explosion-proof hazard location, and the location risk weight table is used to characterize the risk probability of each explosion-proof hazard location in the overall detector. The explosion-proof hazard locations whose predicted probability values ​​in the predicted probability vector of the detector's explosion-proof hazard locations are greater than the probability sensitivity threshold are identified as the suspected hazard locations. Based on the aforementioned risk weight table for specific locations, risk factors corresponding to locations with potential hazards are obtained. The dynamic feature values ​​in the dynamic feature data of the detector are normalized to obtain a normalized dynamic anomaly index. Calculate the product between the risk factor corresponding to each suspected hidden danger location and the normalized dynamic abnormal indicator corresponding to each suspected hidden danger location to obtain the product result corresponding to each suspected hidden danger location, and sum up the product results corresponding to each suspected hidden danger location to obtain the comprehensive hidden danger risk score; The comprehensive hazard risk score is converted into the risk assessment level, wherein the risk assessment level includes high risk, medium risk and low risk.

7. The method as described in claim 6, characterized in that, The explosion-proof compliance test results for the detector are obtained based on the risk assessment results of the detector's explosion-proof hazard locations and the detector's historical explosion-proof compliance test records, including: Each historical explosion-proof compliance test record in the detector's historical explosion-proof compliance test record data is weighted to obtain a time decay weight, wherein the time decay weight is used to characterize the time distance between the historical explosion-proof compliance test record and the current time; Based on the time decay weight, the weighted compatibility score between the suspected hidden danger location and the unqualified hidden danger location in the historical explosion-proof compliance test record data of the detector is calculated; Based on the risk assessment level and the weighted consistency score, the target explosion-proof compliance test result is determined. If the risk assessment level is high risk and the weighted consistency score is greater than a first preset threshold, the target explosion-proof compliance test result is determined to be non-compliant. If the risk assessment level is medium risk and the weighted consistency score is between the first preset threshold and a second preset threshold, the target explosion-proof compliance test result is determined to be pending review. If the risk assessment level is low risk and the weighted consistency score is less than a second preset threshold, the target explosion-proof compliance test result is determined to be compliant. Wherein, the first preset threshold value is greater than the second preset threshold value, and both the first preset threshold value and the second preset threshold value are positive numbers.

8. An explosion-proof compliance testing device for detectors based on the Internet of Things, characterized in that, The device includes: The first unit is used to preprocess the static feature data and dynamic feature data of the detector respectively to obtain the static feature vector corresponding to the static feature data and the dynamic feature vector corresponding to the dynamic feature data. The second unit is used to perform deep feature encoding processing on the static feature vector and the dynamic feature vector to obtain a deep feature encoding vector, wherein the deep feature encoding vector is used to characterize the deep hidden danger features after the interaction and fusion of the static attributes and dynamic operating state of the detector. The third unit is used to classify and predict the explosion-proof hazard parts of the detector based on the deep feature encoding vector, and obtain the explosion-proof hazard part prediction probability vector of the detector, wherein the explosion-proof hazard part prediction probability vector of the detector is used to characterize the prediction probability of each explosion-proof hazard part of the detector. The fourth unit is used to determine the risk assessment result of the explosion-proof hazard location of the detector based on the predicted probability vector of the explosion-proof hazard location of the detector and the dynamic feature data of the detector. The fifth unit is used to obtain the target explosion-proof compliance test result of the detector based on the risk assessment results of the explosion-proof hazard parts of the detector and the historical explosion-proof compliance test record data of the detector.

9. A terminal, characterized in that, The method includes a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to perform the steps of the method as described in any one of claims 1 to 7.

10. A medium, characterized in that, The medium stores a plurality of instructions adapted for loading by a processor to execute the steps of the method according to any one of claims 1 to 7.

Images