Multi-dimensional perception oil tank leakage risk early warning method

Abstract

The invention belongs to the technical field of oil tank safety monitoring, and particularly discloses and provides a multi-dimensional perception oil tank leakage risk early warning method, which can simultaneously cover leakage real-time parameters and tank bottom corrosion states through a dual-network architecture, traces leakage risk causes from the source, and improves the safety of oil tank safety monitoring. The problems that the monitoring dimension is single, and the corrosion-leakage incidence relation is difficult to establish due to the fact that leakage sensing data and tank bottom corrosion state data cannot be synchronously obtained are solved. Through advanced identification of a tank bottom corrosion high-risk area and cooperative verification of leakage abnormal parameters, a corrosion correlation type leakage risk can be pre-judged in advance, the problems that early warning is delayed and an alarm can only be triggered after leakage occurs are solved, and the accident expansion probability is reduced; the comprehensive early warning information including the risk source coordinate, the risk source type, the leakage diffusion direction and the leakage propagation intensity trend is constructed, so that field personnel can be guided to formulate a targeted scheme, and the disposal efficiency and accuracy are improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of oil tank safety monitoring, and particularly relates to a multi-dimensional perception oil tank leakage risk early warning method. BACKGROUND

[0002] As the core equipment for storing flammable, explosive, toxic and hazardous media in the fields of petroleum and chemical industry, the operation safety of oil tank is directly related to the safety of personnel life, enterprise property and the stability of surrounding ecological environment. In the long-term service process of oil tank, corrosion hazards are prone to occur in the tank bottom due to medium corrosion, weld aging and other factors. If not found in time, it is easy to develop into a leakage accident, which not only may cause major safety accidents such as fire and explosion, but also may cause irreversible environmental damage such as soil pollution and groundwater pollution. Therefore, timely and accurate early warning of oil tank leakage risk is a key link to ensure the safe and stable operation of tank area.

[0003] The multi-dimensional perception oil tank leakage risk early warning technology can break through the limitations of traditional single-dimensional monitoring by integrating multi-dimensional monitoring data and realizing collaborative analysis of leakage parameters and equipment corrosion state, thereby providing technical support for early identification, accurate determination and efficient disposal of oil tank leakage risk, and having important practical significance for improving the safety prevention and control capability of tank area, reducing the accident rate and reducing environmental damage.

[0004] However, the current existing oil tank leakage risk early warning technology still has many limitations and cannot meet the high-precision and high-timeliness safety monitoring needs of tank area. The specific problems are as follows: (1) The monitoring dimension in the existing oil tank leakage monitoring scheme is single, and the corrosion-leakage correlation cannot be established, which leads to the inability to identify the causal relationship between corrosion hazards and leakage risks from the source and to predict potential risks in advance.

[0005] (2) The existing oil tank leakage monitoring scheme identifies abnormalities and triggers alarms only after the leakage medium has spread to the sensing node, lacks early monitoring of high-risk areas of tank bottom corrosion, and leads to early warning lagging behind the risk development process, which increases the difficulty of accident control for field personnel.

[0006] (3) The early warning information in the existing oil tank leakage monitoring scheme mainly includes a single prompt of leakage parameter abnormalities, lacks comprehensive determination of risk source coordinates, risk types, leakage diffusion direction and intensity trend, and field personnel cannot clearly determine the core position and development trend of risks, which may lead to insufficient pertinence of disposal measures and resource waste. SUMMARY

[0007] In view of this, in order to solve the problems proposed in the background art, a multi-dimensional perception oil tank leakage risk early warning method is provided.

[0008] The objective of this invention can be achieved through the following technical solution: This invention provides a multi-dimensional sensing method for early warning of oil tank leakage risks, including: S1, deploying a biomimetic sensing network in the body of several oil tanks and the surrounding space of the target monitoring area, and simultaneously deploying a magnetic marker imaging network on the bottom plate of the several oil tanks.

[0009] S2. The biomimetic sensor network collects leakage sensing data from each node in real time, while the magnetic marker imaging network simultaneously collects the adsorption state distribution map characterizing the corrosion state of the tank bottom plate.

[0010] S3. Perform dynamic alarm path analysis based on the leak detection data; perform tank bottom corrosion risk assessment based on the adsorption state distribution map.

[0011] The specific steps of S3 include: S31, when an abnormal parameter is identified in the leakage sensing data of any node in the bionic sensing network, the bionic communication rules are activated, and a dynamic alarm path is formed through the force transmission and sensitivity enhancement mechanism between nodes.

[0012] S32. Simultaneously, based on the adsorption state distribution map, identify high-risk areas for bottom corrosion and generate bottom corrosion risk assessment results.

[0013] S4. Based on the dynamic alarm path and the tank bottom corrosion risk assessment results, generate comprehensive early warning information for oil tank leakage and provide feedback.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention can simultaneously cover real-time leakage parameters and tank bottom corrosion status through a dual network architecture, making the monitoring data more comprehensive and the correlation higher, which is conducive to tracing the cause of leakage risk from the source and solving the problem of difficulty in establishing corrosion-leakage correlation caused by the single monitoring dimension and the inability to simultaneously obtain leakage perception data and tank bottom corrosion status data.

[0015] 2. This invention helps to predict corrosion-related leakage risks in advance by identifying high-risk areas of tank bottom corrosion in advance and verifying abnormal leakage parameters. It solves the problem of delayed early warning and only triggering alarms after leakage occurs, thus reducing the probability of accident escalation.

[0016] 3. This invention constructs comprehensive early warning information that includes the coordinates of the risk source, the type of the risk source, the direction of leakage and diffusion, and the trend of leakage propagation intensity. This helps guide on-site personnel to formulate targeted solutions and improves the efficiency and accuracy of handling. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram illustrating the implementation steps of the method of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 As shown, the present invention provides a multi-dimensional sensing method for early warning of oil tank leakage risks. The specific steps are as follows: S1, deploy a biomimetic sensor network in the main body of several oil tanks and the surrounding space of the target monitoring area, and at the same time deploy a magnetic marker imaging network on the bottom plate of the several oil tanks.

[0021] In one feasible embodiment of the present invention, the specific process of deploying the biomimetic sensor network includes: surveying the basic information of the oil tank and the surrounding environment of the target monitoring area, and determining the deployment parameters of the biomimetic sensor network and the magnetic marker imaging network accordingly.

[0022] It should be noted that the basic information of the oil tanks includes the number of oil tanks in the target monitoring area, as well as the volume of each oil tank, material such as carbon steel or stainless steel, tank type such as vertical or horizontal, bottom plate area, thickness, recorded bottom plate weld location, and previous corrosion maintenance records.

[0023] The surrounding environmental information includes the spatial range around each oil tank within the target monitoring area, such as the distance between tank areas and road locations, ambient temperature and humidity, prevailing wind direction, and the presence of interference sources such as strong electromagnetic fields and vibrating equipment, to avoid interference with subsequent sensing signals.

[0024] The deployment parameters of the biomimetic sensor network include node density such as 10m. 2 / node, high-risk areas encrypted to 5m 2 / and the types of monitoring parameters such as oil and gas concentration, temperature and humidity, and micro-vibration.

[0025] The deployment parameters of the magnetic marker imaging network include the marker coverage area, which must cover the entire bottom plate of the tank, including edge blind areas, and the field of view angle of the imaging unit to avoid obstruction.

[0026] In the target monitoring area, a biomimetic sensor network is deployed in the form of several oil tank bodies and surrounding spaces according to the node sequence of first the body and then the surroundings, and first the key points and then the regular ones. Specifically, it includes: selecting the corresponding sensor for each node according to the medium of the several oil tanks, deploying nodes for the oil tank body according to the principle of vertical layering and horizontal uniformity, and deploying nodes for the surrounding environment according to the principle of equal distance.

[0027] Specifically, the specific process of deploying the biomimetic sensor network is as follows: (1) Sensor node selection and preprocessing: Select suitable nodes according to the oil tank medium such as gasoline, diesel and crude oil. For example, to monitor gasoline leakage, a high-sensitivity oil and gas concentration sensor is required, and to monitor tank wall deformation, a micro-vibration sensor is required. At the same time, anti-corrosion treatment is carried out on the nodes, such as coating the outer shell with polytetrafluoroethylene to avoid oil and gas corrosion.

[0028] (2) Tank body node layout: For the tank wall: Layout according to the principle of vertical layering + horizontal uniformity. For example, for vertical tanks, one layer is set every 2m from the top to the bottom, and 3-5 nodes are evenly distributed along the circumference of each layer, focusing on covering the junction of the tank wall and the bottom plate, and the weld seam, which are high-risk leakage areas. For the tank top: Layout 2-3 additional nodes around the breather valve, manhole and other easily leaking parts to monitor the escape of oil and gas.

[0029] (3) Node layout in the surrounding space: For the ground: a linear node strip is set up within a range of 1-3m around the oil tank, with one node every 1.5m, to monitor the diffusion of the leaked medium to the ground. For the air: if there are supports in the tank area, a horizontal node network is set up at a height of 2m above the ground to capture the evaporation trajectory of oil and gas in the air.

[0030] (4) Network networking and power supply: All nodes are connected through wireless self-organizing network technology such as LoRa to ensure stable data transmission; dual power supply of solar energy and lithium battery is adopted to avoid the wiring risks of wired power supply, such as sparks caused by aging wires.

[0031] In one feasible embodiment of the present invention, the specific process of deploying the magnetic marker imaging network includes: pre-processing the bottom plates of several oil tanks belonging to the target monitoring area by cleaning the oil stains and rust on the surface of the bottom plate and smoothing the local protrusions or depressions to ensure that the markers can fit tightly against the bottom plate and avoid misjudgment of the adsorption state due to gaps.

[0032] It should be noted that the dispersion contains uniformly distributed magnetic marker particles, such as nanocrystalline alloy magnetic sheets, with a diameter of 5-10 mm.

[0033] The selected magnetic markers are arranged in a grid array on the bottom plates of several oil tank bodies, and are fixed by strong magnetic adsorption and high-temperature resistant adhesive. The coordinates of each magnetic marker are recorded after it is pasted, as shown on the bottom plate of tank 1. .

[0034] For example, markings are placed in regular areas at 20cm×20cm intervals, and in key areas such as base plate welds and previously corroded areas, markings are placed at 10cm×10cm intervals.

[0035] Above the bottom plate of the oil tank, such as the internal support beam, a magnetic imaging sensor array, such as a magnetic resonance imaging unit, is arranged in a ring to ensure that the field of view of each magnetic imaging sensor can cover several, such as 20-30, magnetic markers within the corresponding range, and that the fields of view of all magnetic imaging sensors have no overlap or blind spots.

[0036] S2. The biomimetic sensor network collects leakage sensing data from each node in real time, while the magnetic marker imaging network simultaneously collects the adsorption state distribution map characterizing the corrosion state of the tank bottom plate.

[0037] In one feasible embodiment of the present invention, the specific acquisition process of the adsorption state distribution map characterizing the corrosion state of the tank bottom plate includes: activating the excitation field generating device deployed outside the plurality of oil tanks, applying the excitation field according to the preset frequency and intensity, synchronously acquiring the magnetic field signal detected by the magnetic imaging sensor array in the magnetic marker imaging network, and storing the acquired signal in a time sequence.

[0038] The acquired signal is filtered to remove noise interference, and the filtered signal is recorded as the spatial distribution signal.

[0039] The spatial distribution signal is input into a preset inversion algorithm to calculate the initial adsorption state distribution map, which contains adsorption intensity values ​​of multiple partitions.

[0040] It should be noted that the goal of the inversion algorithm is to deduce the adsorption density distribution of magnetic markers inside the tank bottom based on the total magnetic field distribution measured externally. Specifically, it employs specialized inversion algorithms such as Tikhonov regularization and Bayesian inversion algorithms, combined with the physical model and constraints of the oil tank body, to calculate the magnetic source distribution at the tank bottom that is most likely to generate the observed magnetic field.

[0041] The initial adsorption state distribution map is smoothed to eliminate local anomalies, resulting in an adsorption state distribution map characterizing the corrosion state of the tank bottom plate.

[0042] This invention, through a dual-network architecture, can simultaneously cover real-time leakage parameters and tank bottom corrosion status, resulting in more comprehensive and correlated monitoring data. This facilitates tracing the causes of leakage risks from their source and solves the problem of difficulty in establishing a corrosion-leakage correlation caused by a single monitoring dimension and the inability to simultaneously acquire leakage perception data and tank bottom corrosion status data.

[0043] S3. Perform dynamic alarm path analysis based on the leak detection data; perform tank bottom corrosion risk assessment based on the adsorption state distribution map.

[0044] The specific steps of S3 include: S31, when an abnormal parameter is identified in the leakage sensing data of any node in the bionic sensing network, the bionic communication rules are activated, and a dynamic alarm path is formed through the force transmission and sensitivity enhancement mechanism between nodes.

[0045] In one feasible embodiment of the present invention, the specific identification process of abnormal parameters in the leakage sensing data of any node in the biomimetic sensor network includes: comparing each parameter of the leakage sensing data of each node in the biomimetic sensor network with its corresponding safety threshold, filtering out each parameter of the leakage sensing data of each node in the biomimetic sensor network that exceeds the corresponding safety threshold, and recording them as the marked abnormal parameters of each node.

[0046] When the abnormality verification conditions are met, the abnormal parameter is marked as an abnormal parameter. The specific contents of the abnormality verification conditions include: (1) the abnormal parameter is continuously abnormal within the set sampling period.

[0047] For example, the set sampling period can be 3 sampling periods, and the sampling period can be 1 minute, that is, the abnormal parameter is marked as abnormal for 3 minutes to exclude the instantaneous over-threshold caused by electromagnetic interference of the sensor, such as pulse-type abnormality within 1 second.

[0048] (2) The node corresponding to the marked abnormal parameter has at least two marked abnormal parameters or the marked abnormal parameter exceeds the predefined percentage of its corresponding security threshold.

[0049] In a specific example, if the abnormal parameter is oil and gas concentration, and this parameter continuously exceeds the corresponding safety threshold within a 3-minute monitoring cycle, and there are also abnormal parameters for temperature and humidity at the node where it is located, then the oil and gas concentration parameter is recorded as an abnormal parameter.

[0050] For example, the predefined percentage of the safety threshold can be: 40% for gasoline medium and 60% for diesel medium.

[0051] This is used to identify anomalous parameters in the leaked sensing data of any node in a biomimetic sensor network.

[0052] It should be added that when there are abnormal parameters in the leakage sensing data of a node in the bionic sensor network, the node will immediately send an abnormal trigger signal to the coordination node in the network, such as setting up a coordination node for each oil tank area. The signal includes the node ID, abnormal parameter type, abnormal value, and trigger timestamp.

[0053] The coordinating node is responsible for coordinating communication.

[0054] In one feasible embodiment of the present invention, the specific content of the biomimetic communication rules includes: the coordinating node in the biomimetic sensor network calls the node topology relation library to identify the first-level neighbor nodes and second-level neighbor nodes of the initial abnormal node, forming a node layer of abnormal core area-buffer zone, and at the same time, the coordinating node sends a communication activation instruction to all identified neighbor nodes.

[0055] It should be added that the node topology database is pre-built during the network setup in step S1, and is used to record the physical location of each node and the list of adjacent nodes.

[0056] It should be noted that the communication activation command includes the initial abnormal node coordinates and key parameters to be monitored, aiming to activate the low-latency communication link between nodes, enabling real-time downlink communication response, replacing conventional timed polling communication, and ensuring rapid signal transmission.

[0057] It should also be noted that the initial abnormal node is the signal source node, which continuously sends real-time change data of abnormal parameters to neighboring nodes; the first-level neighbor node is the signal transmission node, which is dynamically determined based on signal strength and quality, and its initial physical distance from the initial abnormal node is ≤10m. The first-level neighbor node prioritizes receiving and forwarding signal source data, and at the same time activates its own sensitivity enhancement; the second-level neighbor node is the signal monitoring node, which is dynamically determined based on signal strength and quality, and its initial physical distance from the first-level neighbor node is ≤20m. The second-level neighbor node increases the sampling frequency, monitors whether secondary anomalies occur, and avoids the risk of missed reports spreading.

[0058] After receiving the signal source data transmitted by the initial abnormal node, the first-level neighbor node and the second-level neighbor node calculate the weight of the signal they transmit to their neighboring nodes. When the weight is greater than a preset weight limit, the neighboring node transmits the signal to its neighboring node.

[0059] It should be noted that the specific process of transmitting the signal weight to adjacent nodes includes: (1) calculating the distance weight, the specific formula of which is: This indicates that the closer the physical distance to the initial abnormal node, the higher the weight.

[0060] (2) Calculate the signal strength weight, the specific formula is as follows: This indicates that the higher the strength of the abnormal signal received by the node, the higher its weight.

[0061] (3) Calculate the health weight, the specific formula is as follows: This indicates that the higher the node's health, the higher its weight.

[0062] The weights for transmitting signals to neighboring nodes are obtained by weighting and summing the distance weight, signal strength weight, and health weight.

[0063] It should also be noted that the weight of the signal transmitted to the adjacent node is in the range of 0-1, and the larger the weight, the higher the priority.

[0064] For example, assuming the preset weight limit is 0.6, the first-level neighbor nodes B and C of the initial abnormal node A have a weight of 0.85 for B to transmit signals to neighboring nodes and a weight of 0.55 for C to transmit signals to neighboring nodes. Thus, only node B transmits signals to its neighbor node D, while node C does not transmit signals to its neighbor nodes.

[0065] Based on the node layers, a node sensitivity grading enhancement strategy is implemented.

[0066] In one specific example, for the initial abnormal node, the sampling frequency is increased from the usual 5 minutes / time to 10 seconds / time, and the monitoring accuracy of abnormal parameters is doubled. For example, the resolution of oil and gas concentration detection is increased from 1ppm to 0.5ppm. At the same time, multi-parameter linkage monitoring is enabled, such as monitoring the capacitance change of liquid leakage while detecting oil and gas concentration.

[0067] For primary neighbor nodes, the sampling frequency is increased to 30 seconds / time, and the alarm threshold for abnormal parameters is reduced by 20%, such as the original threshold of 100ppm, which is reduced to 80ppm, so as to detect secondary anomalies in advance.

[0068] For secondary neighbor nodes, the sampling frequency is increased to once every minute, and signal fluctuation monitoring is enabled. For example, if the rate of change of oil and gas concentration is monitored, and the concentration increases by ≥10ppm within 1 minute, it is immediately marked as a high-risk fluctuation.

[0069] It should also be noted that the enhanced sensitivity nodes will monitor data in real time and feed it back to the coordination node. If a node detects a new anomaly, such as a neighboring node detecting a concentration exceeding 80 ppm, the node will automatically be upgraded to a new signal source node.

[0070] In one feasible embodiment of the present invention, the specific process of forming a dynamic alarm path includes: the coordinating node connects the coordinates of the initial abnormal node, the transmitting node, and the new signal source node in chronological order according to the signal transmission sequence and weight priority, to form an initial dynamic alarm path as follows: Each node includes an anomaly value and a transmission timestamp.

[0071] Invalid information in the initial dynamic alarm path is removed to ensure the accuracy and reliability of the path.

[0072] For example, if a node transmits a signal only once and there is no subsequent abnormal feedback, such as node E not detecting any abnormality after receiving the signal, it is removed from the path.

[0073] After removing invalid information, compare the abnormal parameter change trends of all nodes on the initial dynamic alarm path. For example, if the oil and gas concentrations all show an upward trend, and if the trend of a certain node is opposite to the abnormal parameter change, such as a decrease in concentration, then mark it as an interfering node and adjust the initial dynamic alarm path to bypass that node.

[0074] This is used to obtain the dynamic alarm path.

[0075] It should be noted that the dynamic alarm path includes node coordinates, abnormal parameters, transmission time, and risk level.

[0076] S32. Simultaneously, based on the adsorption state distribution map, identify high-risk areas for bottom corrosion and generate bottom corrosion risk assessment results.

[0077] In one feasible embodiment of the present invention, the specific process of generating the tank bottom corrosion risk assessment result includes: image preprocessing and feature enhancement of the adsorption state distribution map.

[0078] In the enhanced adsorption state distribution map, pixel areas where the adsorption density exceeds the set global high-risk threshold are designated as high-risk areas for tank bottom corrosion.

[0079] Morphological analysis was performed on the high-risk area of ​​tank bottom corrosion. The adsorption density fraction, area fraction and shape risk factor of the high-risk area of ​​tank bottom corrosion were calculated respectively. The adsorption density fraction, area fraction and shape risk factor were weighted and summed to obtain the comprehensive risk score.

[0080] For example, the weight corresponding to the adsorption density fraction can be set to 0.5, the weight corresponding to the area fraction can be set to 0.3, and the weight corresponding to the shape risk factor can be set to 0.2. The weights are based on the contribution rate analysis of corrosion-induced leakage, where adsorption density has the greatest impact on leakage.

[0081] Several high-risk areas of tank bottom corrosion, along with their corresponding adsorption density fraction, area fraction, shape risk factor, and comprehensive risk score, are collectively referred to as the tank bottom corrosion risk assessment results, which are then sent to the monitoring unit in real time.

[0082] This invention helps to predict corrosion-related leakage risks in advance by identifying high-risk areas of tank bottom corrosion in advance and verifying abnormal leakage parameters. It solves the problem of delayed early warning and only triggering alarms after leakage occurs, thus reducing the probability of accident escalation.

[0083] S4. Based on the dynamic alarm path and the tank bottom corrosion risk assessment results, generate comprehensive early warning information for oil tank leakage and provide feedback.

[0084] In one feasible embodiment of the present invention, the comprehensive early warning information for oil tank leakage includes the coordinates of the risk source, the type of the risk source, the direction of leakage diffusion, and the trend of leakage propagation intensity.

[0085] In one feasible embodiment of the present invention, the specific process of generating comprehensive early warning information for oil tank leaks includes: determining the coordinates of the risk source based on the coordinates and signal strength weights of nodes in the dynamic alarm path using a weighted average method. It should be noted that the specific formula for determining the coordinates of the risk source using the weighted average method is as follows: .

[0086] Calculate the spatial distance between the risk source coordinates and the center point of the high-risk corrosion area at the bottom of the tank. Based on this, determine the risk source type. Specifically: if the risk source coordinates are located at the bottom of the tank and the distance from the corrosion area is less than a preset distance value, such as 1m, then the risk type is recorded as corrosion-related risk; if the risk source is located on the tank wall or top of the tank and the corrosion area is directly below it, then the risk type is recorded as corrosion-indirect risk; if the distance between the risk source and the corrosion area is greater than a set distance value, such as 3m, and there is no spatial correlation, then the risk type is recorded as no corrosion risk.

[0087] The direction of leakage diffusion is determined based on the changing trend of node coordinates in the dynamic alarm path. Specifically, if the node coordinates extend from the bottom of the tank to the ground, the direction of leakage diffusion is recorded as seepage to the ground; if the node coordinates extend from the tank wall to the air, the direction of leakage diffusion is recorded as evaporation into the air.

[0088] The signal strength weights of each node in the dynamic alarm path are extracted, and the leakage propagation intensity trend is analyzed. Specifically, if the signal strength weights of each node show an increasing trend over time, the leakage propagation intensity trend is recorded as an increase in leakage intensity and an acceleration of risk spread; if the signal strength weights of each node show a decreasing trend over time, the leakage propagation intensity trend is recorded as a decrease in leakage intensity and a slowdown in spread speed.

[0089] This invention constructs comprehensive early warning information that includes the coordinates of the risk source, the type of the risk source, the direction of leakage and diffusion, and the trend of leakage propagation intensity. This helps guide on-site personnel to develop targeted solutions and improves the efficiency and accuracy of handling.

[0090] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0091] Those skilled in the art will recognize that the algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0092] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0093] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0094] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A multi-dimensional sensing method for early warning of oil tank leakage risks, characterized in that: include: S1. A biomimetic sensor network is deployed in the main body of several oil tanks and the surrounding space of the target monitoring area, and a magnetic marker imaging network is deployed on the bottom plate of the several oil tanks. S2. The biomimetic sensor network collects leakage sensing data from each node in real time, while the magnetic marker imaging network simultaneously collects the adsorption state distribution map characterizing the corrosion state of the tank bottom plate. S3. Perform dynamic alarm path analysis based on the leak detection data; perform tank bottom corrosion risk assessment based on the adsorption state distribution map; The specific steps of S3 include: S31, when an abnormal parameter is identified in the leakage sensing data of any node in the bionic sensing network, the bionic communication rules are activated, and a dynamic alarm path is formed through the force transmission and sensitivity enhancement mechanism between nodes. S32. Simultaneously, based on the adsorption state distribution map, identify high-risk areas for tank bottom corrosion and generate tank bottom corrosion risk assessment results. S4. Based on the dynamic alarm path and the tank bottom corrosion risk assessment results, generate comprehensive early warning information for oil tank leakage and provide feedback.

2. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 1, characterized in that: The specific process of deploying the biomimetic sensor network includes: The basic information of the oil tanks and the surrounding environment in the target monitoring area are surveyed, and the deployment parameters of the bionic sensor network and the magnetic marker imaging network are determined accordingly. In the target monitoring area, a biomimetic sensor network is deployed in the form of several oil tank bodies and surrounding spaces according to the node sequence of first the body and then the surroundings, and first the key points and then the regular ones. Specifically, it includes: selecting the corresponding sensor for each node according to the medium of the several oil tanks, deploying nodes for the oil tank body according to the principle of vertical layering and horizontal uniformity, and deploying nodes for the surrounding environment according to the principle of equal distance.

3. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 1, characterized in that: The specific process of deploying the magnetic marker imaging network includes: Pre-treatment was carried out on the bottom plates of several oil tanks in the target monitoring area to clean the oil and rust on the surface of the bottom plates, and to flatten the local protrusions or depressions to ensure that the markers could fit tightly against the bottom plates. The selected magnetic markers are arranged in a grid array on the bottom plates of several oil tank bodies, and the magnetic markers are fixed with the help of strong magnetic adsorption and high temperature resistant adhesive. The coordinates of each magnetic marker are recorded after it is pasted. A magnetic imaging sensor array is arranged in a ring above the bottom plate of the oil tank to ensure that the field of view of each magnetic imaging sensor can cover several magnetic markers within the corresponding range.

4. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 3, characterized in that: The specific process for collecting the adsorption state distribution map characterizing the corrosion state of the tank bottom plate includes: The excitation field generating device deployed outside the oil tanks is activated, and the excitation field is applied according to the preset frequency and intensity. The magnetic field signal detected by the magnetic imaging sensor array in the magnetic marker imaging network is collected synchronously, and the collected signal is stored in time sequence. The acquired signal is filtered to remove noise interference, and the filtered signal is recorded as the spatial distribution signal. The spatial distribution signal is input into a preset inversion algorithm to calculate an initial adsorption state distribution map, which contains adsorption intensity values ​​of multiple partitions. The initial adsorption state distribution map is smoothed to eliminate local anomalies, resulting in an adsorption state distribution map characterizing the corrosion state of the tank bottom plate.

5. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 1, characterized in that: The specific process for identifying abnormal parameters in the leaked sensing data of any node in the biomimetic sensor network includes: The parameters of the leakage sensing data of each node in the bionic sensor network are compared with their corresponding safety thresholds. The parameters of the leakage sensing data of each node in the bionic sensor network that exceed the corresponding safety thresholds are selected and recorded as the marked abnormal parameters of each node. When the abnormality verification condition is met, the abnormal parameter is marked as an abnormal parameter. The specific contents of the abnormality verification condition include: (1) the abnormal parameter is continuously abnormal within the set sampling period; (2) The node corresponding to the marked abnormal parameter has at least two marked abnormal parameters at the same time or the marked abnormal parameter exceeds the predefined percentage of its corresponding security threshold; This is used to identify anomalous parameters in the leaked sensing data of any node in a biomimetic sensor network.

6. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 5, characterized in that: The specific content of the biomimetic communication rules includes: In the biomimetic sensor network, the coordinating node calls the node topology relation library to identify the first-level and second-level neighbor nodes of the initial abnormal node, forming a node ring of abnormal core area-buffer zone. At the same time, the coordinating node sends a communication activation command to all identified neighbor nodes. After receiving the signal source data transmitted by the initial abnormal node, the first-level neighbor node and the second-level neighbor node calculate the weight of the signal they transmit to the neighboring node. When the weight is greater than the preset weight limit, the neighboring node transmits the signal to its neighboring node. Based on the node layers, a node sensitivity grading enhancement strategy is implemented.

7. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 6, characterized in that: The specific process for forming the dynamic alarm path includes: The coordination node connects the coordinates of the initial abnormal node, the transmission node, and the new signal source node in chronological order according to the signal transmission sequence and weight priority, forming an initial dynamic alarm path. Invalid information in the initial dynamic alarm path is removed to ensure the accuracy and reliability of the path; After removing invalid information, compare the abnormal parameter change trends of all nodes on the initial dynamic alarm path. If the trend of a certain node is opposite to the abnormal parameter change, it is marked as an interference node, and the initial dynamic alarm path is adjusted to bypass the node. This is used to obtain the dynamic alarm path.

8. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 1, characterized in that: The specific process for generating the tank bottom corrosion risk assessment results includes: The adsorption state distribution map is subjected to image preprocessing and feature enhancement; In the enhanced adsorption state distribution map, pixel regions where the adsorption density exceeds the set global high-risk threshold are defined as high-risk areas for tank bottom corrosion. Morphological analysis was performed on the high-risk area of ​​tank bottom corrosion. The adsorption density fraction, area fraction and shape risk factor of the high-risk area of ​​tank bottom corrosion were calculated respectively. The adsorption density fraction, area fraction and shape risk factor were then weighted and summed to obtain the comprehensive risk score. Several high-risk areas of tank bottom corrosion, along with their corresponding adsorption density fraction, area fraction, shape risk factor, and comprehensive risk score, are collectively referred to as the tank bottom corrosion risk assessment results, which are then sent to the monitoring unit in real time.

9. The method for multi-dimensional sensing of oil tank leakage risk early warning according to claim 1, characterized in that: The comprehensive early warning information for oil tank leaks includes the coordinates of the risk source, the type of the risk source, the direction of leak spread, and the trend of leak propagation intensity.

10. The multi-dimensional sensing method for early warning of oil tank leakage risk according to claim 9, characterized in that: The specific process for generating comprehensive early warning information on oil tank leaks includes: Based on the coordinates and signal strength weights of nodes in the dynamic alarm path, the coordinates of the risk source are determined by a weighted average method. Calculate the spatial distance between the risk source coordinates and the center point of the high-risk corrosion area at the bottom of the tank. Based on this, determine the type of risk source: if the risk source coordinates are located at the bottom of the tank and the distance from the corrosion area is less than a preset distance value, then the risk type is recorded as corrosion-related risk; if the risk source is located on the tank wall or top of the tank and the corrosion area is directly below it, then the risk type is recorded as corrosion-indirect risk; if the distance between the risk source and the corrosion area is greater than a set distance value and there is no spatial correlation, then the risk type is recorded as no corrosion risk. The direction of leakage diffusion is determined based on the changing trend of node coordinates in the dynamic alarm path. Specifically, if the node coordinates extend from the bottom of the tank to the ground, the direction of leakage diffusion is recorded as seepage to the ground; if the node coordinates extend from the tank wall to the air, the direction of leakage diffusion is recorded as evaporation into the air. The signal strength weights of each node in the dynamic alarm path are extracted, and the leakage propagation intensity trend is analyzed. Specifically, if the signal strength weights of each node show an increasing trend over time, the leakage propagation intensity trend is recorded as an increase in leakage intensity and an acceleration of risk spread; if the signal strength weights of each node show a decreasing trend over time, the leakage propagation intensity trend is recorded as a decrease in leakage intensity and a slowdown in spread speed.

Images