Multi-sensor collaborative intelligent diagnosis system, method and equipment for leakage of storage tank

Abstract

The invention relates to the technical field of product oil cave depot tank bottom plate leakage online monitoring and diagnosis, and particularly discloses a multi-sensor collaborative storage tank leakage intelligent diagnosis system, method and equipment, and the system comprises a signal collection assembly which comprises a plurality of sensors used for monitoring storage tank information data and an upper computer capable of receiving sensor signals; the software function component comprises a storage tank structure modeling module for establishing a virtual storage tank model according to the storage tank structure data, a regular detection data module for regularly detecting and storing the storage tank data, a real-time monitoring module for monitoring the real-time data of the storage tank, and a leakage diagnosis module for diagnosing the real-time data; the real-time monitoring module comprises a liquid level monitoring module, an oil gas concentration monitoring module and an acoustic emission monitoring module; and the risk early warning model receives the diagnosis result of the leakage diagnosis module and outputs a storage tank leakage diagnosis early warning signal. According to the invention, real-time dynamic monitoring data and regular detection records can be fused to carry out joint probabilistic judgment on the leakage of the storage tank.

Description

Technical Field

[0001] The present application relates to the technical field of online monitoring and diagnosis of bottom plate leakage of finished oil cavern storage tanks, and specifically discloses a multi-sensor collaborative intelligent tank leakage diagnosis system, method and equipment. Background Art

[0002] Refined oil storage tanks are typically atmospheric pressure tanks and are essential energy storage equipment for the petrochemical industry and military logistics units. Compared to conventional surface storage tanks, cavernous refined oil storage tanks offer a more sealed environment and the ability to maintain oil storage for extended periods. The tank floor is the most likely location for leaks. Tank floor leak detection technologies can be categorized into two types: leak detection and leak monitoring, depending on the detection cycle.

[0003] Currently, leak detection is generally conducted during tank overhauls or when a leak is suspected. Traditional detection techniques are divided into two types. The first is open-tank inspection, which involves draining the oil from the tank and cleaning it. Inspectors then enter the tank to conduct inspections. Inspection methods include visual inspection, bubble test-pressure method, bubble test-vacuum method, tracer gas method, penetration test, and magnetic particle inspection. The second is inspection while the oil is stored. This involves using equipment outside the tank to inspect the tank. Key methods include metrology, basic drilling leak detection, basic leak detection layer method, infrared thermal imaging, and acoustic emission testing. Both open-tank inspection and external inspection while the oil is stored rely on methods that can be affected by various factors. For example, manual visual inspection or nondestructive testing after opening a tank can miss tiny cracks or hidden leaks located in complex structures (such as weld roots and tank bottom edge plates). External detection methods such as metrology are susceptible to temperature fluctuations, evaporation losses, and instrument errors. Basic drilling and layer detection methods rely on the probability of leaks reaching specific monitoring points. Infrared thermal imaging and acoustic emission testing have high requirements for leak rate, environmental conditions, and equipment sensitivity, and are prone to false positives or omissions. This uncertainty in accuracy means that even if manpower and resources are invested in regular testing, there is still a possibility of failing to identify real leaks or issuing false alarms, which not only reduces the credibility of test results but also makes decisions and resource allocation based on test results risky. Therefore, the dual deficiencies of regular testing in terms of timeliness and reliability make it a significant shortcoming in effectively preventing and early controlling the risk of tank leakage.

[0004] Traditional leakage monitoring technology relies on pre-embedded components. This method primarily involves deploying sensors during foundation excavation for new tanks or renovations of existing ones. Sensors are embedded in the tank's foundation layer and, by monitoring the oil content in the layer, determine whether the tank is leaking. This approach is completely inapplicable to the vast number of existing, in-service tanks, significantly limiting its scope of application. Even when adopted in new projects, it significantly increases initial construction costs, including the sensors themselves, complex wiring systems, and installation and commissioning fees. More critically, if a sensor fails or performance degrades during tank operation, repair or replacement requires vacating the tank. This forced excavation and repair is costly and seriously compromises the tank's structural safety and normal operation, making it difficult to ensure the reliability and continuity of long-term monitoring. Furthermore, the coverage, sensitivity, and long-term stability of pre-embedded sensors may be limited by factors such as installation location, soil conditions, and material aging. They may not fully capture all potential leaks or even minute leak signals, making their monitoring effectiveness uncertain. Currently, in the open and widely recognized international or national standards and specifications, there is no unified, mature and mandatory standard technology and method system for the real-time, online and systematic leakage monitoring of in-service storage tanks. When storage tank managers face suspected tank leaks, they do not know how to judge and there are no clear judgment indicators. Faced with the blind selection of many detection methods and the lack of a maintenance method decision-making system to support them, some people have to blindly use tank opening inspection, which leads to increased detection costs. Therefore, in view of this, the present invention provides a multi-sensor collaborative storage tank leakage intelligent diagnosis system, method and equipment to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent diagnosis system, method and equipment for combining real-time dynamic monitoring data with periodic inspection records to make joint probabilistic judgments on tank leakage.

[0006] To achieve the above objectives, the basic solution of the present invention provides a multi-sensor coordinated intelligent tank leakage diagnosis system, method and equipment, including: Signal acquisition components, including several sensors for monitoring tank information data and a host computer that can receive sensor signals; The software functional components include a tank structure modeling module that establishes a virtual tank model based on the tank structure data, a periodic detection data module that regularly detects and stores the tank data, a real-time monitoring module that monitors the real-time data of the tank, and a leakage diagnosis module that diagnoses the real-time data. The real-time monitoring module includes a liquid level monitoring module, an oil and gas concentration monitoring module, and an acoustic emission monitoring module. The risk warning model receives the diagnosis results of the leakage diagnosis module and outputs a tank leakage diagnosis warning signal.

[0007] Furthermore, the tank structure modeling module presets a relationship function between the tank liquid level and the tank body deformation and a tank volatilization function related to the temperature and the liquid level. The liquid level monitoring module obtains real-time liquid level monitoring data and temperature of the tank through a signal acquisition component. The leakage diagnosis module synchronously calculates the tank foundation deformation liquid level change, the oil and gas volatilization liquid level change, the oil and gas volatilization liquid level change error, the liquid level monitor precision parameter-determined measurement error, and the real-time liquid level change. When the real-time liquid level change is greater than the sum of the tank foundation deformation liquid level change, the oil and gas volatilization liquid level change, the oil and gas volatilization liquid level change error, and the liquid level monitor precision parameter-determined measurement error, the tank is judged to be leaking.

[0008] Furthermore, an oil and gas concentration threshold is preset in the tank structure modeling module, the oil and gas concentration monitoring module collects real-time concentration data inside and outside the tank through a signal acquisition component, and the leakage diagnosis module compares and calculates the collected real-time concentration data inside and outside the tank with the oil and gas concentration threshold and determines whether the tank is leaking.

[0009] Furthermore, in the signal acquisition component, sensor arrays located inside and outside the tank are deployed in a tank gradient. The oil and gas concentration monitoring module obtains the monitoring data of the sensor array and the leakage diagnosis module determines whether it is abnormal. The risk warning model sorts the sensor arrays corresponding to the abnormal data in time response order, obtains the sensor array with the earliest time response, and marks the tank position corresponding to the sensor array.

[0010] Furthermore, a signal feature library of different types is constructed in the tank structure modeling module, the acoustic emission monitoring module collects acoustic emission signals through a signal acquisition component, and the leakage diagnosis module performs time domain analysis and frequency domain analysis on the collected acoustic emission signals, and combines the time difference positioning method with the depth feature matching to determine whether leakage occurs and the leakage area.

[0011] Furthermore, in the signal acquisition component, 8 acoustic emission sensors are evenly deployed circumferentially at a height of 20 cm on the bottom plate of the tank to form a circular array to collect acoustic emission signals.

[0012] Furthermore, the risk warning model determines the number of modules with abnormal data in the liquid level monitoring module, the oil and gas concentration monitoring module, and the acoustic emission monitoring module, and issues alarm prompts according to the level.

[0013] Based on the same inventive concept, the present invention provides a multi-sensor collaborative intelligent tank leakage diagnosis method, including using the above-mentioned intelligent tank leakage diagnosis system to perform leakage diagnosis on the tank.

[0014] Furthermore, the steps for using the above-mentioned intelligent tank leakage diagnosis system to diagnose tank leakage are as follows: Acquire real-time tank information data through signal acquisition components; The regular detection data module obtains real-time tank information data from the signal acquisition component and transmits it to the leakage diagnosis module; The leakage diagnosis module transmits the diagnosis results to the risk warning model; The risk warning model outputs a tank leakage diagnosis warning signal.

[0015] Based on the same inventive concept, the present invention provides a multi-sensor collaborative intelligent tank leakage diagnosis device, in which the above-mentioned intelligent tank leakage diagnosis system is integrated and the above-mentioned intelligent tank leakage diagnosis method is executed.

[0016] The principle and effect of this solution are: The present invention establishes a comprehensive leakage detection and diagnosis system by integrating the liquid level value of the medium inside the storage tank, the index value of the medium inside and outside the storage tank, and the leakage signal value of the bottom plate of the storage tank, and integrates the regular detection data for intelligent judgment. It can be used simultaneously to make a comprehensive judgment, improve the accuracy of leakage judgment, and improve the safety of storage tank use. Any single module can also be selected for use. For suspected leaking storage tanks, without opening the tank, the internal liquid level can be judged, the external oil and gas concentration can be judged, and finally the acoustic emission leakage detection can be performed in this order. The detection methods are progressive layer by layer to form a set of leakage judgment and maintenance decision-making system. It avoids the human and material costs caused by managers blindly opening the tank for detection or excessive detection, and improves the utilization rate of the storage tank. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A schematic diagram of a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown; Figure 2 A schematic diagram of a signal acquisition component in a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown; Figure 3 A schematic diagram of a storage tank installation in a multi-sensor collaborative storage tank leakage intelligent diagnosis system proposed in an embodiment of the present application is shown; Figure 4 A schematic diagram of sensor installation in a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown; Figure 5A schematic diagram of early warning level assessment in a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown; Figure 6 A schematic diagram of leak location determination in a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown; Figure 7 A flowchart of a multi-sensor collaborative intelligent tank leakage diagnosis method proposed in an embodiment of the present application is shown; Figure 8 A flowchart of early warning level assessment in a multi-sensor collaborative intelligent tank leakage diagnosis system proposed in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0019] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0020] The figure marks in the drawings of the specification include: storage tank 1, sensor 2, liquid level sensor 201, first layer oil and gas concentration sensor 202, second layer oil and gas concentration sensor 203, third layer oil and gas concentration sensor 204, acoustic emission sensor 205, amplifier 3, signal transmission line 4, signal collector 5, host computer 6.

[0021] A multi-sensor collaborative intelligent tank leakage diagnosis system, implementing e.g. Figure 1 As shown, it includes a signal acquisition component for monitoring information data of the storage tank 1, a software function module for receiving signals from the signal acquisition component and performing signal processing and diagnosis, and a risk warning model for receiving the diagnosis results of the leakage diagnosis module and outputting a leakage diagnosis warning signal of the storage tank 1.

[0022] Among them, such as Figure 2 and Figure 3As shown, the signal acquisition system includes several sensors 2 for monitoring information data of the storage tank 1, a signal transmission line 4 for transmitting signals, an amplifier 3 for amplifying the signals, a signal collector 5 for converting analog signals into digital signals, and a host computer 6 for receiving the converted signals. The sensor 2 types include acoustic emission sensors 205, liquid level sensors 201, and oil and gas concentration sensors 2. The acoustic emission sensor 205 uses a piezoelectric sensor 2 with a center frequency of 40kHz and a frequency range of 15kHz to 70kHz, which is used to convert acoustic emission signals into electrical signals. The liquid level sensor 201 uses the MS350V model with an accuracy of ±1mm and can accurately measure the liquid level. The oil and gas concentration sensor 2 uses a PID gas sensor 2 with a measurement range of 0-200ppm; the signal collector 5 uses a SAEPA2 acoustic emission preamplifier 3 to amplify and convert analog signals into digital signals. The industrial computer uses a high-speed 48-channel acquisition card combination of DSP+FPGA to modulate and process the signals; the host computer 6 uses a computer with 16G memory and an i7 processor to display signal acquisition information and fault evaluation; the signal transmission line 4 uses a coaxial cable.

[0023] The software functional modules include a tank structure modeling module that establishes a virtual tank model based on the structural data of tank 1, a periodic detection data module that regularly detects and stores the data of tank 1, a real-time monitoring module that monitors the real-time data of tank 1, and a leakage diagnosis module that diagnoses the real-time data. The real-time monitoring module includes a liquid level monitoring module, an oil and gas concentration monitoring module, and an acoustic emission monitoring module.

[0024] The tank structure modeling module is used to establish a virtual tank model based on the tank structure data; the regular inspection data module is used to enter the regular inspection data into the tank basic data management and build a basic management database for leakage risk control points; the three types of real-time monitoring modules are used to process, display and store online monitoring data, including the liquid level monitoring data, oil and gas concentration monitoring data and acoustic emission monitoring data of tank 1; the leakage diagnosis module uses the risk warning model to diagnose the three types of real-time monitoring indicators.

[0025] The liquid level monitoring and leakage determination process is implemented through a dynamic compensation mechanism. The tank structure modeling module pre-enters the relationship function f1 between the tank level and tank deformation, as well as the tank volatility function f2, which is related to temperature and liquid level. Continuously collected tank data is then subjected to regression analysis to adjust the relevant functional relationships. After acquiring real-time liquid level monitoring data, temperature, and other parameters, two key compensation quantities are simultaneously calculated: the tank foundation deformation liquid level change Δ1 and the oil and gas volatility liquid level change Δ2. The oil and gas volatility liquid level change error is μ1 (0.5mm). The real-time liquid level change Δ0 is also calculated simultaneously, and the liquid level monitor accuracy parameters determine the measurement error μ2 (±1mm).

[0026] Specifically, the relationship function f1 between the tank liquid level and the tank deformation is expressed as follows: Where: ρ is the density of the liquid in the tank (kg / cm 2 ); R is the basic radius or average radius of the tank (cm); E is the elastic modulus of the tank material (kg / cm 2 ); δ is the tank wall thickness of the studied section (cm); h is the liquid level height in the tank (cm).

[0027] The expression of the tank volatility function f2 related to temperature and liquid level is as follows: Where: KE is the oil quality coefficient, which is 24 for gasoline and 14 for crude oil; Pa is the local atmospheric pressure, mmHg; Py is the true vapor pressure at the oil body temperature; R is the radius of the tank (m); H is the height of the tank, including the equivalent height of the tank top (m); h is the liquid level in the tank (m); ΔT is the annual average value of daily atmospheric temperature change (℃); Kp is the coating factor or paint coefficient; C is the tank correction factor; π is the circumference of a circle, which is 3.1415926.

[0028] When the measured liquid level drop value Δ0 exceeds the sum of the oil and gas volatilization level change and the deformation compensation amount (that is, the judgment formula Δ0 > Δ1+ Δ2+ μ1 + μ2 is satisfied), the tank leakage warning is immediately triggered and transmitted to the monitoring center.

[0029] like Figure 4 As shown, the oil and gas concentration monitoring module utilizes a three-layer gradient array of oil and gas concentration sensors 2 deployed in the storage tank 1 and an operating room sensor 2 to identify and locate leaks. Oil and gas concentration detection holes are manually drilled at various locations in the base layer of the storage tank 1. A first layer of oil and gas concentration sensors 202 are installed in each monitoring hole. A second layer of oil and gas concentration sensors 203 are installed at various locations 50 cm above the bottom plate of the first outer wall panel. A third layer of oil and gas concentration sensors 204 are installed at various locations 100 cm above the first outer wall panel. A reference sensor 2 is also installed above the operating room pipe. Real-time data from the storage tanks is collected to generate baseline concentration data, representing the normal concentration range. The alarm time Tx for the oil and gas concentration detected by the oil and gas concentration sensor 2X to reach an abnormally high value is used. The presence and location of a leak are preliminarily determined based on the sensor alarm time sequence. Leakage is further determined by monitoring the concentration value curve of the sensor.

[0030] For example: if the abnormal alarm time T of the oil and gas concentration sensor 2 in the tank room 1 is earlier than the abnormal alarm time T of the oil and gas concentration sensor 2 in the operating room, it is judged as a suspected leak in the storage tank 1. Further, the abnormal alarm time Ta of the first-layer oil and gas concentration sensor 202 in the tank room is earlier than the abnormal alarm time Tb of the second-layer oil and gas concentration sensor 203 and the abnormal alarm time Tc of the third-layer oil and gas concentration sensor 204, it is judged as a leak in the tank bottom plate. Furthermore, the abnormal alarm time Ta1 of the first-layer oil and gas concentration sensor 202 in the aX area is earlier than the abnormal alarm time of other first-layer oil and gas sensors 2, then the leakage location is judged to be the sensor 2 aX area; if the oil and gas concentration value obtained by the oil and gas sensor 2 remains unchanged after increasing, it is judged to be a leak. If it continues to increase and then decreases, it is judged to be a false leak.

[0031] The acoustic emission monitoring module deploys eight acoustic emission sensors 205 evenly distributed around the circumference of tank 1, forming a circular array at a height of 20 cm from the bottom plate. Initially, a basic environmental signal database is constructed by continuously collecting ambient background noise. Frequency domain analysis is then performed on various corrosion and leakage signals. Fast Fourier transform (FFT) is used to generate spectrograms, identifying the energy proportions of different signal characteristic frequency bands and establishing a signature library for different signal types. During real-time monitoring, the collected acoustic emission signals undergo both time and frequency domain analysis. Time domain analysis primarily monitors the effective mean square (RMS) voltage and average signal level (ASL) values. Spectral analysis utilizes a convolutional neural network to extract localized mutation patterns in the waveform. A frequency-domain attention mechanism is used to automatically weight key frequency bands. A feature fusion layer correlates the time and frequency domain information, enabling accurate classification of leaks, corrosion, and other signal types. Finally, a time-difference localization method is combined with deep feature matching to determine the presence and location of leaks.

[0032] Comprehensive diagnosis and rating of tank bottom leakage: see attached Figure 5 , a comprehensive rating is made based on the leakage judgment results obtained by each of the three independent monitoring modules. If the summary shows that only one of the three methods is a leak, the system will ultimately rate it as Level I, prompting a Level I alarm; if the summary shows that two methods are both leaks, the system will ultimately rate it as Level II, prompting a Level II alarm; and when all three monitoring methods consistently judge it as a leak, the system will rate it as the highest level, Level III, prompting a Level III alarm.

[0033] Comprehensive diagnosis of tank bottom leakage location: This function uses the regular detection of magnetic flux leakage detection data module, oil and gas concentration monitoring module, and acoustic emission leakage monitoring module.

[0034] See attached Figure 6The tank floor is divided into eight circumferential zones. Based on regular inspection results, the magnetic flux leakage detection module classifies the eight zones into four risk levels: zero, I, II, and III. Level III is the highest risk. The magnetic flux leakage module can only classify one zone as Level III. The oil and gas concentration determination module classifies the localized zone as Level III, while the remaining zones are rated zero. The acoustic emission monitoring module classifies the localized zone as Level III, while the remaining zones are rated zero. Based on early corrosion signals, the acoustic emission module classifies the eight zones into four risk levels: zero, I, II, and III. Level III is the highest risk. The regional risk value is calculated by taking the average of the module levels for that zone. For zones with three levels, the regional position-related weight coefficient is calculated as follows: the distances from the other three-level zones to the middle of the three-level zone are 1, 0.8, 0.6, 0.4, 0.2, and 0, respectively. The final fault value for that zone is the regional risk value multiplied by the regional position-related weight coefficient.

[0035] Based on the same inventive concept, this embodiment provides a multi-sensor collaborative intelligent diagnosis method for tank leakage, such as Figure 7 As shown, the steps are as follows: Step S1, inputting regular inspection data, which includes basic information data of the storage tank 1, used for establishing the storage tank model, including various structural dimensions, materials, construction years, etc. of the storage tank 1; Step S2 involves liquid level monitoring, oil and gas concentration monitoring, and acoustic emission monitoring. The system records the tank's deformation and liquid level change Δ1, the oil and gas volatilization volume Δ2 and its error μ1, and the liquid level monitoring accuracy μ2. These values ​​are compared with the measured liquid level drop Δ0. If the measured drop Δ0 > Δ1 + Δ2 + μ1 + μ2, a ​​leak is initially determined; otherwise, no leak is detected. For oil and gas concentration monitoring, data is collected under non-process operating conditions to generate baseline concentration data T0. Oil and gas concentration sensors 2 are installed in three layers, with eight sensors per layer and one sensor installed on the pipeline in the operating room, for a total of 25 sensors. Real-time oil and gas concentration data is collected from all sensors: Ta1-8, Tb1-8, Tc1-8, and Td. All collected concentration data is compared with the baseline concentration data. If any concentration exceeds the baseline concentration data, a leak is initially determined; otherwise, no leak is detected. The data identified as leaks is then compared again, and the leak location is preliminarily determined based on the order of concentration. Finally, the trend of oil and gas concentration changes is used to determine whether it is a false leak. If the oil and gas concentration continues to rise and then stabilizes, it is considered a leak; otherwise, it is a false leak. Acoustic emission monitoring preprocesses the waveform of the real-time acoustic emission signal, then performs time domain analysis and spectrum analysis on it to extract core parameters: effective voltage value (reflecting signal strength), average signal level (representing energy mean), arrival time (the moment when the signal triggers the threshold), etc. A fast Fourier transform (FFT) is used to generate a spectrum diagram to identify the energy proportion of the characteristic frequency band of 20-150kHz (>60% is considered a leak-related frequency band). A pre-trained neural network model is established, and the system deeply integrates the waveform's time-domain features with the spectral energy distribution characteristics for intelligent analysis. First, a convolutional neural network is used to extract local mutation patterns in the waveform, while a frequency-domain attention mechanism is used to automatically weight key frequency bands (for example, suppressing low-frequency rain noise interference in heavy rain conditions and focusing on the 20-150kHz effective frequency band). Then, a feature fusion layer is used to correlate the time-domain and frequency-domain information to achieve accurate classification of signal types (including typical conditions such as leakage and corrosion). Finally, based on the spatial correlation of the multi-sensor two signals, combined with the time difference location method (TDOA) and deep feature matching, a leakage probability value in the range of 0-1 and the predicted positioning coordinates are output. Step S3: Comprehensive diagnosis and rating of multi-source data, see Appendix Figure 8 The system first aggregates the leak detection results from each of the three independent monitoring methods. It then assigns a comprehensive rating based on the number of leaks identified in these results. If only one of the three methods identifies a leak, the system assigns a Level I rating. If two methods identify a leak, the system assigns a Level II rating. And if all three methods consistently identify a leak, the system assigns the highest level, Level III. The more leaks identified by the monitoring methods, the higher the risk level.

[0036] Based on the same inventive concept, this embodiment provides a multi-sensor collaborative intelligent tank leakage diagnosis device, which integrates the above-mentioned intelligent tank leakage diagnosis system and executes the above-mentioned intelligent tank leakage diagnosis method.

[0037] This embodiment establishes a comprehensive leakage detection and diagnosis system by integrating the liquid level value of the medium in the storage tank, the index value of the medium inside and outside the storage tank, and the leakage signal value of the tank bottom plate, and integrates the regular detection data for intelligent judgment. It can be used simultaneously for comprehensive judgment to improve the accuracy of leakage judgment and improve the safety of storage tank use. Any single module can also be selected for use. For suspected leaking storage tanks, the internal liquid level can be judged first, then the external oil and gas concentration can be judged, and finally the acoustic emission leakage detection can be performed without opening the tank. The detection method is progressive, forming a set of leakage judgment and maintenance decision-making system. It avoids the human and material costs caused by managers blindly opening the tank for detection or excessive detection, and improves the utilization rate of the tank.

[0038] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to equivalent embodiments using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any indirect modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A multi-sensor coordinated intelligent tank leakage diagnosis system, characterized by: include: Signal acquisition components, including several sensors for monitoring tank information data and a host computer that can receive sensor signals; The software functional components include a tank structure modeling module that establishes a virtual tank model based on the tank structure data, a periodic detection data module that regularly detects and stores the tank data, a real-time monitoring module that monitors the real-time data of the tank, and a leakage diagnosis module that diagnoses the real-time data. The real-time monitoring module includes a liquid level monitoring module, an oil and gas concentration monitoring module, and an acoustic emission monitoring module. The risk warning model receives the diagnosis results of the leakage diagnosis module and outputs a tank leakage diagnosis warning signal.

2. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 1 is characterized in that: The tank structure modeling module presets a relationship function between the tank liquid level and the tank body deformation, and a tank volatilization function related to the temperature and liquid level. The liquid level monitoring module obtains real-time liquid level monitoring data and temperature of the tank through a signal acquisition component. The leakage diagnosis module synchronously calculates the tank foundation deformation liquid level change, the oil and gas volatilization liquid level change, the oil and gas volatilization liquid level change error, the measurement error determined by the accuracy parameters of the liquid level monitor, and the real-time liquid level change. When the real-time liquid level change is greater than the sum of the tank foundation deformation liquid level change, the oil and gas volatilization liquid level change, the oil and gas volatilization liquid level change error, and the measurement error determined by the accuracy parameters of the liquid level monitor, it is determined that the tank is leaking.

3. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 1 is characterized in that: The oil and gas concentration threshold is preset in the tank structure modeling module, the oil and gas concentration monitoring module collects real-time concentration data inside and outside the tank through a signal acquisition component, and the leakage diagnosis module compares and calculates the real-time concentration data inside and outside the tank with the oil and gas concentration threshold and determines whether the tank is leaking.

4. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 3 is characterized in that: In the signal acquisition component, oil and gas concentration sensor arrays are deployed in a tank gradient, located inside and outside the tank respectively. The oil and gas concentration monitoring module obtains the monitoring data of the oil and gas concentration sensor array, and the leakage diagnosis module determines whether it is abnormal. The risk warning model sorts the oil and gas concentration sensors corresponding to the abnormal data in time response order, obtains the sensor array with the earliest time response, and marks the tank position corresponding to the oil and gas concentration sensor.

5. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 1 is characterized in that: The tank structure modeling module constructs a signal feature library of different types. The acoustic emission monitoring module collects acoustic emission signals through a signal acquisition component. The leakage diagnosis module performs time domain analysis and frequency domain analysis on the collected acoustic emission signals, and combines the time difference positioning method with the depth feature matching to determine whether leakage occurs and the leakage area.

6. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 5 is characterized in that: In the signal acquisition component, eight acoustic emission sensors are evenly deployed circumferentially at a height of 20 cm on the bottom plate of the tank to form a circular array to collect acoustic emission signals.

7. The multi-sensor coordinated intelligent tank leakage diagnosis system according to claim 1 is characterized in that: The risk warning model determines the number of modules with abnormal data in the liquid level monitoring module, oil and gas concentration monitoring module, and acoustic emission monitoring module, and issues alarm prompts according to the level.

8. A multi-sensor collaborative intelligent diagnosis method for storage tank leakage, characterized in that: The method comprises using the intelligent tank leakage diagnosis system according to any one of claims 1 to 7 to perform leakage diagnosis on a storage tank.

9. The multi-sensor coordinated intelligent tank leakage diagnosis method according to claim 8, characterized in that: The steps of using the intelligent tank leakage diagnosis system according to any one of claims 1 to 7 to diagnose a tank leakage are as follows: Acquire real-time tank information data through signal acquisition components; The regular detection data module obtains real-time tank information data from the signal acquisition component and transmits it to the leakage diagnosis module; The leakage diagnosis module transmits the diagnosis results to the risk warning model; The risk warning model outputs a tank leakage diagnosis warning signal.

10. A multi-sensor coordinated intelligent diagnostic device for storage tank leakage, characterized in that: The intelligent tank leakage diagnosis device integrates the intelligent tank leakage diagnosis system described in any one of 1-7, and executes the intelligent tank leakage diagnosis method described in claim 8 or 9.

Images