CN116822255B - Light AI model construction method based on autonomous controllable software and hardware platform - Google Patents
Light AI model construction method based on autonomous controllable software and hardware platform Download PDFInfo
- Publication number
- CN116822255B CN116822255B CN202311092675.9A CN202311092675A CN116822255B CN 116822255 B CN116822255 B CN 116822255B CN 202311092675 A CN202311092675 A CN 202311092675A CN 116822255 B CN116822255 B CN 116822255B
- Authority
- CN
- China
- Prior art keywords
- container
- defect
- coefficient
- abnormal
- discrete
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000010276 construction Methods 0.000 title claims abstract description 23
- 230000007547 defect Effects 0.000 claims abstract description 99
- 230000002159 abnormal effect Effects 0.000 claims abstract description 79
- 238000007667 floating Methods 0.000 claims abstract description 46
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000005452 bending Methods 0.000 claims abstract description 16
- 238000004364 calculation method Methods 0.000 claims abstract description 5
- 238000003860 storage Methods 0.000 claims description 17
- 230000005856 abnormality Effects 0.000 claims description 16
- 230000008859 change Effects 0.000 claims description 6
- 239000006185 dispersion Substances 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 230000000903 blocking effect Effects 0.000 claims description 2
- 230000002950 deficient Effects 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 7
- 230000008569 process Effects 0.000 abstract description 7
- 238000012545 processing Methods 0.000 abstract description 6
- 238000001514 detection method Methods 0.000 description 5
- 239000007788 liquid Substances 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- 238000012544 monitoring process Methods 0.000 description 4
- 238000011835 investigation Methods 0.000 description 3
- 230000008439 repair process Effects 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 239000007789 gas Substances 0.000 description 2
- 238000009776 industrial production Methods 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000003345 natural gas Substances 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 238000013105 post hoc analysis Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Abstract
The invention discloses a lightweight AI model construction method based on an autonomous controllable software and hardware platform, which relates to the technical field of model construction, wherein an abnormal coefficient is built by comprehensively calculating a vessel pressure floating coefficient, a defect discrete index and the impact amplitude of a pipeline bending part, the abnormal coefficient is compared with an abnormal threshold value to complete AI model construction, the built AI model is adapted to an embedded hardware platform of a vessel, the state of the vessel is monitored at regular time through the embedded hardware platform, and when the probability of future operation failure of the vessel is predicted by the embedded hardware platform based on the AI model to be large, an early warning signal is sent. According to the invention, in the operation process of the pressure vessel, the edge calculation and analysis are performed based on the AI model, data is not required to be transmitted to the cloud, the data processing efficiency is improved, the fault prediction is performed on the pressure vessel, the judgment and the processing are effectively performed in time before the pressure vessel breaks down, and the safe operation of the pressure vessel is ensured.
Description
Technical Field
The invention relates to the technical field of model construction, in particular to a lightweight AI model construction method based on an autonomous controllable software and hardware platform.
Background
Pressure vessels are a type of vessel with certain pressure for storing and transporting gas, liquid or steam and the like, which are widely applied to various industrial fields such as chemical industry, petroleum, natural gas, energy, pharmacy, food processing and the like, play a vital role in modern industrial production, but bring certain safety risks at the same time, and need strict supervision and reasonable use;
pressure vessels play an important role in industrial production, but due to their special working environment and high-pressure operating conditions, once malfunctions or operational errors occur, serious accidents and casualties may occur, so safety monitoring and fault prediction for pressure vessels are very important.
The prior art has the following defects:
the traditional pressure vessel monitoring method usually relies on a sensor to collect data and then upload the data to a cloud for processing and analysis, however, the method has some problems such as data transmission delay, dependence on network connection and massive occupation of cloud resources;
the existing monitoring method for the pressure vessel is usually fault detection, however, when the pressure vessel fails, problems such as leakage and explosion can occur, and serious safety risks are brought to surrounding staff and environment.
Disclosure of Invention
The invention aims to provide a light AI model construction method based on an autonomous controllable software and hardware platform, so as to solve the defects in the background technology.
In order to achieve the above object, the present invention provides the following technical solutions: the method for constructing the light AI model based on the independently controllable software and hardware platform comprises the following steps:
s1: acquiring a storage medium parameter in the container, and analyzing whether the pressure in the container is abnormal or not after acquiring a pressure floating coefficient of the container based on the storage medium parameter;
s2: detecting defects in a randomly selected area of the outer wall of the container through a mobile ultrasonic sensor, and analyzing whether the outer wall of the container is abnormal or not after obtaining a defect discrete index based on the depth of the defects;
s3: when the internal pressure of the container is abnormal or the outer wall of the container is abnormal, an early warning signal is sent out, and when the internal pressure of the container is not abnormal or the outer wall of the container is not abnormal, the step S4 is executed;
s4: comprehensively calculating the pressure floating coefficient of the container, the defect discrete index and the impact amplitude of the bent part of the pipeline to establish an abnormal coefficient, and comparing the abnormal coefficient with an abnormal threshold value to complete the AI model construction;
s5: the constructed AI model is adapted to an embedded hardware platform of the container, and the state of the container is monitored at regular time through the embedded hardware platform;
s6: and when the probability of the future operation failure of the container is large based on the AI model prediction, the embedded hardware platform sends out an early warning signal.
In a preferred embodiment, the calculated expression of the vessel pressure float coefficient is:
computing containerCoefficient of pressure float，/>For the real-time volume change of the magnet, +.>For the time period of early warning of the temperature of the storage medium inside the container, < + >>The time period of early warning for blocking the inlet and the outlet of the container.
In a preferred embodiment, the vessel pressure float coefficient is obtainedAfter that, if the pressure floating coefficient of the container is +>If the floating threshold value is greater than the floating threshold value, analyzing the pressure inside the container for abnormality, if the floating coefficient of the container pressure is +.>And when the floating threshold value is less than or equal to the threshold value, analyzing that the pressure inside the container is not abnormal.
In a preferred embodiment, the logic for obtaining the defect discrete index is:
calculating defect discrete coefficientsThe expression is:
in the middle of，/>Indicating the lack of the outer wall of the containerThe number of dip sampling points, +.>Is a positive integer>Representing different defect depths, < >>Representing defect depth average values;
if the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->Discrete threshold value or less, defect discrete index or less>；
If the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->> discrete threshold, defect discrete index->；
If the defect depth average value> depth threshold, and defect discrete coefficient +.>> discrete threshold, defect discrete index->；
If the defect depth average value> depth threshold, and defect discrete coefficient +.>Discrete threshold value or less, defect discrete index or less>。
In a preferred embodiment, in step S4, the AI model construction includes the steps of:
the pressure floating coefficient, defect discrete index and impact amplitude of the bent part of the pipeline are comprehensively calculated to establish an abnormal coefficientThe computational expression is:
in the method, in the process of the invention,for defect discrete index>For the pressure floating coefficient of the container->Is the impact amplitude of the bending part of the pipeline, +.>、/>、/>The dispersion index of the defect, the pressure floating coefficient of the container and the proportionality coefficient of the impact amplitude of the bending part of the pipeline are respectively +.>、/>、/>Are all greater than 0;
and comparing the anomaly coefficient with an anomaly threshold value to complete AI model construction.
In a preferred embodiment, the constructed AI model is fitted into the embedded hardware platform of the container, and the container state is monitored periodically by the embedded hardware platform, if abnormal coefficients are presentIf the failure probability is smaller than the abnormality threshold value, the embedded hardware platform predicts that the future operation of the container is large based on the AI model, and sends out an early warning signal, if the failure coefficient is + ->And if the probability of the future operation failure of the container is larger than or equal to the abnormal threshold value, the embedded hardware platform predicts the probability of the future operation failure of the container based on the AI model, and does not send out an early warning signal.
In a preferred embodiment, the calculated expression of the impact amplitude at the pipe bend is:
in the method, in the process of the invention,is the impact amplitude of the bending part of the pipeline, +.>Is impact force>Is the distance of the bend from the point of impact,is the moment of inertia of the section of the pipe,/->Is the area of the duct cross section.
In the technical scheme, the invention has the technical effects and advantages that:
1. according to the method, after the storage medium parameters in the container are obtained, the container pressure floating coefficient is obtained based on the storage medium parameters, whether the container internal pressure is abnormal or not is analyzed, a random selection area of the container outer wall is detected through the mobile ultrasonic sensor, after the defect discrete index is obtained based on the defect depth, whether the container outer wall is abnormal or not is analyzed, the container pressure floating coefficient, the defect discrete index and the impact amplitude of a pipeline bending position are comprehensively calculated to establish the abnormal coefficient, the abnormal coefficient is compared with an abnormal threshold value to complete the AI model construction, the constructed AI model is adapted to an embedded hardware platform of the container, the state of the container is monitored through the embedded hardware platform at regular time, when the probability of the container future operation failure is predicted based on the AI model, an early warning signal is sent out, the probability of the container future operation failure is predicted to be small, and the early warning signal is not sent out.
2. The invention establishes an abnormal coefficient by comprehensively calculating the pressure floating coefficient, the defect discrete index and the impact amplitude of the bent part of the pipeline to obtain the abnormal coefficientAnd then, comparing the abnormal coefficient with an abnormal threshold value to complete AI model construction, effectively improving data processing efficiency, predicting faults of the pressure vessel in operation, and guaranteeing safe use of the pressure vessel.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments described in the present invention, and other drawings may be obtained according to these drawings for a person having ordinary skill in the art.
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1: referring to fig. 1, the method for constructing a lightweight AI model based on an autonomous controllable software and hardware platform according to the present embodiment includes the following steps:
after obtaining the storage medium parameter in the container and the pressure floating coefficient of the container based on the storage medium parameter, analyzing whether the pressure in the container is abnormal or not, specifically comprising the following steps:
firstly, acquiring relevant parameters of a storage medium in a container, wherein the parameters are the basis for calculating a container pressure floating coefficient and analyzing pressure abnormality, calculating the pressure floating coefficient of the container by using the parameters of the storage medium in the container, analyzing whether the pressure in the container is abnormal or not by using the calculated pressure floating coefficient, and continuously monitoring the pressure change in the container in the whole process and recording data in time. These data are important for post-hoc analysis and problem investigation, and if an abnormality is found in the internal pressure of the container, emergency measures such as cutting off the medium supply, shutting down the equipment, etc. are taken in time according to circumstances to ensure safety.
Detecting defects in a randomly selected area of the outer wall of the container through a mobile ultrasonic sensor, and analyzing whether the outer wall of the container is abnormal or not after obtaining a defect discrete index based on the depth of the defects, wherein the method specifically comprises the following steps:
first, the detection area of the outer wall of the container is randomly selected. Thus, the surface of the outer wall of the container can be covered more comprehensively, the detection of potential defects is ensured to be comprehensive and uniform, and the mobile ultrasonic sensor is used for detecting the defects in the randomly selected area. The ultrasonic sensor can detect defects inside the material, and depth data of the detected defects are obtained through ultrasonic detection. The defect depth refers to the extension depth of the defect in the outer wall of the container, the defect discrete index is calculated based on the defect depth data, whether the outer wall of the container is abnormal or not is analyzed by combining the defect depth and the discrete index calculation result, and the acquired data and analysis result are recorded in time and reported in the detection and analysis process. These data and analysis results are important for subsequent maintenance and repair decisions, and if abnormal defects are found on the outer wall of the container, necessary measures such as repair, replacement or other maintenance measures need to be taken to ensure safe operation of the container.
When the internal pressure of the container is abnormal or the outer wall of the container is abnormal, sending an early warning signal, and when the internal pressure of the container is not abnormal or the outer wall of the container is not abnormal, executing the next step;
comprehensively calculating the pressure floating coefficient of the container, the defect discrete index and the impact amplitude of the bent part of the pipeline to establish an abnormal coefficient, and comparing the abnormal coefficient with an abnormal threshold value to complete the AI model construction;
the constructed AI model is adapted to an embedded hardware platform of a container, the state of the container is monitored at regular time through the embedded hardware platform, when the embedded hardware platform predicts that the probability of the future operation of the container is high based on the AI model, an early warning signal is sent out, when the probability of the future operation of the container is small, the early warning signal is not sent out, and when a container manager receives the early warning signal, corresponding management needs to be timely carried out, and the corresponding management comprises the following steps:
1) Immediately, an administrator should conduct an emergency check to check whether the pressure vessel and related equipment are abnormal, which may include checking parameters such as pressure, temperature, liquid level, etc. to ensure that the equipment is operating within safe limits;
2) Isolation risk area: if the early warning signal indicates that serious risks possibly exist, an administrator should timely isolate the risk areas so as to prevent personnel from entering the areas possibly having the risks;
3) Notifying relevant personnel: the manager should timely inform relevant operation staff, security staff or management layer, report early warning signals to them, and jointly make further treatment schemes;
4) Emergency measures are taken: during the investigation, if it is found that there is indeed a problem, the administrator should take emergency measures, such as emergency shut-down of the equipment, cutting off of the medium supply, etc., to prevent further risk expansion.
According to the method, after the storage medium parameters in the container are obtained, the container pressure floating coefficient is obtained based on the storage medium parameters, whether the container internal pressure is abnormal or not is analyzed, a random selection area of the container outer wall is detected through the mobile ultrasonic sensor, after the defect discrete index is obtained based on the defect depth, whether the container outer wall is abnormal or not is analyzed, the container pressure floating coefficient, the defect discrete index and the impact amplitude of a pipeline bending position are comprehensively calculated to establish the abnormal coefficient, the abnormal coefficient is compared with an abnormal threshold value to complete the AI model construction, the constructed AI model is adapted to an embedded hardware platform of the container, the state of the container is monitored through the embedded hardware platform at regular time, when the probability of the container future operation failure is predicted based on the AI model, an early warning signal is sent out, the probability of the container future operation failure is predicted to be small, and the early warning signal is not sent out.
Example 2: and acquiring a storage medium parameter in the container, and analyzing whether the pressure in the container is abnormal or not after acquiring the pressure floating coefficient of the container based on the storage medium parameter.
Wherein:
the calculated expression of the vessel pressure floating coefficient is:
calculating the pressure float coefficient of the container，/>For the real-time volume change of the magnet, +.>For the time period of early warning of the temperature of the storage medium inside the container, < + >>For the period of early warning of the blockage of the inlet and the outlet of the container, when the temperature of the high-temperature medium stored in the pressure container or the working environment is too high, the pressure in the container can be abnormally increased due to the thermal expansion of the medium, when the temperature of the low-temperature liquid stored in the pressure container or the working environment is too low, the pressure in the container can be abnormally reduced due to the shrinkage of the medium, and the outlet or the inlet pipeline of the pressure container is blocked, so that the gas or the liquid cannot flow out or flow in, and the internal pressure is abnormal.
Obtaining vessel pressure float coefficientAfter that, if the pressure floating coefficient of the container is +>If the floating threshold value is greater than the floating threshold value, analyzing the pressure inside the container for abnormality, if the floating coefficient of the container pressure is +.>And when the floating threshold value is less than or equal to the threshold value, analyzing that the pressure inside the container is not abnormal.
Detecting defects in a randomly selected area of the outer wall of the container through a mobile ultrasonic sensor, and analyzing whether the outer wall of the container is abnormal or not after obtaining a defect discrete index based on the depth of the defects.
Wherein:
the acquisition logic of the defect discrete index is as follows:
firstly, calculating defect discrete coefficientsThe computational expression is:
in the middle of，/>Representing the number of defective sampling points of the outer wall of the container, +.>Is a positive integer>Representing different defect depths, < >>Representing defect depth average values;
if the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->Discrete threshold value or less, defect discrete index or less>No abnormality exists on the outer wall of the analysis container, and management is not needed;
if the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->> discrete threshold, defect discrete index->Analyzing the existence of slight abnormality on the outer wall of the container, wherein the slight abnormality may mean that some minor defects or problems exist, and the structural integrity and the safety of the container are not greatly influenced, so that the container needs to be periodically detected and monitored, abnormal change conditions are recorded so as to discover potential problems in time, a periodic inspection system is established, the outer wall of the container is periodically inspected, including detecting the defects and acquiring the defect depth by using an ultrasonic sensor, tracking and analyzing the development of the slight abnormality, and knowing the abnormal change trend so as to take corresponding measures in time;
if the defect depth average value> depth threshold, and defect discrete coefficient +.>> discrete threshold, defect discrete index->Analyzing the existence of moderate abnormality on the outer wall of the container, wherein the moderate abnormality possibly indicates the existence of larger or more defects and possibly has certain influence on the structural integrity of the container, so that the detected moderate abnormality needs to be repaired and maintained, the defects are timely processed, further expansion is prevented, the inspection frequency is increased, the container is inspected more frequently, so that new abnormal conditions can be timely found, the areas with the moderate abnormality are safely isolated, and personnel are prevented from entering the areas possibly with danger;
if the defect depth average value> depth threshold, and lack ofDiscrete coefficient of trapping->Discrete threshold value or less, defect discrete index or less>Analyzing that the outer wall of the container is severely abnormal, wherein the severely abnormal represents that the outer wall of the container has serious defects or structural problems, and possibly has serious threat to the safety of the container, so that the container needs to be immediately shut down, medium supply is cut off, personnel safety is ensured, emergency repair or replacement is carried out on the area with severely abnormal, the safety before the container is restored to normal operation is ensured, accident investigation is carried out on the found severely abnormal, the reason of the abnormal is found, and measures are taken to prevent similar problems from happening again.
Because the pressure vessel belongs to high-risk industrial equipment, when the outer wall of the analysis vessel is slightly abnormal, an early warning signal is required to be sent out, and the safe use of the pressure vessel is further ensured.
Example 3: and comprehensively calculating the pressure floating coefficient of the container, the defect discrete index and the impact amplitude of the bent part of the pipeline to establish an abnormal coefficient, and comparing the abnormal coefficient with an abnormal threshold value to complete the AI model construction.
The constructed AI model is adapted to an embedded hardware platform of the container, the state of the container is monitored at regular time through the embedded hardware platform, when the embedded hardware platform predicts that the probability of the future operation of the container is high based on the AI model, an early warning signal is sent out, and when the probability of the future operation of the container is low, the early warning signal is not sent out.
Wherein:
the calculation expression of the impact amplitude at the bending part of the pipeline is as follows:
in the method, in the process of the invention,the impact amplitude of the bending part of the pipeline,/>is impact force>Is the distance of the bend from the point of impact,is the moment of inertia of the section of the pipe,/->Is the area of the cross section of the pipeline, and the larger the impact amplitude of the bent part of the pipeline is, the more causes:
1) Pipe rupture or leakage: excessive impact amplitudes can cause the material at the pipe bend to be stressed beyond its bearing capacity, thereby causing the pipe to crack or leak; the rupture or leakage may cause danger and environmental pollution, and threaten the safety of equipment and personnel;
2) Pipe deformation and damage: the increase in the impact amplitude may cause deformation and damage to the pipe bend, changing its shape and structure; this will affect the proper operation and performance of the pipeline, possibly resulting in reduced pipeline function or inoperability;
3) Equipment failure: excessive impact amplitude can cause excessive stress on equipment such as connectors, valves, etc. in the pipeline system, thereby causing equipment failure or failure;
4) Potential safety hazard: the impact amplitude at the bending part is too large, so that potential safety hazards can be brought, for example, instability of a pipeline system is caused, and other dangerous accidents are caused;
5) Unstable operation: the bending part is impacted by too large amplitude, so that the running of a pipeline system is unstable, and the problems of unsmooth medium flow, pressure fluctuation and the like are caused.
The pressure floating coefficient, defect discrete index and impact amplitude of the bent part of the pipeline are comprehensively calculated to establish an abnormal coefficientThe computational expression is:
in the method, in the process of the invention,for defect discrete index>For the pressure floating coefficient of the container->Is the impact amplitude of the bending part of the pipeline, +.>、/>、/>The dispersion index of the defect, the pressure floating coefficient of the container and the proportionality coefficient of the impact amplitude of the bending part of the pipeline are respectively +.>、/>、/>Are all greater than 0.
Obtaining anomaly coefficientsAnd then, comparing the anomaly coefficient with an anomaly threshold value to complete AI model construction.
The constructed AI model is adapted to an embedded hardware platform of the container, the state of the container is monitored regularly through the embedded hardware platform, and if abnormal coefficients are generatedLess than an anomaly threshold, embedded hardwareThe platform predicts that the probability of the future operation of the container is high based on an AI model, sends out an early warning signal, and if an abnormality coefficient is ++>And if the probability of the future operation failure of the container is larger than or equal to the abnormal threshold value, the embedded hardware platform predicts the probability of the future operation failure of the container based on the AI model, and does not send out an early warning signal.
The method establishes the abnormal coefficient through comprehensively calculating the pressure floating coefficient, the defect discrete index and the impact amplitude of the bent part of the pipeline to obtain the abnormal coefficientAnd then, comparing the abnormal coefficient with an abnormal threshold value to complete AI model construction, effectively improving data processing efficiency, predicting faults of the pressure vessel in operation, and guaranteeing safe use of the pressure vessel.
The above formulas are all formulas with dimensions removed and numerical values calculated, the formulas are formulas with a large amount of data collected for software simulation to obtain the latest real situation, and preset parameters in the formulas are set by those skilled in the art according to the actual situation.
In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (5)
1. The light AI model construction method based on the autonomous controllable software and hardware platform is characterized in that: the construction method comprises the following steps:
s1: acquiring a storage medium parameter in the container, and analyzing whether the pressure in the container is abnormal or not after acquiring a pressure floating coefficient of the container based on the storage medium parameter;
s2: detecting defects in a randomly selected area of the outer wall of the container through a mobile ultrasonic sensor, and analyzing whether the outer wall of the container is abnormal or not after obtaining a defect discrete index based on the depth of the defects;
s3: when the internal pressure of the container is abnormal or the outer wall of the container is abnormal, an early warning signal is sent out, and when the internal pressure of the container is not abnormal or the outer wall of the container is not abnormal, the step S4 is executed;
s4: comprehensively calculating the pressure floating coefficient of the container, the defect discrete index and the impact amplitude of the bent part of the pipeline to establish an abnormal coefficient, and comparing the abnormal coefficient with an abnormal threshold value to complete the AI model construction;
s5: the constructed AI model is adapted to an embedded hardware platform of the container, and the state of the container is monitored at regular time through the embedded hardware platform;
s6: when the probability of the future operation failure of the container is large based on the AI model prediction, an early warning signal is sent out;
the calculation expression of the pressure floating coefficient of the container is as follows:
calculating the pressure floating coefficient of the container>，/>For the real-time volume change of the magnet, +.>For the time period of early warning of the temperature of the storage medium inside the container, < + >>A time period for early warning of the blocking of the inlet and the outlet of the container;
in step S4, the AI model construction includes the steps of:
the pressure floating coefficient, defect discrete index and impact amplitude of the bent part of the pipeline are comprehensively calculated to establish an abnormal coefficientThe computational expression is:
in (1) the->For defect discrete index>For the pressure floating coefficient of the container->Is the impact amplitude of the bending part of the pipeline, +.>、/>、/>The dispersion index of the defect, the pressure floating coefficient of the container and the proportionality coefficient of the impact amplitude of the bending part of the pipeline are respectively +.>、/>、/>Are all greater than 0;
and comparing the anomaly coefficient with an anomaly threshold value to complete AI model construction.
2. The method for constructing the lightweight AI model based on the autonomous controllable software and hardware platform according to claim 1, wherein the method comprises the following steps: obtaining the pressure floating coefficient of the containerAfter that, if the pressure floating coefficient of the container is +>If the floating threshold value is greater than the floating threshold value, analyzing the pressure inside the container for abnormality, if the floating coefficient of the container pressure is +.>And when the floating threshold value is less than or equal to the threshold value, analyzing that the pressure inside the container is not abnormal.
3. The method for constructing the lightweight AI model based on the autonomous controllable software and hardware platform according to claim 1, wherein the method comprises the following steps: the acquisition logic of the defect discrete index is as follows:
calculating defect discrete coefficientsThe expression is:
in->，/>Representing the number of defective sampling points of the outer wall of the container, +.>Is a positive integer>Representing different defect depths, < >>Representing defect depth average values;
if the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->Discrete threshold value or less, defect discrete index；
If the defect depth average valueDepth threshold value is less than or equal to, and defect discrete coefficient is less than or equal to->> discrete threshold, defect discrete index；
If the defect depth average value> depth threshold, and defect discrete coefficient +.>> discrete threshold, defect discrete index；
If the defect depth average value> depth threshold, and defect discrete coefficient +.>Discrete threshold value or less, defect discrete index。
4. The method for constructing the lightweight AI model based on the autonomous controllable software and hardware platform according to claim 1, wherein the method comprises the following steps: the constructed AI model is adapted to an embedded hardware platform of the container, the state of the container is monitored regularly through the embedded hardware platform, and if abnormal coefficients are generatedIf the failure probability is smaller than the abnormality threshold value, the embedded hardware platform predicts that the future operation of the container is large based on the AI model, and sends out an early warning signal, if the failure coefficient is + ->And if the probability of the future operation failure of the container is larger than or equal to the abnormal threshold value, the embedded hardware platform predicts the probability of the future operation failure of the container based on the AI model, and does not send out an early warning signal.
5. The method for constructing the lightweight AI model based on the autonomous controllable software and hardware platform as claimed in claim 4, wherein the method comprises the following steps: the calculation expression of the impact amplitude at the bending part of the pipeline is as follows:
in (1) the->Is the impact amplitude of the bending part of the pipeline, +.>Is impact force>Is the distance of the bend from the impact point, +.>Is the moment of inertia of the section of the pipe,/->Is the area of the duct cross section.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311092675.9A CN116822255B (en) | 2023-08-29 | 2023-08-29 | Light AI model construction method based on autonomous controllable software and hardware platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311092675.9A CN116822255B (en) | 2023-08-29 | 2023-08-29 | Light AI model construction method based on autonomous controllable software and hardware platform |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116822255A CN116822255A (en) | 2023-09-29 |
| CN116822255B true CN116822255B (en) | 2023-12-29 |
Family
ID=88114870
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311092675.9A Active CN116822255B (en) | 2023-08-29 | 2023-08-29 | Light AI model construction method based on autonomous controllable software and hardware platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116822255B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109783906A (en) * | 2018-12-29 | 2019-05-21 | 东北大学 | A kind of pipeline detection magnetic flux leakage data intelligent analysis system and method |
| CN110501425A (en) * | 2019-09-04 | 2019-11-26 | 江苏鸣帆工程检测有限公司 | A kind of fixed pressure vessel ultrasonic wave detecting system and method |
| CN114137079A (en) * | 2021-11-30 | 2022-03-04 | 东莞理工学院 | Ultrasonic guided wave nondestructive testing method based on combination of deep learning and Duffing system |
| CN114148533A (en) * | 2021-12-15 | 2022-03-08 | 中国航空工业集团公司成都飞机设计研究所 | System and method for monitoring liquid leakage of airplane in real time |
| CN114493393A (en) * | 2021-12-06 | 2022-05-13 | 南京市锅炉压力容器检验研究院 | Multi-dimensional pressure container health monitoring system and method based on artificial intelligence field calculation |
| CN114577418A (en) * | 2022-05-05 | 2022-06-03 | 欧米勒电气有限公司 | Pressure vessel leakage detection system based on Internet of things |
| CN115236196A (en) * | 2021-04-23 | 2022-10-25 | 中国石油化工股份有限公司 | Pressure container monitoring system and method |
| CN115730720A (en) * | 2022-11-21 | 2023-03-03 | 上海橘盒数字科技有限公司 | Product data abnormity prediction method, system, device and storage medium |
| CN116183726A (en) * | 2023-02-22 | 2023-05-30 | 山东丰汇工程检测有限公司 | Ultrasonic detection capability evaluation method and system for stainless steel small-diameter pipe weld joint |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11488010B2 (en) * | 2018-12-29 | 2022-11-01 | Northeastern University | Intelligent analysis system using magnetic flux leakage data in pipeline inner inspection |
-
2023
- 2023-08-29 CN CN202311092675.9A patent/CN116822255B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109783906A (en) * | 2018-12-29 | 2019-05-21 | 东北大学 | A kind of pipeline detection magnetic flux leakage data intelligent analysis system and method |
| CN110501425A (en) * | 2019-09-04 | 2019-11-26 | 江苏鸣帆工程检测有限公司 | A kind of fixed pressure vessel ultrasonic wave detecting system and method |
| CN115236196A (en) * | 2021-04-23 | 2022-10-25 | 中国石油化工股份有限公司 | Pressure container monitoring system and method |
| CN114137079A (en) * | 2021-11-30 | 2022-03-04 | 东莞理工学院 | Ultrasonic guided wave nondestructive testing method based on combination of deep learning and Duffing system |
| CN114493393A (en) * | 2021-12-06 | 2022-05-13 | 南京市锅炉压力容器检验研究院 | Multi-dimensional pressure container health monitoring system and method based on artificial intelligence field calculation |
| CN114148533A (en) * | 2021-12-15 | 2022-03-08 | 中国航空工业集团公司成都飞机设计研究所 | System and method for monitoring liquid leakage of airplane in real time |
| CN114577418A (en) * | 2022-05-05 | 2022-06-03 | 欧米勒电气有限公司 | Pressure vessel leakage detection system based on Internet of things |
| CN115730720A (en) * | 2022-11-21 | 2023-03-03 | 上海橘盒数字科技有限公司 | Product data abnormity prediction method, system, device and storage medium |
| CN116183726A (en) * | 2023-02-22 | 2023-05-30 | 山东丰汇工程检测有限公司 | Ultrasonic detection capability evaluation method and system for stainless steel small-diameter pipe weld joint |
Non-Patent Citations (3)
| Title |
|---|
| Adaptive residual CNN-based fault detection and diagnosis system of small modular reactors;Yuantao Yao灯;《Applied Soft Computing》;全文 * |
| 压力容器检验与危险源的控制措施研究;杜汉伟;《中国设备工程》;全文 * |
| 压力容器气密性故障检测方法研究;廖湖广;《万方硕士学位论文数据库》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116822255A (en) | 2023-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR101341248B1 (en) | Risk based maintenance system for power generation facilities | |
| CN102495935B (en) | Determination method for risk of storage medium leakage of underground natural gas storage reservoir | |
| CN114577418B (en) | Pressure vessel leakage detection system based on Internet of things | |
| Pang et al. | An integrated LSTM-AM and SPRT method for fault early detection of forced-oxidation system in wet flue gas desulfurization | |
| CN116822255B (en) | Light AI model construction method based on autonomous controllable software and hardware platform | |
| CN117629549A (en) | Bridge building health monitoring and safety early warning system | |
| Korshunov et al. | Assurance of reliability and safety in liquid hydrocarbons marine transportation and storing | |
| JP2009115090A (en) | Automatic detection and notification of degradation of turbine internal component | |
| CN111259562B (en) | Dynamic virtual system for analyzing, controlling and controlling failure area of heating surface of boiler | |
| CN114278873B (en) | Remote monitoring method for pipeline fault | |
| Cong et al. | Risk assessment based on fault tree analysis for damaged pipe repair during operation in petrochemical plant | |
| Maturana et al. | Methodology for defining risk indices applicable to subsea equipment | |
| Wang et al. | Optimal inspection and maintenance plans for corroded pipelines | |
| JPH09113673A (en) | Anomaly surveillance method and device for plant | |
| KR20210021857A (en) | Intelligent diagnosing system available of facility and equipment operation by dynamic simulator of marine structure | |
| Yang et al. | Machine Learning in Process Safety and Asset Integrity Management | |
| KR20190083131A (en) | Methode for pipe wall thinning management based on risk ranking | |
| CN102749848A (en) | Water supply pipeline safety monitoring method based on useless component projection analysis | |
| Sun et al. | Fault Diagnosis Method for Ash Silo in Thermal Power Plant Based on Variable Weight Synthesizing Theory | |
| Liu et al. | Construction of monitoring system for pump, pressure vessel and pressure pipeline based on big data analysis | |
| Li et al. | Integrity Management of Large Size Atmospheric Storage Tank | |
| Chen et al. | Research and application of risk assessment and monitoring and early warning technology for oil and gas storage and transportation pipelines | |
| Yasseri | Fragility analysis of corroded pipeline | |
| Dadhich et al. | Fault detection and automation in the liquid storage terminals | |
| Bryla et al. | Real time & long term acoustic emission monitoring: A new way to use acoustic emission—Application to hydroelectric penstocks and paper machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |