CN116579663B - Abnormal early warning method in unloading process of powder tank truck - Google Patents

Abnormal early warning method in unloading process of powder tank truck Download PDF

Info

Publication number
CN116579663B
CN116579663B CN202310814665.5A CN202310814665A CN116579663B CN 116579663 B CN116579663 B CN 116579663B CN 202310814665 A CN202310814665 A CN 202310814665A CN 116579663 B CN116579663 B CN 116579663B
Authority
CN
China
Prior art keywords
sequence
time period
group
stability evaluation
calculating
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
Application number
CN202310814665.5A
Other languages
Chinese (zh)
Other versions
CN116579663A (en
Inventor
刘霞云
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Huiyuan Intelligent Technology Co ltd
Original Assignee
Jiangsu Huiyuan Intelligent Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Huiyuan Intelligent Technology Co ltd filed Critical Jiangsu Huiyuan Intelligent Technology Co ltd
Priority to CN202310814665.5A priority Critical patent/CN116579663B/en
Publication of CN116579663A publication Critical patent/CN116579663A/en
Application granted granted Critical
Publication of CN116579663B publication Critical patent/CN116579663B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Physics & Mathematics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • Educational Administration (AREA)
  • General Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Game Theory and Decision Science (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The application relates to the technical field of containers for material transportation, in particular to an abnormality early warning method in the unloading process of a powder tank truck; acquiring a pressure sequence, a new vibration sequence, an air flow sequence and a temperature sequence corresponding to each time period, and calculating stability evaluation of each time period; acquiring the difference distances of any two stability evaluations, and grouping all time periods according to the difference distances to obtain a plurality of groups; calculating the correlation of each group, and calculating the health degree of each time period according to the correlation; and training the LSTM network by taking the pressure sequence, the new vibration sequence, the air flow sequence, the temperature sequence, the correlation and the health degree of each time period as samples to obtain a trained LSTM network, then obtaining the predicted health degree of the future time period after the current time period, and judging whether the powder tank truck is abnormal according to the predicted health degree. The method can accurately predict whether the powder tank truck is abnormal in the future time period, and the air compressor is protected to the greatest extent.

Description

Abnormal early warning method in unloading process of powder tank truck
Technical Field
The application relates to the technical field of containers for material transportation, in particular to an abnormality early warning method in the unloading process of a powder tank truck.
Background
The powder tank truck is easy to cause the condition of explosion of the tank body due to overlarge pressure in the unloading operation process, and the personnel safety problem and the economic loss problem are extremely easy to cause; therefore, the unloading process of the powder tank truck needs to be monitored so as to avoid accidents; the method for monitoring the powder tank truck commonly used at present comprises the following steps: judging whether the powder tank truck needs to carry out pressure relief adjustment operation according to the pressure in the tank body so as to avoid accidents, wherein an operator is required to monitor the pressure in the tank at any time; meanwhile, when the pressure relief adjustment operation is needed to be carried out on the powder tank truck, partial damage is caused to the tank body and the air compressor under the condition, and the service performance of the tank body and the air compressor is affected, so that early warning is needed to be carried out in advance in the unloading process, and the tank body and the air compressor are protected.
Disclosure of Invention
In order to solve the technical problems, the application aims to provide an abnormal early warning method in the unloading process of a powder tank truck, and the adopted technical scheme is as follows:
collecting pressure, vibration acceleration and air flow corresponding to each detection moment of the powder tank truck in a plurality of time periods; obtaining a pressure sequence, a vibration acceleration sequence and an air flow sequence corresponding to each time period; acquiring the temperature of the external environment corresponding to each detection moment in a plurality of time periods to obtain a temperature sequence corresponding to each time period;
in each vibration acceleration sequence, calculating a variance for every z vibration acceleration according to the sequence from front to back, and obtaining a new vibration sequence corresponding to each vibration acceleration sequence according to the variance; wherein z is greater than 2;
calculating stability evaluation corresponding to each time period based on the pressure sequence, the air flow sequence and the new vibration sequence;
recording Euclidean distances of any two stability evaluations as difference distances corresponding to the two stability evaluations, and grouping all time periods based on the difference distances to obtain a plurality of groups;
acquiring a stability evaluation sequence and a temperature change sequence corresponding to each group; calculating the correlation corresponding to each group based on the stability evaluation sequence and the temperature change sequence;
calculating the health degree corresponding to each time period according to the stability evaluation and the average temperature corresponding to each time period and the correlation corresponding to the group to which each time period belongs;
taking a pressure sequence, a new vibration sequence, an air flow sequence, a temperature sequence, correlation and health degree corresponding to each time period as samples, and training an LSTM network by using the samples to obtain a trained LSTM network;
and obtaining the predicted health degree corresponding to the future time period after the current time period according to the trained LSTM network, and judging whether the powder tank truck is abnormal in the future time period according to the predicted health degree.
Preferably, the stability is evaluated as:
wherein ,for stability evaluation; />Is a pressure sequence; />Is a new vibration sequence; />Is a sequence of air flow;as a function of the maximum value; />As a function of the average; />As a function of variance; />Is an exponential function based on a natural constant e.
Preferably, the method for obtaining the stability evaluation sequence and the temperature change sequence corresponding to each group specifically comprises the following steps: the stability evaluation corresponding to each time period in the group forms a stability evaluation sequence corresponding to the group; and calculating the average temperature corresponding to the temperature sequence of each time period in the group, wherein the average temperature corresponding to each time period in the group forms the temperature change sequence corresponding to the group.
Preferably, the method for acquiring the correlation comprises the following steps:
wherein ,for relevance, ->For the temperature change sequence,/->For stability evaluation sequence, ++>Is->And->Pearson correlation coefficient of (b); />As a function of the maximum value; />Function for minimum;as a function of the variance.
Preferably, the health degree is:
wherein ,for health degree, the->For stability evaluation; />Is a temperature sequence; />Is a correlation; />As a hyperbolic tangent function; />As a function of the average; />For adjusting parameters, and->, in the formula ,/>Is the optimum temperature.
Preferably, the method for acquiring the corresponding overall loss function of the LSTM network in the training process comprises the following steps: randomly selecting a time period in any one group, and calculating the membership degree corresponding to the time period based on the difference distance between the stability evaluation corresponding to the time period and the stability evaluation corresponding to the rest of the time periods in the group; obtaining the membership degree corresponding to each time period in the group, and taking the time period corresponding to the maximum membership degree as the representative time period of the group; obtaining the representative time periods of each group;
calculating the accumulated sum of the difference distances between the stability evaluation corresponding to the representative time period of the group and the stability evaluation corresponding to the representative time periods of the rest other groups to obtain the inter-group difference corresponding to the group; thereby obtaining the difference between the groups corresponding to each group;
arranging all the time periods according to the content marked in advance to obtain a plurality of discharging sequences; according to the group-to-group difference of the groups to which each time period belongs in the unloading sequence, calculating the weight corresponding to each time period in the unloading sequence and the sequence weight corresponding to the unloading sequence; obtaining an overall loss function according to the weight and the sequence weight;
the overall loss function is:
wherein ,for the whole loss function->Is the total number of discharge sequences; />Is the total number of time periods in the discharge sequence; />The sequence weight corresponding to the nth unloading sequence; />The weight corresponding to the ith time period in the nth unloading sequence;is the loss corresponding to the ith time period in the nth discharge sequence.
Preferably, the membership obtaining method includes: and calculating the accumulated sum of the difference distances between the stability evaluation corresponding to any one time period in the group and the stability evaluation corresponding to other time periods remaining in the group, calculating the ratio of the accumulated sum to the total number of the time periods in the group minus 1, and recording the value normalized by the ratio as the membership degree of the time period.
Preferably, the weight obtaining method includes: the normalized value of the inter-group difference corresponding to the group to which each time period belongs is recorded as the weight corresponding to each time period;
the acquisition method of the sequence weight comprises the following steps: calculating the accumulated sum of the differences among the groups corresponding to the groups to which all the time periods belong in the unloading sequence; and the accumulated and normalized value is recorded as the sequence weight corresponding to the unloading sequence.
The embodiment of the application has at least the following beneficial effects:
according to the application, stability evaluation of each time period is calculated through a pressure sequence, a new vibration sequence and an air flow sequence corresponding to each time period; the stability evaluation is obtained through three aspects of vibration, pressure and air flow, and the stability evaluation is obtained by integrating multi-angle information, so that the obtained stability evaluation can more accurately represent the state of the unloading process corresponding to the time period. According to the application, the health degree of each time period is calculated according to the correlation; the aim of introducing the correlation in calculating the health degree is that the sensitivity degree of different tank trucks and different types of faults to temperature is different, and the integrated result is corrected by introducing the correlation; the obtained health degree is more accurate. Meanwhile, the application obtains the predicted health degree of the future time period after the current time period through the trained LSTM network, and judges whether the powder tank truck is abnormal in the future time period according to the predicted health degree. The method can accurately predict whether the powder tank truck is abnormal in the future time period, and the air compressor is protected to the greatest extent.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart illustrating steps of an embodiment of an abnormality early warning method in the unloading process of a powder tank truck according to the present application.
Detailed Description
In order to further describe the technical means and effects adopted by the present application to achieve the preset purpose, the following detailed description of the specific embodiments, structures, features and effects thereof according to the present application is given with reference to the accompanying drawings and the preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
Referring to fig. 1, a flowchart of a method for early warning of abnormality in a powder tank truck unloading process according to an embodiment of the present application is shown, and the method includes the following steps:
step 1, collecting pressure, vibration acceleration and air flow corresponding to each detection moment of a powder tank truck in a plurality of time periods; obtaining a pressure sequence, a vibration acceleration sequence and an air flow sequence corresponding to each time period; and acquiring the temperature of the external environment corresponding to each detection moment in a plurality of time periods to obtain a temperature sequence corresponding to each time period.
In the unloading process of the powder tank truck, the pressure is generally added into the tank body of the powder tank truck through an air compressor, namely compressed air (or external compressed air) generated by the air compressor enters the tank, the powder material is fluidized through a fluidization device, and the powder material is conveyed to a designated position outside the tank along an unloading pipeline along with air by means of the pressure difference between the inside and outside of the tank body. If the pressure in the tank body is abnormal, dangerous situations can occur, and meanwhile, the air compressor can be damaged.
Therefore, in this embodiment, the shock-resistant pressure gauge is used to collect the pressure corresponding to the powder tank truck, that is, the pressure in the tank body of the powder tank truck, so as to obtain the pressure sequence corresponding to each time period，/>, wherein ,/>For the pressure corresponding to the 1 st detection moment of the powder tank truck in any time period,/for the powder tank truck>For the pressure corresponding to the 2 nd detection moment of the powder tank truck in the time period,/for the powder tank truck>Corresponding to the mth detection moment of the powder tank truck in the time periodM is the total number of detection instants in the time period.
In the unloading process of the powder tank truck, when the output power of the air compressor is insufficient, the vibration condition of the air compressor has obvious difference from the vibration condition corresponding to the normal operation of the air compressor, and the concrete manifestation is that the vibration acceleration of the air compressor is unstable and has larger fluctuation; therefore, whether the air compressor is in a normal state can be represented by the vibration acceleration of the air compressor, namely, the embodiment obtains the corresponding vibration acceleration by arranging the MEMS vibration sensor on the air compressor, and obtains the vibration acceleration sequences corresponding to each time period，/>, wherein ,/>Vibration acceleration corresponding to 1 st detection moment of the powder tank truck in any time period is +.>Vibration acceleration corresponding to the 2 nd detection moment of the powder tank truck in the time period is +.>The vibration acceleration corresponding to the kth detection time in the time period is obtained for the powder tank truck, and k is the total number of detection times in the time period.
In the unloading process of the powder tank truck, when the opening angle of the ball valve is large, the material-gas ratio (the ratio of the mass of materials in a pipeline to the mass of gas) can be reduced, the unloading distance is increased, the conveying height is increased, but the unloading efficiency is reduced, and the unloading time is increased; conversely, the feed gas ratio can be increased, the conveying distance is shortened, the conveying height is reduced, the discharging efficiency is high, and the discharging time can be shorter; however, when the opening angle of the ball valve is too small, the discharge pipeline is easy to be blocked and explode, so that the opening angle of the ball valve needs to be measured for safety, and the opening angle of the ball valve cannot be directly measured, but the ball valveThe opening angle of the discharge pipe and the air flow in the discharge pipe are in positive correlation, so the opening angle of the side reaction ball valve is measured by measuring the air flow of the discharge pipe; specifically, an air flow meter is arranged at the front end of the ball valve, and the air flow is collected through the air flow meter, so that an air flow sequence is obtained，/>, wherein ,/>The air flow corresponding to the 1 st detection moment of the powder tank truck in any time period is +.>For the air flow corresponding to the 2 nd detection moment of the powder tank truck in the time period, the air flow is +.>The air flow corresponding to the mth detection moment in the period of the powder tank truck is obtained, and m is the total number of detection moments in the period of the powder tank truck.
Because the powder dry materials with the particle diameter not more than 0.1mm are stored in the tank body of the powder tank truck; the temperature of the external environment can influence the pressure in the tank body, and when the temperature of the external environment is higher, the gas sucked into the tank body can cause the temperature in the tank body to rise at the moment, so that the pressure in the tank body is further caused to rise, and the safety of the tank body is further influenced to a certain extent; if the heat insulation performance of the tank body is good, the temperature of the external environment cannot cause excessive influence on the tank body within a certain temperature range, but once the temperature of the external environment is higher than the temperature range, the safety of the tank body is still threatened, the corresponding performance conditions of different powder tank trucks are different, and the influence degree of the temperature of the external environment on the tank body is also different; therefore, under the same temperature of the external environment, the influence of the temperature of the external environment on the safety of the tank body can show different influence degrees due to the difference of the tank body.
In the embodiment, a temperature sensor is arranged outside the tank body, and the temperature sensor is used for collecting the temperature corresponding to the external environment to obtain a temperature sequence corresponding to each time period，/>, wherein ,/>For the temperature corresponding to the 1 st detection moment of the external environment in any time period,/for the external environment>The temperature corresponding to the 2 nd detection moment in the time period is the temperature of the external environment; />And (3) for the temperature of the external environment corresponding to the mth detection time in the time period, wherein m is the total number of detection times in the time period.
It should be noted that, the collection of the information is all in a plurality of historic unloading processes corresponding to different powder tank trucks; collecting information from the beginning of unloading; one history of the discharge process includes a plurality of time periods, and the time length of the time period in this embodiment is 10 seconds.
When the pressure, the air flow and the temperature are collected, the time interval between two adjacent detection moments in the time period is 0.5 seconds, and when the vibration acceleration is collected, the time interval between two adjacent detection moments in the time period is 0.1 seconds, and an implementer can adjust the time length corresponding to the time period and the time interval between two adjacent detections according to actual conditions.
Step 2, calculating a variance for every z vibration acceleration in the vibration acceleration sequences according to the sequence from front to back, and obtaining a new vibration sequence corresponding to each vibration acceleration sequence according to the variance; wherein z is greater than 2.
Specifically, z takes 10, and an implementer can adjust the value of z; i.e. every 10 oscillations in the sequence of oscillation accelerations in front to back orderCalculating a variance of the dynamic acceleration to obtain a new vibration sequence corresponding to the vibration acceleration sequence，, wherein ,/>For the variance calculated from the 1 st to 10 th vibration acceleration in the vibration acceleration sequence,/->For the variance calculated from the 11 th to 20 th vibration acceleration in the vibration acceleration sequence,/->The variance calculated from the kth-10 vibration acceleration to the kth vibration acceleration in the vibration acceleration sequence. The new vibration sequence can represent the shaking condition of the air compressor.
It should be noted that, in the present embodiment, when vibration acceleration is collected, the time interval between two adjacent detection moments in the time period is 0.1 second, and when pressure, air flow and temperature are collected, the time interval between two adjacent detection moments in the time period is 0.5 second; in addition, in the embodiment, every 10 vibration acceleration is selected to calculate a variance, so that the number of variances in the obtained new vibration sequence is consistent with the number of detection moments in a time period when the pressure, the air flow and the temperature are acquired, namely, the number of variances in the new vibration sequence is m.
Step 3, calculating stability evaluation corresponding to each time period based on the pressure sequence, the air flow sequence and the new vibration sequence; and recording Euclidean distances of any two stability evaluations as difference distances corresponding to the two stability evaluations, and grouping all time periods based on the difference distances to obtain a plurality of groups.
When the powder tank truck reaches a designated position, firstly, an air compressor needs to be started to increase the pressure in the tank body to about 0.18Mpa, and then the unloading operation is carried out; in the normal unloading process, the pointer of the air compressor is maintained in a stable state, namely, the pressure in the tank body is maintained at about 0.18 Mpa; at the same time, the air flow rate and the vibration acceleration are maintained in a steady state as well, so in this embodiment, the stability evaluation corresponding to each time zone is calculated from the pressure sequence, the air flow rate sequence, and the new vibration sequence.
Specifically, the stability was evaluated as:
wherein ,for stability evaluation; />Is a pressure sequence; />Is a new vibration sequence; />Is a sequence of air flow;as a function of the maximum value; />As a function of the average; />As a function of variance; />Is an exponential function based on a natural constant e.
Characterizing the stability of the air flow during the corresponding period,/->The larger the value of the corresponding time period is, the more unstable the air flow in the corresponding time period is, and the smaller the value of the stability evaluation in the corresponding time period is; />Characterizing the degree of stability of the vibration acceleration in the corresponding time period,/->The larger the value of (2) is, the more unstable the vibration acceleration in the corresponding time period is, and the smaller the value of the stability evaluation in the corresponding time period is; />Characterizing the degree of stability of the pressure in the corresponding time period, +.>The value of (2) and->The closer the value of (c) is, the more stable the pressure in the corresponding time zone is, and the larger the value of the stability evaluation in the corresponding time zone is.
And then obtaining the difference distance between any two stability evaluations, and grouping all time periods according to the difference distance to obtain a plurality of groups.
Specifically, grouping all time periods by adopting a DBSCAN clustering algorithm to obtain a plurality of groups; in the grouping, the present embodiment sets the search radius eps to 0.1, and the minimum value mints in the cluster to 5, that is, at least 5 time periods are included in one group; in the actual operation process, the implementer can set the searching radius and the value of the minimum value in the cluster according to the actual situation. The DBSCAN clustering algorithm is a well-known technique and will not be described in detail.
Before grouping all the time periods, marking all the time periods, wherein the marked content is the discharging process and discharging time corresponding to each time period.
Further, in order to accurately perform corresponding treatment on the powder tank truck in the subsequent process, the embodiment further comprises obtaining the types of each group. The types include a normal group, an air compressor failure group, a temperature abnormality group and a blowpipe blockage group.
Specifically, carrying out averaging treatment on the new vibration sequences, the air flow sequences and the temperature sequences corresponding to all time periods in each group to obtain average vibration, average air flow and average temperature in each group corresponding to each group; when the average vibration in the group is larger than the vibration threshold value, the group is an air compressor fault group; when the average air flow in the group is smaller than the air flow threshold value, the group is a blowing-assisting pipe blocking group; when the average temperature in the group is greater than the temperature threshold, the group is a temperature abnormal group; wherein, vibration threshold value, air flow threshold value and temperature threshold value all are set according to actual conditions by the practitioner.
Step 4, obtaining a stability evaluation sequence and a temperature change sequence corresponding to each group; and calculating the correlation corresponding to each group based on the stability evaluation sequence and the temperature change sequence.
The method for acquiring the stability evaluation sequences and the temperature change sequences corresponding to each group specifically comprises the following steps: the stability evaluation corresponding to each time period in the group forms a stability evaluation sequence corresponding to the group; and calculating the average temperature corresponding to the temperature sequence of each time period in the group, wherein the average temperature corresponding to each time period in the group forms the temperature change sequence corresponding to the group.
The correlation among the above is:
wherein ,for relevance, ->For the temperature change sequence,/->For stability evaluation sequence, ++>Is->And->Pearson correlation coefficient of (b); />As a function of the maximum value; />Function for minimum;as a function of the variance.
The correlation represents the influence degree of temperature change in the corresponding time period on the stability in the unloading process; when the value of the correlation is close to 0, the influence degree of the current temperature change on the stability in the unloading process is smaller or no influence is caused; when the value of the correlation is close to 1, the influence degree of the current temperature change on the stability in the unloading process is larger. The purpose of calculating the correlation here is: the temperature change within a certain range does not cause obvious influence on the current unloading process, and once the temperature exceeds the range, stronger correlation can appear, namely the influence degree of the temperature change on the stability in the unloading process is larger, thereby influencing the judgment of the health degree in the unloading process in the subsequent step.
It should be noted that, the pearson phase relationship is a known technology and is not in the protection scope of the present application, and will not be described in detail.
And 5, calculating the health degree corresponding to each time period according to the stability evaluation and the average temperature corresponding to each time period and the correlation corresponding to the group to which each time period belongs.
The health degree is as follows:
wherein ,for health degree, the->For stability evaluation; />Is a temperature sequence; />Is a correlation; />As a hyperbolic tangent function; />As a function of the average; />For adjusting parameters, and->, in the formula ,/>Is the optimum temperature. The optimum temperature in this example is 25 ℃, and the practitioner can adjust according to the actual situation.
The comprehensive evaluation of the discharging process corresponding to the time period is represented by the health degree, and the larger the value of the health degree is, the higher the comprehensive evaluation of the discharging process corresponding to the time period is, namely, all indexes of the discharging process are in a stable health state; conversely, the lower the comprehensive evaluation of the discharging process corresponding to the time period is, namely, each index of the discharging process is not stable enough; comprehensively judging with the temperature change condition according to the correlation, and determining the proper condition of the temperature in the unloading process so as to obtain the health degree in the unloading process; and completing the comprehensive evaluation of the discharging process corresponding to the time period. The effect of introducing the correlation in calculating the health degree is that the sensitivity degree of different tank trucks and different types of faults to the temperature is different, and the comprehensive result is corrected by introducing the correlation; the obtained health degree is more accurate.
Step 6, taking a pressure sequence, a new vibration sequence, an air flow sequence, a temperature sequence, a correlation and a health degree corresponding to each time period as samples, and training an LSTM network by using the samples to obtain a trained LSTM network; and obtaining the predicted health degree corresponding to the future time period after the current time period according to the trained LSTM network, and judging whether the powder tank truck is abnormal in the future time period according to the predicted health degree.
Specifically, the method for acquiring the corresponding overall loss function of the LSTM network in the training process comprises the following steps: firstly, randomly selecting a time period in any one group, and calculating membership corresponding to the time period based on the difference distance between stability evaluation corresponding to the time period and stability evaluation corresponding to other time periods remained in the group; obtaining the membership degree corresponding to each time period in the group, and taking the time period corresponding to the maximum membership degree as the representative time period of the group; and thus the representative time periods of the respective groups.
The membership obtaining method comprises the following steps: and calculating the accumulated sum of the difference distances between the stability evaluation corresponding to any one time period in the group and the stability evaluation corresponding to other time periods remaining in the group, calculating the ratio of the accumulated sum to the total number of the time periods in the group minus 1, and recording the value normalized by the ratio as the membership degree of the time period. The greater the membership, the more representative the time period is within the group; the smaller the membership, the less representative the time period is, and the weaker the representativeness is.
The calculation formula of the membership degree is specifically as follows:, wherein ,/>For the membership of time period A, +.>Evaluation of stability for time period AA difference distance between stability evaluations corresponding to time period X; n is the total number of time periods in the group; />Is a cosine function; />Is the circumference ratio. In the embodiment, the cosine function is adopted for normalization, so that the membership degree and the difference distance are in a negative correlation relationship, and the smaller the difference distance is, the larger the membership degree is; the implementer may also choose other functions to normalize.
Then calculating the accumulated sum of the difference distances between the stability evaluation corresponding to the representative time period of the group and the stability evaluation corresponding to the representative time periods of the rest other groups to obtain the inter-group difference corresponding to the group; and further obtaining the difference between the groups corresponding to the groups.
Finally, arranging all the time periods according to the content marked in advance to obtain a plurality of discharging sequences; according to the group-to-group difference of the groups to which each time period belongs in the unloading sequence, calculating the weight corresponding to each time period in the unloading sequence and the sequence weight corresponding to the unloading sequence; and obtaining an overall loss function according to the weight and the sequence weight.
The weight acquisition method comprises the following steps: the normalized value of the inter-group difference corresponding to the group to which each time period belongs is recorded as the weight corresponding to each time period; the normalization not only enables the value of the weight corresponding to each time period in the same unloading sequence to be between 0 and 1, but also enables the sum of the weights corresponding to all the time periods in one unloading sequence to be 1.
The acquisition method of the sequence weight comprises the following steps: calculating the accumulated sum of the differences among the groups corresponding to the groups to which all the time periods belong in the unloading sequence; and the accumulated and normalized value is recorded as the sequence weight corresponding to the unloading sequence; a discharging sequence has a sequence weight, the value of the sequence weight is between 0 and 1, and the sum of all the sequence weights is 1.
The overall loss function is:
wherein ,for the whole loss function->Is the total number of discharge sequences; />Is the total number of time periods in the discharge sequence; />The sequence weight corresponding to the nth unloading sequence; />The weight corresponding to the ith time period in the nth unloading sequence;is the loss corresponding to the ith time period in the nth discharge sequence.
The LSTM network is input into a pressure sequence, a new vibration sequence, an air flow sequence, a temperature sequence, a correlation and a health degree corresponding to a time period; the outputs of the LSTM network are a pressure sequence, an airflow sequence, and a new vibration sequence corresponding to the future time period for the predicted health corresponding to the future time period.
It should be noted that, according to the mean square error loss function, the loss corresponding to each time period is calculated; training of LSTM networks is a well-known technique and will not be described in detail.
According to the trained LSTM network, the method for obtaining the predicted health degree corresponding to the future time period after the current time period comprises the following specific steps: and inputting the current time period and the pressure sequence, the new vibration sequence, the air flow sequence, the temperature sequence, the correlation and the health degree corresponding to each time period before the current time period into a trained LSTM network, and outputting the predicted health degree corresponding to the future time period after the current time period and the pressure sequence, the air flow sequence and the new vibration sequence corresponding to the future time period.
And comparing the predicted health degree with a health threshold value, and when the predicted health degree is smaller than the health threshold value, giving an abnormal early warning, and taking corresponding measures by an implementer to process the unloading process of the powder tank truck, namely carrying out pressure relief operation or shutdown treatment in advance, so as to avoid damage to the air conditioner and danger. When the predicted health degree is greater than the health threshold value, the powder tank truck is indicated to be free of abnormality in a future time period. In this embodiment, the health threshold has a value of 0.8, and the practitioner can adjust the health threshold during the actual operation.
Further, in order to more accurately take corresponding measures on the powder tank truck, when the predicted health degree is smaller than the health threshold value, a group corresponding to the future time period is obtained according to a pressure sequence, an air flow sequence and a new vibration sequence corresponding to the future time period, and the abnormal type of the powder tank truck in the future time period is obtained according to the category corresponding to the group to which the future time period belongs; the method can remind related staff of abnormal types of the powder tank truck in a future time period, provides enough preparation time for the related staff, and further accurately takes corresponding measures on the powder tank truck, and can protect the air compressor and the powder tank truck to the greatest extent.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the application and are intended to be included within the scope of the application.

Claims (7)

1. An abnormality early warning method in the unloading process of a powder tank truck is characterized by comprising the following steps:
collecting pressure, vibration acceleration and air flow corresponding to each detection moment of the powder tank truck in a plurality of time periods; obtaining a pressure sequence, a vibration acceleration sequence and an air flow sequence corresponding to each time period; acquiring the temperature of the external environment corresponding to each detection moment in a plurality of time periods to obtain a temperature sequence corresponding to each time period;
in each vibration acceleration sequence, calculating a variance for every z vibration acceleration according to the sequence from front to back, and obtaining a new vibration sequence corresponding to each vibration acceleration sequence according to the variance; wherein z is greater than 2;
calculating stability evaluation corresponding to each time period based on the pressure sequence, the air flow sequence and the new vibration sequence;
recording Euclidean distances of any two stability evaluations as difference distances corresponding to the two stability evaluations, and grouping all time periods based on the difference distances to obtain a plurality of groups;
acquiring a stability evaluation sequence and a temperature change sequence corresponding to each group; calculating the correlation corresponding to each group based on the stability evaluation sequence and the temperature change sequence;
calculating the health degree corresponding to each time period according to the stability evaluation and the average temperature corresponding to each time period and the correlation corresponding to the group to which each time period belongs;
taking a pressure sequence, a new vibration sequence, an air flow sequence, a temperature sequence, correlation and health degree corresponding to each time period as samples, and training an LSTM network by using the samples to obtain a trained LSTM network;
obtaining a predicted health degree corresponding to a future time period after the current time period according to the trained LSTM network, and judging whether the powder tank truck is abnormal in the future time period according to the predicted health degree;
the stability evaluation was:
wherein ,for stability evaluation; />Is a pressure sequence; />Is a new vibration sequence; />Is a sequence of air flow; />As a function of the maximum value; />As a function of the average; />As a function of variance; />Is an exponential function based on a natural constant e.
2. The method for early warning of abnormality in unloading process of powder tank truck according to claim 1, wherein the method for obtaining the stability evaluation sequence and the temperature change sequence corresponding to each group specifically comprises: the stability evaluation corresponding to each time period in the group forms a stability evaluation sequence corresponding to the group; and calculating the average temperature corresponding to the temperature sequence of each time period in the group, wherein the average temperature corresponding to each time period in the group forms the temperature change sequence corresponding to the group.
3. The method for early warning of abnormality in the unloading process of a powder tank truck according to claim 1, wherein the method for acquiring the correlation is as follows:
wherein ,for relevance, ->For the temperature change sequence,/->For stability evaluation sequence, ++>Is thatAnd->Pearson correlation coefficient of (b); />As a function of the maximum value; />Function for minimum;as a function of the variance.
4. The method for early warning of abnormality in the unloading process of a powder tank truck according to claim 1, wherein the health degree is:
wherein ,for health degree, the->For stability evaluation; />Is a temperature sequence; />Is a correlation; />As a hyperbolic tangent function; />As a function of the average; />For adjusting parameters, and->, in the formula ,/>Is the optimum temperature.
5. The method for early warning of abnormality in the unloading process of a powder tank truck according to claim 1, wherein the method for acquiring the corresponding integral loss function of the LSTM network in the training process is as follows: randomly selecting a time period in any one group, and calculating the membership degree corresponding to the time period based on the difference distance between the stability evaluation corresponding to the time period and the stability evaluation corresponding to the rest of the time periods in the group; obtaining the membership degree corresponding to each time period in the group, and taking the time period corresponding to the maximum membership degree as the representative time period of the group; obtaining the representative time periods of each group;
calculating the accumulated sum of the difference distances between the stability evaluation corresponding to the representative time period of the group and the stability evaluation corresponding to the representative time periods of the rest other groups to obtain the inter-group difference corresponding to the group; thereby obtaining the difference between the groups corresponding to each group;
arranging all the time periods according to the content marked in advance to obtain a plurality of discharging sequences; according to the group-to-group difference of the groups to which each time period belongs in the unloading sequence, calculating the weight corresponding to each time period in the unloading sequence and the sequence weight corresponding to the unloading sequence; obtaining an overall loss function according to the weight and the sequence weight;
the overall loss function is:
wherein ,for the whole loss function->Is the total number of discharge sequences; />Is the total number of time periods in the discharge sequence;the sequence weight corresponding to the nth unloading sequence; />The weight corresponding to the ith time period in the nth unloading sequence;is the loss corresponding to the ith time period in the nth discharge sequence.
6. The method for early warning of abnormality in unloading a powder tank truck according to claim 5, wherein the membership obtaining method is as follows: and calculating the accumulated sum of the difference distances between the stability evaluation corresponding to any one time period in the group and the stability evaluation corresponding to other time periods remaining in the group, calculating the ratio of the accumulated sum to the total number of the time periods in the group minus 1, and recording the value normalized by the ratio as the membership degree of the time period.
7. The method for early warning of abnormality in unloading a powder tank truck according to claim 5, wherein the weight obtaining method is as follows: the normalized value of the inter-group difference corresponding to the group to which each time period belongs is recorded as the weight corresponding to each time period;
the acquisition method of the sequence weight comprises the following steps: calculating the accumulated sum of the differences among the groups corresponding to the groups to which all the time periods belong in the unloading sequence; and the accumulated and normalized value is recorded as the sequence weight corresponding to the unloading sequence.
CN202310814665.5A 2023-07-05 2023-07-05 Abnormal early warning method in unloading process of powder tank truck Active CN116579663B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310814665.5A CN116579663B (en) 2023-07-05 2023-07-05 Abnormal early warning method in unloading process of powder tank truck

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310814665.5A CN116579663B (en) 2023-07-05 2023-07-05 Abnormal early warning method in unloading process of powder tank truck

Publications (2)

Publication Number Publication Date
CN116579663A CN116579663A (en) 2023-08-11
CN116579663B true CN116579663B (en) 2023-10-20

Family

ID=87536081

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310814665.5A Active CN116579663B (en) 2023-07-05 2023-07-05 Abnormal early warning method in unloading process of powder tank truck

Country Status (1)

Country Link
CN (1) CN116579663B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116881673B (en) * 2023-09-06 2023-12-19 山东济矿鲁能煤电股份有限公司阳城煤矿 Shield tunneling machine operation and maintenance method based on big data analysis
CN117433591B (en) * 2023-12-21 2024-02-27 河北省农林科学院农业信息与经济研究所 Automatic equipment monitoring system based on multi-sensor data

Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE19904219A1 (en) * 1998-07-16 2000-01-20 Continental Teves Ag & Co Ohg Detection of critical cases when motor vehicles are likely to overturn
CN109583570A (en) * 2018-11-30 2019-04-05 重庆大学 The method for determining bridge health monitoring system abnormal data source based on deep learning
CN110059844A (en) * 2019-02-01 2019-07-26 东华大学 Energy storage device control method based on set empirical mode decomposition and LSTM
CN113507150A (en) * 2021-06-29 2021-10-15 连君昊 Energy storage battery control method and control system
CN114252749A (en) * 2022-02-28 2022-03-29 国网湖北省电力有限公司检修公司 Transformer partial discharge detection method and device based on multiple sensors
CN114638280A (en) * 2022-01-30 2022-06-17 江苏东佳电气有限公司 Explosion-proof motor bearing temperature anomaly monitoring system based on local equipment network
CN114925943A (en) * 2022-07-22 2022-08-19 江苏布罗信息技术有限公司 Transformer abnormal state prediction method based on power Internet of things
CN115086795A (en) * 2022-05-31 2022-09-20 国网河南省电力公司南召县供电公司 RFID data acquisition device and method for acceptance of unmanned aerial vehicle
CN115290368A (en) * 2022-09-30 2022-11-04 南通润厚设备工程有限公司 Fault diagnosis method in unloading process of powder tanker
CN115420437A (en) * 2022-11-07 2022-12-02 南京邮电大学 Ultrasonic leak detection device based on deep learning
CN115440242A (en) * 2022-09-02 2022-12-06 天津市恒一机电科技有限公司 Method for detecting longitudinal tearing of conveying belt and related equipment
CN115615540A (en) * 2022-12-20 2023-01-17 潍坊百特磁电科技有限公司 Carrier roller fault identification method, equipment and medium of permanent magnet self-discharging iron remover

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113987834B (en) * 2021-11-15 2022-07-15 华东交通大学 CAN-LSTM-based railway train bearing residual life prediction method

Patent Citations (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE19904219A1 (en) * 1998-07-16 2000-01-20 Continental Teves Ag & Co Ohg Detection of critical cases when motor vehicles are likely to overturn
CN109583570A (en) * 2018-11-30 2019-04-05 重庆大学 The method for determining bridge health monitoring system abnormal data source based on deep learning
CN110059844A (en) * 2019-02-01 2019-07-26 东华大学 Energy storage device control method based on set empirical mode decomposition and LSTM
CN113507150A (en) * 2021-06-29 2021-10-15 连君昊 Energy storage battery control method and control system
CN114638280A (en) * 2022-01-30 2022-06-17 江苏东佳电气有限公司 Explosion-proof motor bearing temperature anomaly monitoring system based on local equipment network
CN114252749A (en) * 2022-02-28 2022-03-29 国网湖北省电力有限公司检修公司 Transformer partial discharge detection method and device based on multiple sensors
CN115086795A (en) * 2022-05-31 2022-09-20 国网河南省电力公司南召县供电公司 RFID data acquisition device and method for acceptance of unmanned aerial vehicle
CN114925943A (en) * 2022-07-22 2022-08-19 江苏布罗信息技术有限公司 Transformer abnormal state prediction method based on power Internet of things
CN115440242A (en) * 2022-09-02 2022-12-06 天津市恒一机电科技有限公司 Method for detecting longitudinal tearing of conveying belt and related equipment
CN115290368A (en) * 2022-09-30 2022-11-04 南通润厚设备工程有限公司 Fault diagnosis method in unloading process of powder tanker
CN115420437A (en) * 2022-11-07 2022-12-02 南京邮电大学 Ultrasonic leak detection device based on deep learning
CN115615540A (en) * 2022-12-20 2023-01-17 潍坊百特磁电科技有限公司 Carrier roller fault identification method, equipment and medium of permanent magnet self-discharging iron remover

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Rolling Bearing Initial Fault Detection Using Long Short-Term Memory Recurrent Network;HUAITAO SHI 等;IEEE Access;第7卷;第171559-171569页 *
矿用自卸车异常振动及其改进的试验研究;盛云 等;农业装备与车辆工程(第4期);第72-76页 *

Also Published As

Publication number Publication date
CN116579663A (en) 2023-08-11

Similar Documents

Publication Publication Date Title
CN116579663B (en) Abnormal early warning method in unloading process of powder tank truck
US6799951B2 (en) Compressor degradation detection system
CN115290368B (en) Fault diagnosis method in unloading process of powder tanker
US8471702B2 (en) Method and system for compressor health monitoring
US9862495B2 (en) Aircraft air-conditioning heat exchanger contamination detection
US7603242B2 (en) Fuel leak estimator
CN116735804A (en) Intelligent sensor precision monitoring system based on Internet of things
WO2017175678A1 (en) Abnormality detection device, abnormality detection method, and abnormality detection system
US7461544B2 (en) Methods for detecting water induction in steam turbines
EP3648237A1 (en) System and method for detecting battery cell swelling
CN110582626B (en) System and method for monitoring a turbomachine with anomaly detection corrected by a wear factor
CN109707653A (en) A kind of Long-distance Transmission Pipeline Systems of Centrifugal Compressor Unit intelligent monitor system
JP6806576B2 (en) Anomaly detection method for traveling motors for railway vehicles
US20140229121A1 (en) Systems and methods for determining a level of fouling of compressors
CN112197167B (en) Vehicle-mounted hydrogen storage bottle air tightness early warning monitoring system
EP1792158A1 (en) Fuel leak estimator
JP2014043795A (en) Surge generation preventing device and compressor incorporating the same
US20130246002A1 (en) Method of measuring health index of plant in which condition of lower level component is reflected and computer-readable storage medium in which program to perform the method is stored
US20150110650A1 (en) Exhaust-gas turbocharger
CN116610915A (en) Pneumatic ash conveying abnormal state diagnosis and fault early warning method
EP3179165B1 (en) Steam using facility management method, and steam using facility
JP2018137967A (en) Method for detecting abnormality of compressor for railway vehicle
CN115355729B (en) On-line monitoring method for gas mixture and non-condensable gas of condenser and vacuum system
US11378943B2 (en) Information processing device, information processing method, and program
CN117734347B (en) Hub unit, monitoring method and application thereof

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