CN117570039B - Mining flameproof submersible sand discharge electric pump fault monitoring system and method - Google Patents
Mining flameproof submersible sand discharge electric pump fault monitoring system and method Download PDFInfo
- Publication number
- CN117570039B CN117570039B CN202410057857.0A CN202410057857A CN117570039B CN 117570039 B CN117570039 B CN 117570039B CN 202410057857 A CN202410057857 A CN 202410057857A CN 117570039 B CN117570039 B CN 117570039B
- Authority
- CN
- China
- Prior art keywords
- electric pump
- data
- fault
- signal
- wireless
- 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
- 238000000034 method Methods 0.000 title claims abstract description 93
- 238000012544 monitoring process Methods 0.000 title claims abstract description 34
- 238000005065 mining Methods 0.000 title claims abstract description 29
- 239000004576 sand Substances 0.000 title claims abstract description 17
- 238000012545 processing Methods 0.000 claims abstract description 56
- 238000012423 maintenance Methods 0.000 claims abstract description 52
- 238000004891 communication Methods 0.000 claims abstract description 37
- 238000007405 data analysis Methods 0.000 claims abstract description 19
- 238000013500 data storage Methods 0.000 claims abstract description 18
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 68
- 230000005540 biological transmission Effects 0.000 claims description 40
- 238000003745 diagnosis Methods 0.000 claims description 29
- 230000006870 function Effects 0.000 claims description 26
- 230000008569 process Effects 0.000 claims description 24
- 230000035945 sensitivity Effects 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 238000012706 support-vector machine Methods 0.000 claims description 15
- 238000007418 data mining Methods 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 13
- 230000000694 effects Effects 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 230000002159 abnormal effect Effects 0.000 claims description 5
- 238000003066 decision tree Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000007637 random forest analysis Methods 0.000 claims description 5
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000012856 packing Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 2
- 239000004820 Pressure-sensitive adhesive Substances 0.000 claims 1
- 238000002790 cross-validation Methods 0.000 claims 1
- 230000005611 electricity Effects 0.000 claims 1
- 238000010606 normalization Methods 0.000 claims 1
- 239000013049 sediment Substances 0.000 claims 1
- 238000004458 analytical method Methods 0.000 description 9
- 238000005457 optimization Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 7
- 230000006872 improvement Effects 0.000 description 7
- 239000013598 vector Substances 0.000 description 7
- 239000002184 metal Substances 0.000 description 6
- 229910052751 metal Inorganic materials 0.000 description 6
- 238000007599 discharging Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 5
- 238000005299 abrasion Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 150000002739 metals Chemical class 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 238000001228 spectrum Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 210000000476 body water Anatomy 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011478 gradient descent method Methods 0.000 description 2
- 238000013021 overheating Methods 0.000 description 2
- 238000010183 spectrum analysis Methods 0.000 description 2
- 230000005678 Seebeck effect Effects 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 239000002817 coal dust Substances 0.000 description 1
- 239000004020 conductor Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000011499 joint compound Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F04—POSITIVE - DISPLACEMENT MACHINES FOR LIQUIDS; PUMPS FOR LIQUIDS OR ELASTIC FLUIDS
- F04D—NON-POSITIVE-DISPLACEMENT PUMPS
- F04D15/00—Control, e.g. regulation, of pumps, pumping installations or systems
- F04D15/0088—Testing machines
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/211—Selection of the most significant subset of features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Mechanical Engineering (AREA)
- Control Of Non-Positive-Displacement Pumps (AREA)
Abstract
The invention discloses a fault monitoring system and method for a mining flameproof submersible sand discharge electric pump, belonging to the technical field of monitoring systems, and comprising the following steps: the sensor is used for acquiring the operation parameters and state information of the electric pump in real time; the wireless communication module is used for transmitting the acquired data to a data receiver of the wellhead through a wireless signal; a data receiver for receiving data from the wireless communication module; the data processing unit is used for analyzing and modeling the received data; the display unit is used for providing the running state and fault alarm of the electric pump, and fault reasons and maintenance suggestions for operators and maintenance personnel; and the data storage unit is used for storing the operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction. The invention is used for solving the technical problem that the fault monitoring and diagnosing method of the mining explosion-proof submersible sand discharge electric pump in the prior art cannot meet the requirements of safety and efficiency of mine drainage.
Description
Technical Field
The invention belongs to the technical field of monitoring systems, and particularly relates to a fault monitoring system and method for a mining explosion-proof submersible sand discharge electric pump.
Background
An electric pump is a device for driving an impeller to rotate by using electric energy so as to convey liquid to a required place, and is widely applied to the fields of industry, agriculture, construction, mine and the like. The performance and the service life of the electric pump are closely related to the running state and the fault condition of the electric pump, so that the running parameters and the state information of the electric pump are monitored and analyzed in real time, and faults of the electric pump are timely found and diagnosed, so that the electric pump is an important means for ensuring the safe, efficient and stable running of the electric pump.
The explosion-proof submerged sand-discharging electric pump for mine is a special electric pump for discharging water and sand in mine, and has the characteristics of compact structure, reliable performance, safety, explosion prevention and the like, and is one of main equipment for mine drainage. Because the mining explosion-proof submersible sand-removing electric pump works in a severe environment, the mining explosion-proof submersible sand-removing electric pump is often corroded and impacted by substances such as water, sand, mud, coal dust and the like, faults such as motor overheating, impeller blockage, bearing abrasion, water seal damage, cable breakage and the like easily occur, the normal operation of the electric pump is influenced, and even the electric pump is damaged and accidents of a mine are caused. Therefore, the operation state and faults of the mining flameproof submersible sand-discharging electric pump are effectively monitored and diagnosed, and the method is a key for guaranteeing the safety and efficiency of mine drainage.
At present, the fault monitoring and diagnosing methods for the mining flameproof submersible sand discharge electric pump mainly comprise the following steps:
manual inspection method: the method is that operators regularly observe and check the operation condition of the electric pump, and judge whether the electric pump has faults, such as temperature, sound, vibration and the like according to experience. The method has the advantages of simplicity, easiness, no need of additional equipment and technology, low efficiency, poor accuracy, incapability of realizing real-time monitoring and remote control of the electric pump, incapability of timely finding and processing faults of the electric pump and easiness in damaging the electric pump and accidents of a mine.
The instrument method comprises the following steps: the method is to measure and record the running parameters and state information of the electric pump by using various instruments and meters, such as a thermometer, an ammeter, a vibrator, a water level meter and the like, and then judge whether the electric pump has faults, such as whether the temperature, the current, the vibration, the water level and the like exceed the normal range or not according to the measurement results. The method has the advantages that the operation condition of the electric pump can be accurately reflected, but the method has the defects that the installation and maintenance of instruments are complex, professional personnel and equipment are needed, the measurement results of the instruments are often discrete and static, the continuous monitoring and dynamic analysis of the electric pump cannot be realized, and the fault reasons and maintenance suggestions of the electric pump cannot be provided.
The signal processing method comprises the following steps: the method utilizes the signal processing technology to analyze and process the operation data of the electric pump, such as time domain analysis, frequency domain analysis, time-frequency domain analysis and the like, extracts characteristic signals reflecting the operation condition and fault condition of the electric pump, such as waveforms, amplitude, frequency, energy, frequency spectrum and the like of temperature, current, vibration, water level and the like, and judges whether the electric pump has faults, such as whether the waveforms, amplitude, frequency, energy, frequency spectrum and the like of temperature, current, vibration, water level and the like are abnormal or not according to the characteristic signals. The method has the advantages that the operation condition of the electric pump can be comprehensively reflected, but has the defects that the signal processing technology is complex, professional personnel and equipment are needed, the signal processing result is often qualitative and subjective, the quantitative evaluation and objective diagnosis of the electric pump cannot be realized, and the fault reason and maintenance suggestion of the electric pump cannot be provided.
In summary, the existing fault monitoring and diagnosing methods for the mining flameproof submersible sand-discharging electric pump have certain defects and cannot meet the requirements for safety and efficiency of mine drainage, so that a system and a method for realizing real-time monitoring and intelligent diagnosis on the running state and faults of the mining flameproof submersible sand-discharging electric pump are urgently needed, performance and service life of the electric pump are improved, and safety and efficiency of mine drainage are guaranteed.
Disclosure of Invention
Aiming at the problems, the invention provides a fault monitoring system and a fault monitoring method for a mining flameproof submersible sand discharge electric pump, which are used for solving the technical problem that the fault monitoring and diagnosing method for the mining flameproof submersible sand discharge electric pump in the prior art cannot meet the requirements on safety and efficiency of mine drainage.
In order to achieve the above purpose, the invention adopts the following technical scheme:
a mining flameproof submersible sand discharge electric pump fault monitoring system, comprising:
the sensor is arranged on a key component of the electric pump and is used for acquiring the operation parameters and state information of the electric pump in real time, including temperature, current, vibration and water level information;
the wireless communication module is connected with the sensor and used for sending the acquired data to a data receiver of the wellhead through a wireless signal;
the data receiver is arranged at the wellhead and is used for receiving data from the wireless communication module;
the data processing unit is connected with the data receiver and used for analyzing and modeling the received data, extracting normal working characteristics and fault characteristics of the electric pump and establishing a fault diagnosis model of the electric pump;
the display unit is connected with the data processing unit and used for providing the running state and fault alarm of the electric pump, and fault reasons and maintenance suggestions for operators and maintenance personnel;
The data storage unit is connected with the data processing unit and used for storing the operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction;
compared with the prior art, the invention has the beneficial effects that: the running parameters and state information of the electric pump, such as temperature, current, vibration, water level and the like, are acquired in real time through the sensor, so that the running condition of the electric pump can be comprehensively reflected, the abnormal phenomenon of the electric pump can be timely found, and the monitoring precision and instantaneity of the electric pump are improved;
the wireless communication module is used for transmitting the acquired data to the data receiver of the wellhead through the wireless signal, so that the limitation of a cable can be overcome, the remote monitoring and control of the electric pump are realized, and the communication efficiency and safety of the electric pump are improved;
the data processing unit is used for analyzing and modeling the received data, extracting normal working characteristics and fault characteristics of the electric pump, establishing a fault diagnosis model of the electric pump, realizing intelligent diagnosis and prediction of the electric pump by using data mining and machine learning methods, and improving diagnosis accuracy and reliability of the electric pump;
the display unit is used for providing the running state and fault alarm, fault reason and maintenance suggestion of the electric pump for operators and maintenance personnel, and friendly interaction and guidance of the electric pump can be realized by using a graphical and phonetic mode, so that the maintenance efficiency and quality of the electric pump are improved;
The operation data and fault information of the electric pump are stored in the database through the data storage unit and used for subsequent data analysis and fault prediction, the technology of the relational database can be utilized to realize the inquiry, statistics, analysis and prediction of the historical data of the electric pump, and the data management and optimization capacity of the electric pump are improved.
As a further improvement of the scheme, the sensors are a temperature sensor, a current sensor, a vibration sensor and a water level sensor, and are used for collecting temperature, current, vibration, water level parameters and state information of the electric pump in real time;
the temperature sensor is used for measuring the temperature of a motor, an impeller, a bearing, a pump body and other parts of the electric pump, the temperature is one of the operation parameters of the electric pump, and an output signal of the temperature sensor is a voltage signal and is in direct proportion to the temperature, and the temperature sensor can be expressed by the following formula:
wherein->The output voltage of the temperature sensor is T, the temperature of the electric pump is T, and k is the sensitivity coefficient of the temperature sensor;
the current sensor is used for measuring the current of the motor of the electric pump, the current is one of the operation parameters of the electric pump, the output signal of the current sensor is a voltage signal, and the output signal is proportional to the current and can be expressed by the following formula:
Wherein->For the output of the current sensorVoltage (V)>The motor current of the electric pump is represented by k, and the sensitivity coefficient of the current sensor is represented by k;
the output signal of the vibration sensor is a voltage signal, and is related to the amplitude and frequency of vibration, and the output signal can be expressed by the following formula:
wherein->For the output voltage of the vibration sensor, A is the amplitude of the bearing vibration of the electric pump, +.>Is the frequency of the bearing vibration of the electric pump, t is time, < ->And->Is the sensitivity coefficient of the vibration sensor;
the water level sensor is used for measuring the water level of the pump body of the electric pump, the water level is one of the state information of the electric pump, the output signal of the water level sensor is a resistance signal, and the output signal is in direct proportion to the water level and can be expressed by the following formula:
，
wherein,is the output resistance of the water level sensor, +.>Is the pump body water level of the electric pump->Is the water level sensorSensitivity coefficient.
The improved technical effects are as follows: parameters such as temperature, current, vibration, water level and the like of the electric pump and state information can be acquired in real time, the running condition and fault condition of the electric pump can be found in time, the safety and reliability of the electric pump are improved, fault shutdown and maintenance cost are reduced, and the service life of the electric pump is prolonged.
As a further improvement of the scheme, the wireless communication module is a node in the wireless sensor network, and a ZigBee protocol is adopted for transmitting data acquired by the sensor to a data receiver at a wellhead through wireless signals.
The improved technical effects are as follows: through the technology of the wireless sensor network, self-organization, self-configuration, self-repair and self-adaptation among a plurality of wireless communication modules can be realized, and the network stability and expandability of the wireless communication modules are improved;
by the ZigBee protocol technology, the communication with low power consumption, low cost, low complexity and high safety among the wireless communication modules can be realized, and the communication efficiency and safety of the wireless communication modules are improved;
through the wireless signal technology, the remote and wireless data transmission between the wireless communication module and the data receiver can be realized, the limitation of cables is overcome, and the data transmission efficiency and flexibility of the wireless communication module are improved.
As a further improvement of the above scheme, the data receiver is a coordinator in the wireless sensor network, and is configured to receive a wireless signal from the wireless communication module, that is, wireless transmission of operation data of the electric pump, convert the wireless signal into a wired signal, that is, wired transmission of operation data of the electric pump, and connect the wired signal with the data processing unit through the wired communication interface, and transmit the operation data of the electric pump to the data processing unit, so as to implement data acquisition and processing of the electric pump;
The data receiver is a device with wireless receiving and wired transmitting capabilities, can realize conversion and transmission of wireless signals and wired signals, and has the following working principle:
(1) The data receiver receives a wireless signal from the wireless communication module through the wireless antenna, that is, wireless transmission of operation data of the electric pump, which can be represented by the following formula:
，
wherein,for wireless signals, A is the amplitude of the wireless signal, < >>Is the frequency of the radio signal, t is time, < >>Is the phase of the wireless signal;
(2) The data receiver demodulates and decodes the wireless signal through the wireless receiving module, and restores the wireless signal into a digital signal, namely the operation data of the electric pump, wherein the digital signal is a binary signal and can be expressed by the following formula:
，
wherein,digital signal>Is the nth bit of the digital signal, +.>The pulse shape of the digital signal is that T is the bit interval of the digital signal and N is the bit number of the digital signal;
(3) The data receiver encodes and modulates the digital signal through the wired transmitting module, and converts the digital signal into a wired signal, namely, the wired transmission of the operation data of the electric pump, wherein the wired signal is a voltage signal and can be expressed by the following formula:
，
Wherein,is a wired signal>Is the nth symbol of the cable signal, < >>The symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal;
(4) The data receiver sends the wired signal to the data processing unit through the wired communication interface, and the conversion and transmission from the wireless transmission of the operation data of the electric pump to the wired transmission are completed.
The improved technical effects are as follows: and the conversion and transmission from the wireless transmission to the wired transmission of the operation data of the electric pump are completed, and the efficiency and the stability of the data transmission of the data receiver are improved.
As a further improvement of the scheme, the data processing unit is an industrial personal computer or other equipment with data processing capability, is internally provided with data processing software and is used for analyzing and modeling the received operation data of the electric pump, extracting normal working characteristics and fault characteristics of the electric pump and establishing a fault diagnosis model of the electric pump.
The improved technical effects are as follows: the high-speed processing and the large-capacity storage of the operation data of the electric pump can be realized, and the data processing capacity and the data storage capacity of the data processing unit are improved.
As a further improvement of the above solution, the display unit is a display or other equipment with display capability, and a speaker or other equipment with voice broadcasting capability, and is used for providing the operating status and fault alarm, and fault cause and maintenance advice of the electric pump for operators and maintenance personnel, and the display unit is connected with the data processing unit, and receives the operating data, fault type, fault position and maintenance advice information of the electric pump from the data processing unit.
The improved technical effects are as follows: through the display or other equipment with display capability, the operation data, fault type, fault position and maintenance advice information of the electric pump can be graphically displayed, and the display effect and the visibility of the display unit are improved.
As a further improvement of the above solution, the data storage unit is a database server or other device with data storage capability, and is used for storing the operation data and fault information of the electric pump in the database for subsequent data analysis and fault prediction.
The improved technical effects are as follows: the high-efficiency storage and the safety protection of the operation data and the fault information of the electric pump can be realized, and the data storage efficiency and the safety of the data storage unit are improved.
A fault monitoring method for a mining flameproof submersible sand discharge electric pump comprises the following steps:
(1) The sensor is arranged on a key component of the electric pump and used for collecting the operation parameters and state information of the electric pump in real time, including temperature, current, vibration and water level;
(2) Connecting the sensor with a wireless communication module to form a wireless sensor network, and transmitting acquired data to a data receiver of a wellhead through wireless signals;
(3) A data processing unit and a display unit are arranged on a data receiver of the wellhead and are used for receiving and processing data from the wireless sensor network and displaying and storing the running condition and fault information of the electric pump;
(4) In the data processing unit, the operation data of the electric pump is analyzed and modeled by adopting a data mining and machine learning method, the normal working characteristics and the fault characteristics of the electric pump are extracted, and a fault diagnosis model of the electric pump is established;
(5) In the display unit, a graphical and phonetic mode is adopted to provide the operation state and fault alarm of the electric pump, as well as fault reasons and maintenance suggestions for operators and maintenance personnel;
(6) In the data storage unit, the operating data and fault information of the electric pump are stored in a database for subsequent data analysis and fault prediction.
As a further improvement of the above scheme, in the step (4), the data mining and machine learning method is adopted to analyze and model the operation data of the electric pump, extract the normal working characteristics and fault characteristics of the electric pump, and build a fault diagnosis model of the electric pump, which specifically comprises the following sub-steps:
(4.1) preprocessing operation data of the electric pump;
(4.2) extracting characteristics of the operation data of the electric pump;
(4.3) performing feature selection on the operation data of the electric pump;
(4.4) performing classifier training on the operation data of the electric pump, and training a multi-classifier capable of identifying multiple fault types of the electric pump by using the operation characteristics and fault types of the electric pump with labels and adopting a supervised learning method to serve as a fault diagnosis model of the electric pump; the classifier training process is represented by the following formula:
，
wherein,indicating the type of fault that is predicted,representing a function of a multi-classifier,parameters representing multiple classifiers; parameters of multiple classifiersBy minimizing a loss functionObtained by a method in whichRepresenting the true fault type, the loss function L representing the difference between the predicted value and the true value; the minimization process of the loss function is expressed by the following formula:
，
wherein,parameters representing the optimal multiple classifier, +.>Representing a process of solving parameters minimizing a loss function, such as gradient descent, newton's method;
(4.5) performing classifier tests on the operation data of the electric pump.
The improved technical effects are as follows: and analyzing and modeling the operation data of the electric pump by using a data mining and machine learning method, extracting the normal working characteristics and the fault characteristics of the electric pump, and establishing a fault diagnosis model of the electric pump. The processes and formulas for data preprocessing, feature extraction, feature selection, classifier training, and classifier testing are described in detail herein to enable identification and prediction of multiple fault types for an electric pump. The fault monitoring device has the beneficial effects that the precision and the reliability of fault monitoring of the electric pump are improved.
Drawings
FIG. 1 is a schematic diagram of a fault monitoring system for a mining flameproof submersible sand-removing electric pump.
Fig. 2 is a schematic diagram of a fault monitoring method of a mining explosion-proof submersible sand discharge electric pump.
Fig. 3 is a schematic flow chart of conversion and transmission of data receiver number signals in the mining flameproof submersible sand discharge electric pump fault monitoring system.
Detailed Description
The following detailed description of the invention, in conjunction with the examples, is intended to be merely exemplary and explanatory and should not be construed as limiting the scope of the invention in any way, in order to provide a better understanding of the invention as claimed.
As shown in fig. 1, a mining flameproof submersible sand discharge electric pump fault monitoring system includes:
the sensor is arranged on a key component of the electric pump and is used for acquiring the operation parameters and state information of the electric pump in real time, including temperature, current, vibration and water level information;
the wireless communication module is connected with the sensor and used for sending the acquired data to a data receiver of the wellhead through a wireless signal;
the data receiver is arranged at the wellhead and is used for receiving data from the wireless communication module;
The data processing unit is connected with the data receiver and used for analyzing and modeling the received data, extracting normal working characteristics and fault characteristics of the electric pump and establishing a fault diagnosis model of the electric pump;
the display unit is connected with the data processing unit and used for providing the running state and fault alarm of the electric pump, and fault reasons and maintenance suggestions for operators and maintenance personnel;
and the data storage unit is connected with the data processing unit and is used for storing the operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction.
As a preferable mode of the above embodiment, the sensors are a temperature sensor, a current sensor, a vibration sensor, and a water level sensor, and are configured to collect temperature, current, vibration, water level parameters, and status information of the electric pump in real time;
the temperature sensor is used for measuring the temperature of a motor, an impeller, a bearing, a pump body and other parts of the electric pump, the temperature is one of the operation parameters of the electric pump, and an output signal of the temperature sensor is a voltage signal and is in direct proportion to the temperature, and the temperature sensor can be expressed by the following formula:
wherein->The output voltage of the temperature sensor is T, the temperature of the electric pump is T, and k is the sensitivity coefficient of the temperature sensor;
The current sensor is used for measuring the current of the motor of the electric pump, the current is one of the operation parameters of the electric pump, the output signal of the current sensor is a voltage signal, and the output signal is proportional to the current and can be expressed by the following formula:
wherein->For the output voltage of the current sensor, +.>The motor current of the electric pump is represented by k, and the sensitivity coefficient of the current sensor is represented by k;
the output signal of the vibration sensor is a voltage signal, and is related to the amplitude and frequency of vibration, and the output signal can be expressed by the following formula:
wherein->For the output voltage of the vibration sensor, A is the amplitude of the bearing vibration of the electric pump, +.>Is the frequency of the bearing vibration of the electric pump, t is time, < ->And->Is the sensitivity coefficient of the vibration sensor;
the water level sensor is used for measuring the water level of the pump body of the electric pump, the water level is one of the state information of the electric pump, the output signal of the water level sensor is a resistance signal, and the output signal is in direct proportion to the water level and can be expressed by the following formula:
，
wherein,is the output resistance of the water level sensor, +.>Is the pump body water level of the electric pump->Is a sensitivity coefficient of the water level sensor.
Temperature sensor:
the output voltage VT of the temperature sensor is proportional to the temperature T, expressed by the following formula:
VT=kT
Where k is the sensitivity coefficient of the temperature sensor, and the unit is V/. Degree.C.and represents the amplitude of change of the output voltage when the temperature changes by 1 ℃.
This formula exploits the seebeck effect of a thermocouple, i.e. when two different metals are connected together to form a circuit, if the temperatures of the two junctions differ, an electromotive force is generated in the circuit, thus forming an electric current. This electromotive force is related to the materials of the two metals and the temperatures of the two junctions, expressed by the following formula:
，
wherein E is electromotive force, E 1 And E is 2 Is the thermoelectric potential of two metals, k 1 And k 2 Seebeck coefficient, T, for two metals 1 And T 2 T is the temperature of two joints 0 For reference temperature, 0 ℃ is typically taken.
One end of one metal is connected to a voltmeter, and the other end is connected to a constant temperature source, and the temperature is kept at T 0 Then the voltage displayed on the voltmeter is the output voltage VT of the temperature sensor, one end of the other metal is connected to the measured object, and the other end is connected to the voltmeter, then the temperature of the measured object is T 1 And T is 2 Namely T 0 Substituting the formula to obtain:
，
let k=k 1 ，T=T 1 -T 0 The formula of the temperature sensor is obtained:
VT=kT，
specific numerical value deduction process of the temperature sensor:
When the sensitivity coefficient k of the temperature sensor is set to be 0.01V/DEG C, and the temperature of the measured object is 50 ℃, the output voltage is as follows:
VT=kT=0.01×50=0.5V，
when the temperature of the measured object is 100 ℃, the output voltage is as follows:
VT=kT=0.01×100=1V。
a current sensor:
the output voltage VI of the current sensor is proportional to the current I, expressed by the following formula:
，
where k is the sensitivity coefficient of the current sensor, and the unit is V/a, and represents the amplitude of the output voltage change when the current changes by 1A.
This voltage is related to current, magnetic field strength and material properties, expressed by the following formula:
，
wherein VH is Hall voltage, B is magnetic field intensity,i is current, l is thickness of material, R H Is the hall coefficient, which is an inherent property of the material.
And connecting two ends of the conductive material to a voltmeter, wherein the voltage displayed on the voltmeter is the output voltage VI of the current sensor, and if the magnetic field strength B and the thickness l of the material are kept unchanged, the Hall voltage is only in direct proportion to the current I and is substituted into the formula to obtain the magnetic field strength B:
，
let k= BlR _h, the formula of the current sensor is obtained:
，
setting the sensitivity coefficient k of the current sensor to be 0.1V/a, then when the current is 10A, the output voltage is:
，
when the current is 20A, the output voltage is:
。
A water level sensor:
the output resistance Rh of the water level sensor is proportional to the water level h and is expressed by the following formula:
，
where k is the sensitivity coefficient of the water level sensor, and the unit is Ω/m, and represents the amplitude of the output resistance change when the water level changes by 1 m.
I.e. when the water level changes, the position of the float changes as well, thereby changing the contact point of the float with the resistor and thus the resistance value of the resistor. This resistance value is related to the position of the float and the length of the resistor, expressed by the following formula:
，
where R is the resistance value, ρ is the resistivity, L is the length of the resistor, and A is the cross-sectional area of the resistor.
If one end of the resistor is grounded and the other end is connected to the voltmeter, the resistance displayed on the voltmeter is the output resistance Rh of the water level sensor, if the resistivity rho and the cross-sectional area A are kept unchanged, the resistance value is only proportional to the length L of the resistor, and the length L of the resistor is proportional to the position of the floater, namely the water level h, and the position is substituted into the formula to obtain the water level sensor:
，
let k=frac { ρ } { a }, the formula of the water level sensor is obtained:
，
the specific numerical value deducing process of the water level sensor comprises the following steps:
setting the sensitivity coefficient k of the water level sensor to be 100 Ω/m, then when the water level is 1m, the output resistance is:
，
When the water level is 2m, the output resistance is:
。
vibration sensor:
first, from the relationship between the output voltage of the vibration sensor and the amplitude and frequency of the vibration, the following is obtained:
，
wherein,for the output voltage of the vibration sensor, A is the amplitude of the bearing vibration of the electric pump, +.>Is the frequency of the bearing vibration of the electric pump, t is time, < ->And->Is the sensitivity coefficient of the vibration sensor;
then, the above formula is simplified into the following form according to the identity of the trigonometric function:
，
wherein,for the total sensitivity coefficient of the vibration sensor, < +.>A phase difference for the vibration sensor;
next, the above formula is transformed into the following form according to the principle of fourier transform:
，
wherein,for a frequency domain representation of the output voltage of the vibration sensor, < >>For angular frequency +.>As a dirac function, +.>Is an imaginary unit;
finally, according to the method of spectrum analysis, the amplitude and frequency of the bearing vibration of the electric pump and the phase difference of the vibration sensor are extracted from the above formula, and are used for judging the running stability and the fault type of the electric pump, as follows:
，
，
。
the process and results for demonstrating the measurement of bearing vibrations and fault diagnosis of an electric pump are presented below, in combination with specific values:
The amplitude of vibration of the bearing of the electric pump is set to be 0.1mm, the frequency is set to be 50Hz, and the sensitivity coefficient of the vibration sensor is set to be k 1 =0.01V/mm，k 2 =0.02V/mm, the output voltage of the vibration sensor is calculated according to the above formula as:
，
，
the above formula is then transformed into the following form according to the principle of fourier transform:
，
then, according to the method of spectrum analysis, the amplitude and frequency of the bearing vibration of the electric pump and the phase difference of the vibration sensor are extracted from the above formula, as follows:
mm，HZ，rad。
finally, according to the amplitude and frequency of bearing vibration of the electric pump and the phase difference of the vibration sensor, the running stability and the fault type of the electric pump are judged, as follows:
if the amplitude of the vibration of the bearing of the electric pump is smaller than 0.2mm, the frequency is equal to the rotating speed of the electric pump, and the phase difference is 0, the electric pump is stable to operate and has no faults;
if the amplitude of the vibration of the bearing of the electric pump is larger than 0.2mm, the frequency is equal to the rotating speed of the electric pump, and the phase difference is not 0, the electric pump has the fault of impeller unbalance, and the balance degree of the impeller needs to be adjusted;
if the amplitude of the vibration of the bearing of the electric pump is larger than 0.2mm, the frequency is larger than the rotating speed of the electric pump, and the phase difference is not 0, the bearing abrasion fault of the electric pump is indicated, and the bearing needs to be replaced;
If the amplitude of the vibration of the bearing of the electric pump is larger than 0.2mm, the frequency is smaller than the rotating speed of the electric pump, and the phase difference is not 0, the electric pump has the fault of water seal damage, and the water seal needs to be replaced.
As a preferred mode of the foregoing embodiment, the wireless communication module is a node in the wireless sensor network, and uses a ZigBee protocol or other low-power consumption, high-reliability, long-distance wireless communication protocols to send the data collected by the sensor to the data receiver of the wellhead through a wireless signal, where the wireless signal is wireless transmission of operation data of the electric pump, and is used to realize remote monitoring and control of the electric pump.
As shown in fig. 3, as a preferred mode of the above embodiment, the data receiver is a coordinator in the wireless sensor network, and is configured to receive a wireless signal from the wireless communication module, that is, wireless transmission of operation data of the electric pump, and convert the wireless signal into a wired signal, that is, wired transmission of operation data of the electric pump, and connect the wired communication interface with the data processing unit, and transmit the operation data of the electric pump to the data processing unit, so as to implement data acquisition and processing of the electric pump;
the data receiver is a device with wireless receiving and wired transmitting capabilities, can realize conversion and transmission of wireless signals and wired signals, and has the following working principle:
(1) The data receiver receives a wireless signal from the wireless communication module, i.e. wireless transmission of operation data of the electric pump, through the wireless antenna, the wireless signal being an electromagnetic wave signal, which can be expressed by the following formula:
，
wherein,for wireless signals, A is the amplitude of the wireless signal, < >>Is the frequency of the radio signal, t is time, < >>Is the phase of the wireless signal;
(2) The data receiver demodulates and decodes the wireless signal through the wireless receiving module, and restores the wireless signal into a digital signal, namely the operation data of the electric pump, wherein the digital signal is a binary signal and can be expressed by the following formula:
，
wherein,digital signal>Is the nth bit of the digital signal, +.>The pulse shape of the digital signal is that T is the bit interval of the digital signal and N is the bit number of the digital signal;
(3) The data receiver encodes and modulates the digital signal through the wired transmitting module, and converts the digital signal into a wired signal, namely, the wired transmission of the operation data of the electric pump, wherein the wired signal is a voltage signal and can be expressed by the following formula:
，
wherein,is a wired signal>Is the nth symbol of the cable signal, < >>The symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal;
(4) The data receiver sends the wired signal to the data processing unit through the wired communication interface, and the conversion and transmission from the wireless transmission of the operation data of the electric pump to the wired transmission are completed.
Technical purpose for implementing the conversion and transmission of the wireless transmission of the operating data of an electric pump to a wired transmission:
first, according to the relationship between the wireless signal and the electromagnetic wave signal, the following formula is obtained:
，
wherein,for wireless signals, A is the amplitude of the wireless signal, < >>Is the frequency of the radio signal, t is time, < >>Is the phase of the wireless signal;
then, according to the relation between the digital signal and the binary signal, the following formula is obtained:
，
wherein,digital signal>Is the nth bit of the digital signal, +.>The pulse shape of the digital signal is that T is the bit interval of the digital signal and N is the bit number of the digital signal;
then, according to the conversion of the wireless signal and the digital signal, the following formula is obtained:
，
wherein,for wireless signals, A is the amplitude of the wireless signal, < >>Is the frequency of the radio signal, t is time, < >>For the phase of the radio signal>Digital signal>Is the nth bit of the digital signal, +.>The pulse shape of the digital signal is that T is the bit interval of the digital signal and N is the bit number of the digital signal;
Then, according to the relation between the wired signal and the voltage signal, the following formula is obtained:
，/>
wherein,is a wired signal>Is the nth symbol of the cable signal, < >>The symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal;
then, according to the conversion of the digital signal and the wired signal, the following formula is obtained:
，
wherein,digital signal>Is the nth bit of the digital signal, +.>For the pulse shape of the digital signal, T is the bit interval of the digital signal, N is the bit number of the digital signal, < >>Is a wired signal>Is the nth symbol of the cable signal, < >>The symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal;
finally, according to the conversion of the wireless signal and the wired signal, the following formula is obtained:
，
wherein,for wireless signals, A is the amplitude of the wireless signal, < >>Is the frequency of the radio signal, t is time, < >>For the phase of the radio signal>Is a wired signal>Is the nth symbol of the cable signal, < >>The symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal.
The process and results for the conversion and transmission of wireless to wired transmission of operational data for an electric pump are presented and are combined with specific values as follows:
Setting the operation data of the electric pump as parameters of temperature, current, vibration, water level and the like, wherein each parameter is an 8-bit binary number, for example, the temperature is 01011011, the current is 10010110, the vibration is 00101101, the water level is 11100010, and the digital signal is calculated according to the mathematical formula:
，
wherein,an nth bit of the digital signal, in turn 01011011100101100010110111100010,setting a square wave for the pulse shape of the digital signal, setting 0.1ms for the bit interval of the digital signal as T, and setting 32 for the bit number of the digital signal as N;
then, the amplitude of the radio signal was set to 1V, the frequency was set to 100kHz, and the phase was set to 0, and the radio signal was calculated as:
，
wherein, A=1V,，；
then, assuming that the symbol shape of the wired signal is a sine wave, the symbol interval is 0.2ms, and the number of symbols is 16, the wired signal is calculated according to the above formula as:
，
wherein,the nth symbol of the wired signal is 0101101110010110, < >>Is the symbol shape of the wired signal, set to +.>T is the symbol interval of the wired signal, which is 0.2ms, N is the number of symbols of the wired signal, which is 16;
finally, according to the conversion of the wireless signal and the wired signal, the following formula is obtained:
，
Wherein,for wireless signals, a=1v,，，in the form of a wired signal, the signal is,the nth symbol of the wired signal, in turn 0101101110010110,the symbol shape of the wired signal is T, the symbol interval of the wired signal is T, and N is the number of symbols of the wired signal.
The data receiver sends the wired signal to the data processing unit through the wired communication interface, and the conversion and transmission from the wireless transmission of the operation data of the electric pump to the wired transmission are completed. The data processing unit restores various parameters of the operation data of the electric pump, such as temperature, current, vibration, water level and the like, according to the symbols of the wired signals, analyzes and processes the parameters, and realizes the functions of data acquisition and processing of the electric pump.
As the preferred mode of the embodiment, the data processing unit is an industrial personal computer or other equipment with data processing capability, is internally provided with data processing software and is used for analyzing and modeling the received operation data of the electric pump, extracting normal operation characteristics and failure characteristics of the electric pump, establishing a failure diagnosis model of the electric pump, the data processing software adopts the data mining and machine learning methods to preprocess the operation data of the electric pump, extract the characteristics, select the characteristics, train the classifier, test the classifier and the like, so as to obtain a multi-classifier capable of identifying multiple failure types of the electric pump, the multi-classifier is an artificial neural network or other models with classification capability, can output the failure types of the electric pump according to the input of the operation data of the electric pump, including motor overheat, impeller blockage, bearing abrasion, water seal damage, cable breakage and the like, and gives corresponding maintenance suggestions according to the failure types, and the maintenance scheme and maintenance steps are output according to the failure types and the failure positions, so as to guide operators and maintenance staff to conduct the failure removal and maintenance.
As a preferred mode of the above embodiment, the display unit is a display or other device with display capability, and a speaker or other device with voice broadcasting capability, and is used for providing the operating status and fault alarm, and fault cause and maintenance advice of the electric pump for operators and maintenance personnel, and the display unit is connected with the data processing unit, and receives the operating data, fault type, fault location and maintenance advice information of the electric pump from the data processing unit.
As a preferred mode of the above embodiment, the data storage unit is a database server or other devices with data storage capability, and is used for storing the operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction, where the database is a relational database, and includes a basic information table, an operation parameter table, a state information table, a fault information table, a maintenance record table and the like of the electric pump, so as to implement query, statistics, analysis and prediction of historical data of the electric pump, and the data analysis and fault prediction is a method based on data mining and machine learning, and can discover the operation rule and fault mode of the electric pump according to the historical data and current data of the electric pump, predict the future state and fault risk of the electric pump, and provide support for operation optimization and maintenance decision of the electric pump.
As shown in fig. 2, the method for monitoring faults of the mining flameproof submersible sand discharge electric pump comprises the following steps:
(1) The sensor is arranged on a key component of the electric pump and used for acquiring the operation parameters and state information of the electric pump in real time, including temperature, current, vibration, water level and the like;
(2) Connecting the sensor with a wireless communication module to form a wireless sensor network, and transmitting acquired data to a data receiver of a wellhead through wireless signals;
(3) A data processing unit and a display unit are arranged on a data receiver of the wellhead and are used for receiving and processing data from the wireless sensor network and displaying and storing the running condition and fault information of the electric pump;
(4) In the data processing unit, the operation data of the electric pump is analyzed and modeled by adopting a data mining and machine learning method, the normal working characteristics and the fault characteristics of the electric pump are extracted, a fault diagnosis model of the electric pump is established, the model is a multi-classifier, the fault types of the electric pump can be output according to the input of the operation data of the electric pump, including motor overheating, impeller blockage, bearing abrasion, water seal damage, cable breakage and the like, corresponding maintenance suggestions are given according to the fault types, and the maintenance schemes and the maintenance steps can be output according to the fault types and the fault positions;
(5) In the display unit, a graphical and phonetic mode is adopted to provide the operation state and fault alarm of the electric pump, as well as the fault reason and maintenance advice for operators and maintenance personnel, the display unit comprises a display screen and a loudspeaker, the display screen is used for displaying the information of the operation parameters, state curves, fault types, fault positions, maintenance advice and the like of the electric pump, and the loudspeaker is used for voice broadcasting the fault alarm and the maintenance advice of the electric pump;
(6) In the data storage unit, the operation data and fault information of the electric pump are stored in a database for subsequent data analysis and fault prediction, the database is a relational database comprising a basic information table, an operation parameter table, a state information table, a fault information table, a maintenance record table and the like of the electric pump, the historical data of the electric pump can be inquired, counted, analyzed and predicted, the data analysis and fault prediction is a method based on data mining and machine learning, the operation rule and fault mode of the electric pump can be found according to the historical data and current data of the electric pump, the future state and fault risk of the electric pump are predicted, and support is provided for operation optimization and maintenance decision of the electric pump.
As a preferred mode of the above embodiment, in the step (4), the data mining and machine learning method is adopted to analyze and model the operation data of the electric pump, extract the normal working characteristics and the fault characteristics of the electric pump, and build the fault diagnosis model of the electric pump, which specifically includes the following sub-steps:
(4.1) preprocessing operation data of the electric pump, including data cleaning, data standardization, data dimension reduction and the like, removing noise, abnormal values, redundant features and the like in the data, and improving the quality and usability of the data;
(4.2) extracting the characteristics of the operation data of the electric pump, including time domain characteristics, frequency domain characteristics, time-frequency domain characteristics and the like, and extracting characteristic values reflecting the operation condition and fault condition of the electric pump, such as mean value, variance, peak value, energy, frequency spectrum and the like, from the data by utilizing methods of mathematical transformation, statistical analysis and the like;
(4.3) selecting the characteristics of the operation data of the electric pump, including a filtering method, a packing method, an embedding method and the like, and selecting the most distinguishable characteristic subset with the representativeness from the extracted characteristics by utilizing methods such as correlation analysis, information gain, a support vector machine and the like as the operation characteristics of the electric pump, so that the dimension and the complexity of the characteristics are reduced;
(4.4) performing classifier training on the operation data of the electric pump, wherein the classifier training comprises an artificial neural network, a support vector machine, a decision tree, a random forest and the like, and training a multi-classifier capable of identifying multiple fault types of the electric pump by using the operation characteristics and the fault types of the electric pump with labels and adopting a supervised learning method as a fault diagnosis model of the electric pump; the classifier training process is represented by the following formula:
，
wherein,indicating the type of fault that is predicted,representing a function of a multi-classifier, such as an artificial neural network, a support vector machine, a decision tree, a random forest, etc.,parameters representing multiple classifiers, such as weights, biases, kernel functions, etc. Parameters of multiple classifiersBy minimizing a loss functionObtained by a method in whichRepresenting the true fault type, the loss function L represents the difference between the predicted value and the true value, such as cross entropy, mean square error, etc. The minimization process of the loss function is expressed by the following formula:
，
wherein,parameters representing the optimal multiple classifier, +.>Representing a process of solving parameters minimizing a loss function, such as gradient descent, newton's method;
technical purpose for implementing classifier training of operational data of an electric pump:
Firstly, according to the relation between the characteristics of the operation data of the electric pump and the fault type, the following formula is obtained:
wherein,a feature vector representing operational data of the electric pump,an original vector representing operational data of the electric pump,representing a feature extraction function, such as principal component analysis, linear discriminant analysis, etc.,parameters representing feature extraction, such as the number of principal components, the number of discrimination directions, etc.;
then, according to the relation between the characteristic vector of the operation data of the electric pump and the fault type, the following formula is obtained:
，
wherein,indicating the type of fault that is predicted,representing a function of a multi-classifier, such as an artificial neural network, a support vector machine, a decision tree, a random forest, etc.,parameters representing multiple classifiers, such as weights, biases, kernel functions, etc.;
then, according to the relation between the actual fault type and the predicted fault type of the operation data of the electric pump, the following formula is obtained:，
wherein the method comprises the steps ofRepresenting the true fault type, L representing a loss function, such as cross entropy, mean square error, etc., for measuring the difference between the predicted value and the true value;
finally, according to the target of classifier training of the operation data of the electric pump, the following formula is obtained:
，
Wherein,parameters representing the optimal multiple classifier, +.>Representing the process of solving the parameters that minimize the loss function, such as gradient descent, newton's method, etc.
The process and results of classifier training for presenting operational data of an electric pump are described in connection with the following embodiments:
setting operation data of the electric pump as parameters of temperature, current, vibration, water level and the like, wherein each parameter is a floating point number, such as the temperature is 36.5 ℃, the current is 12.3A, the vibration is 0.8g, the water level is 1.2m, the fault types of the electric pump are normal, overheat, overload, locked rotor, water leakage and the like, each fault type is represented by an integer, such as normal 0, overheat is 1, overload is 2, locked rotor is 3, and water leakage is 4, and calculating the following results according to the mathematical formula:
first, an original vector of operation data of an electric pump is set using Principal Component Analysis (PCA) as a function of feature extractionConversion to feature vectorsWherein, the method comprises the steps of, wherein,setting the number of the main components to be 2, namely reserving the variance of 95% of the original data;
then, setting a feature vector of the operation data of the electric pump by using a Support Vector Machine (SVM) as a function of the multi-classifierInput into SVM to obtain predicted fault type I.e., normal, wherein,parameters of the SVM, such as a kernel function is a Gaussian kernel, a penalty coefficient is 1, and the like;
then, setting the true fault type of the operation data of the electric pump by using the cross entropy as a loss functionAnd predicted fault typeInput into cross entropy to obtain loss valueThe predicted value is completely consistent with the true value, and no error exists;
finally, setting a gradient descent method as a process for solving the optimal parameters, and setting a loss value of the operation data of the electric pumpInputting into gradient descent method to obtain optimal SVM parametersWherein, the method comprises the steps of, wherein,is a weight vector for the SVM,is a bias term for the SVM,is a parameter of the gaussian kernel and is,is a penalty coefficient.
And (4.5) performing classifier tests on the operation data of the electric pump, wherein the classifier tests comprise cross verification, confusion matrix, accuracy, recall rate and the like, and using the operation characteristics of the electric pump without labels, adopting an unsupervised learning method to test the performance and effect of the fault diagnosis model of the electric pump and evaluate the accuracy and reliability of the fault diagnosis model of the electric pump.
As a preferred mode of the above embodiment, in the step (6), the operation data and the fault information of the electric pump are stored in a database for subsequent data analysis and fault prediction, specifically including the following sub-steps:
(6.1) in the data storage unit, the operation data and fault information of the electric pump are stored in a database, wherein the database is a relational database, and comprises a basic information table, an operation parameter table, a state information table, a fault information table, a maintenance record table and the like of the electric pump, so that the inquiry, statistics, analysis and prediction of the historical data of the electric pump can be realized;
in the data analysis unit, a data mining and machine learning method is adopted to analyze and predict historical data and current data of the electric pump, an operation rule and a fault mode of the electric pump are found, future states and fault risks of the electric pump are predicted, support is provided for operation optimization and maintenance decision of the electric pump, the data analysis unit is an industrial personal computer or other equipment with data analysis capability, data analysis software is built in the data analysis unit and is used for processing and mining data in a database, the data analysis software adopts the data mining and machine learning method to perform clustering, association, classification, regression, prediction and other steps on the data of the electric pump, and a data analysis model capable of reflecting the operation trend and the fault trend of the electric pump is obtained, and is a manual neural network or other models with analysis capability and capable of outputting the operation state and the fault risk of the electric pump and operation optimization and maintenance advice according to the historical data and the current data of the electric pump;
(6.3) providing the operating state and the fault risk of the electric pump and the operation optimization and maintenance advice to operators and maintenance personnel in a display unit in a graphical and phonetic mode, wherein the display unit comprises a display screen and a loudspeaker, the display screen is used for displaying information such as the operating parameters, the state curve, the fault trend, the operation optimization and the maintenance advice of the electric pump, and the loudspeaker is used for phonetic broadcasting the operating state and the fault risk of the electric pump and the operation optimization and the maintenance advice.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to facilitate an understanding of the method of the present invention and its core ideas. The foregoing is merely a preferred embodiment of the invention, and it should be noted that, due to the limited text expressions, there is objectively no limit to the specific structure, and that, for a person skilled in the art, modifications, adaptations or variations may be made without departing from the principles of the present invention, and the above technical features may be combined in any suitable manner; such modifications, variations and combinations, or the direct application of the inventive concepts and aspects to other applications without modification, are contemplated as falling within the scope of the present invention.
Claims (8)
1. The utility model provides a mining flame proof formula dive sediment outflow charge pump fault monitoring system which characterized in that includes:
the sensor is arranged on a component of the electric pump and is used for acquiring the operation parameters and state information of the electric pump in real time, including temperature, current, vibration and water level information;
the wireless communication module is connected with the sensor and used for sending the acquired data to a data receiver of the wellhead through a wireless signal;
the data receiver is arranged at the wellhead and is used for receiving data from the wireless communication module;
the data processing unit is connected with the data receiver and used for analyzing and modeling the received data, extracting normal working characteristics and fault characteristics of the electric pump and establishing a fault diagnosis model of the electric pump;
in the data processing unit, the operation data of the electric pump is analyzed and modeled by adopting a data mining and machine learning method, the normal working characteristics and the fault characteristics of the electric pump are extracted, and a fault diagnosis model of the electric pump is established; the method specifically comprises the following substeps:
(1) Preprocessing operation data of the electric pump, including data cleaning, data normalization and data dimension reduction, removing noise, abnormal values and redundant characteristics in the data, and improving the quality and usability of the data;
(2) Extracting the characteristics of the operation data of the electric pump, including time domain characteristics, frequency domain characteristics and time-frequency domain characteristics, and extracting characteristic values reflecting the operation condition and fault condition of the electric pump from the data by utilizing a mathematical transformation and statistical analysis method;
(3) The method comprises the steps of selecting the operation data of the electric pump, including a filtering method, a packing method and an embedding method, and selecting a feature subset with the most distinguishing capability and the representativeness from the extracted features by utilizing a correlation analysis, an information gain and a support vector machine method as the operation features of the electric pump, so that the dimension and the complexity of the features are reduced;
(4) The method comprises the steps of training operation data of an electric pump by using a classifier, wherein the classifier comprises an artificial neural network, a support vector machine, a decision tree and a random forest, and training a multi-classifier capable of identifying multiple fault types of the electric pump by using the operation characteristics and the fault types of the electric pump with labels by using a supervised learning method to serve as a fault diagnosis model of the electric pump; the classifier training process is represented by the following formula:
，
where y represents the predicted type of fault,representing a function of a multi-classifier,parameters representing multiple classifiers; parameters of multiple classifiers By minimizing a loss functionFor learning, whereinRepresenting the true fault type, loss functionRepresenting the difference between the predicted value and the actual value; the minimization process of the loss function is expressed by the following formula:
，
wherein,the parameters representing the optimal multi-classifier are presented,representing a process of solving a parameter that minimizes a loss function;
(5) The method comprises the steps of performing classifier tests on operation data of an electric pump, including cross verification, confusion matrix, accuracy, recall rate and the like, using operation characteristics of the electric pump without labels, testing performance and effect of a fault diagnosis model of the electric pump by an unsupervised learning method, and evaluating accuracy and reliability of the fault diagnosis model of the electric pump;
the display unit is connected with the data processing unit and used for providing the running state and fault alarm of the electric pump, and fault reasons and maintenance suggestions for operators and maintenance personnel;
and the data storage unit is connected with the data processing unit and is used for storing the operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction.
2. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 1, wherein the sensors are a temperature sensor, a current sensor, a vibration sensor and a water level sensor, and are used for collecting temperature, current, vibration, water level parameters and state information of the electric pump in real time.
3. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 2, wherein the temperature sensor is used for measuring the temperature of a motor, an impeller, a bearing and a pump body of the electric pump, the temperature is one of operation parameters of the electric pump, and the output signal of the temperature sensor is a voltage signal, is proportional to the temperature and can be expressed by the following formula:
wherein, the method comprises the steps of, wherein,for the output voltage of the temperature sensor,is the temperature of the electric pump,is the sensitivity coefficient of the temperature sensor;
the current sensor is used for measuring the current of the motor of the electric pump, the current is one of the operation parameters of the electric pump, the output signal of the current sensor is a voltage signal, and the output signal is proportional to the current, and can be expressed by the following formula:
wherein, the method comprises the steps of, wherein,for the output voltage of the current sensor,is the motor current of the electric pump,is the sensitivity coefficient of the current sensor;
the output signal of the vibration sensor is a voltage signal, and is related to the amplitude and frequency of vibration, and can be expressed by the following formula:
wherein, the method comprises the steps of, wherein,for outputting electricity from vibration sensorThe pressure is applied to the pressure-sensitive adhesive,is the amplitude of the bearing vibration of the electric pump,is the frequency of the bearing vibration of the electric pump,in order to be able to take time,andis the sensitivity coefficient of the vibration sensor;
The water level sensor is used for measuring the water level of the pump body of the electric pump, the water level is one of the state information of the electric pump, the output signal of the water level sensor is a resistance signal, and the output signal is in direct proportion to the water level and can be expressed by the following formula:
，
wherein,is an output resistor of the water level sensor,is the water level of the pump body of the electric pump,is a sensitivity coefficient of the water level sensor.
4. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 1, wherein the data receiver is a coordinator in a wireless sensor network and is used for receiving wireless signals from a wireless communication module, namely wireless transmission of operation data of the electric pump, converting the wireless signals into wired signals, namely wired transmission of the operation data of the electric pump, connecting the wired signals with a data processing unit through a wired communication interface, and transmitting the operation data of the electric pump to the data processing unit for realizing data acquisition and processing of the electric pump;
the data receiver is a device with wireless receiving and wired transmitting capabilities, can realize conversion and transmission of wireless signals and wired signals, and comprises the following steps:
(1) The data receiver receives a wireless signal from the wireless communication module through the wireless antenna, that is, wireless transmission of operation data of the electric pump, which can be represented by the following formula:
，
Wherein,in the form of a wireless signal,for the amplitude of the wireless signal,for the frequency of the wireless signal,in order to be able to take time,is the phase of the wireless signal;
(2) The data receiver demodulates and decodes the wireless signal through the wireless receiving module, and restores the wireless signal into a digital signal, namely the operation data of the electric pump, wherein the digital signal is a binary signal and can be expressed by the following formula:
，
wherein,as a result of the digital signal,is the first of the digital signalsA number of bits of a bit,is the pulse shape of the digital signal,for the bit interval of the digital signal,the number of bits of the digital signal;
(3) The data receiver encodes and modulates the digital signal through the wired transmitting module, and converts the digital signal into a wired signal, namely, the wired transmission of the operation data of the electric pump, wherein the wired signal is a voltage signal and can be expressed by the following formula:
，
wherein,in the form of a wired signal, the signal is,is the first of the wired signalsThe number of symbols to be used in a symbol,symbol shape of wired signalIn the shape of a sheet, the shape of the sheet,for the symbol interval of the cable signal,the number of symbols being a wired signal;
(4) The data receiver sends the wired signal to the data processing unit through the wired communication interface, and the conversion and transmission from the wireless transmission of the operation data of the electric pump to the wired transmission are completed.
5. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 1, wherein the data processing unit is an industrial personal computer or equipment with data processing capability, and is internally provided with data processing software for analyzing and modeling the received operation data of the electric pump, extracting normal working characteristics and fault characteristics of the electric pump, and establishing a fault diagnosis model of the electric pump.
6. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 1, wherein the display unit is a display or equipment with display capability and a loudspeaker or equipment with voice broadcasting capability, and is used for providing the operating state and fault alarm, fault reason and maintenance advice of the electric pump for operators and maintenance staff, and the display unit is connected with the data processing unit and receives the operating data, fault type, fault position and maintenance advice information of the electric pump from the data processing unit.
7. The mining explosion-proof submersible sand-removing electric pump fault monitoring system according to claim 1, wherein the data storage unit is a database server or equipment with data storage capacity and is used for storing operation data and fault information of the electric pump in a database for subsequent data analysis and fault prediction.
8. The fault monitoring method for the mining explosion-proof submersible sand discharge electric pump is characterized by comprising the following steps of:
(1) The sensor is arranged on a key component of the electric pump and used for collecting the operation parameters and state information of the electric pump in real time, including temperature, current, vibration and water level;
(2) Connecting the sensor with a wireless communication module to form a wireless sensor network, and transmitting acquired data to a data receiver of a wellhead through wireless signals;
(3) A data processing unit and a display unit are arranged on a data receiver of the wellhead and are used for receiving and processing data from the wireless sensor network and displaying and storing the running condition and fault information of the electric pump;
(4) In the data processing unit, the operation data of the electric pump is analyzed and modeled by adopting a data mining and machine learning method, the normal working characteristics and the fault characteristics of the electric pump are extracted, and a fault diagnosis model of the electric pump is established; the method specifically comprises the following substeps:
(4.1) preprocessing operation data of the electric pump, including data cleaning, data standardization and data dimension reduction, removing noise, abnormal values and redundant characteristics in the data, and improving the quality and usability of the data;
(4.2) extracting the characteristics of the operation data of the electric pump, including time domain characteristics, frequency domain characteristics and time-frequency domain characteristics, and extracting characteristic values reflecting the operation condition and fault condition of the electric pump from the data by utilizing a mathematical transformation and statistical analysis method;
(4.3) selecting the characteristics of the operation data of the electric pump, including a filtering method, a packing method and an embedding method, and selecting the most distinguishable characteristic subset with the representative characteristic from the extracted characteristics by utilizing a correlation analysis, an information gain and a support vector machine method as the operation characteristics of the electric pump, thereby reducing the dimension and the complexity of the characteristics;
(4.4) performing classifier training on the operation data of the electric pump, wherein the classifier training comprises an artificial neural network, a support vector machine, a decision tree and a random forest, and a multi-classifier capable of identifying multiple fault types of the electric pump is trained by using the operation characteristics and the fault types of the electric pump with labels and a supervised learning method to serve as a fault diagnosis model of the electric pump; the classifier training process is represented by the following formula:
，
where y represents the predicted type of fault,representing a function of a multi-classifier,parameters representing multiple classifiers; parameters of multiple classifiers By minimizing a loss functionFor learning, whereinRepresenting the true fault type, loss functionRepresenting the difference between the predicted value and the actual value; the minimization process of the loss function is expressed by the following formula:
，
wherein,the parameters representing the optimal multi-classifier are presented,representing solution such that lossA process of minimizing a parameter of the function;
(4.5) performing classifier tests on the operation data of the electric pump, including cross-validation, confusion matrix, accuracy, recall rate and the like, using the operation characteristics of the electric pump without labels, adopting an unsupervised learning method to test the performance and effect of the fault diagnosis model of the electric pump, and evaluating the accuracy and reliability of the fault diagnosis model of the electric pump;
(5) In the display unit, a graphical and phonetic mode is adopted to provide the operation state and fault alarm of the electric pump, as well as fault reasons and maintenance suggestions for operators and maintenance personnel;
(6) In the data storage unit, the operating data and fault information of the electric pump are stored in a database for subsequent data analysis and fault prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410057857.0A CN117570039B (en) | 2024-01-16 | 2024-01-16 | Mining flameproof submersible sand discharge electric pump fault monitoring system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410057857.0A CN117570039B (en) | 2024-01-16 | 2024-01-16 | Mining flameproof submersible sand discharge electric pump fault monitoring system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117570039A CN117570039A (en) | 2024-02-20 |
| CN117570039B true CN117570039B (en) | 2024-04-12 |
Family
ID=89864753
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410057857.0A Active CN117570039B (en) | 2024-01-16 | 2024-01-16 | Mining flameproof submersible sand discharge electric pump fault monitoring system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117570039B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118088431B (en) * | 2024-04-28 | 2024-08-06 | 山东海纳智能装备科技股份有限公司 | Explosion-proof submersible sand-discharging electric pump fault monitoring system and method for mine |
| CN118277939B (en) * | 2024-05-31 | 2024-08-02 | 东营鑫奥船舶设备制造有限公司 | Ship cutting machine operation fault diagnosis system based on machine vision |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103341396A (en) * | 2013-07-28 | 2013-10-09 | 洛阳理工学院 | Dual temperature measurement and control protective device for main bearing of cement mill |
| CN104514731A (en) * | 2013-09-27 | 2015-04-15 | 天津市滨海鼎鑫机械制造有限公司 | Intelligent submerged pump controller |
| CN106089754A (en) * | 2016-07-04 | 2016-11-09 | 宁波巨神制泵实业有限公司 | Immersible pump real-time fault diagnosis based on " the Internet+" and method for maintaining |
| CN107843287A (en) * | 2017-10-26 | 2018-03-27 | 苏州数言信息技术有限公司 | Integrated sensor device and the environment event recognition methods based on it |
| CN109635483A (en) * | 2018-12-24 | 2019-04-16 | 上海商然数据服务有限公司 | A kind of motor and failure of pump hypothesis analysis system based on electromagnetic detection |
| CN111340238A (en) * | 2020-03-12 | 2020-06-26 | 中南大学 | Fault diagnosis method, device, equipment and storage medium of industrial system |
| CN111365251A (en) * | 2020-03-27 | 2020-07-03 | 北京天泽智云科技有限公司 | Intelligent diagnosis method for faults of centrifugal pump unit |
| CN111898634A (en) * | 2020-06-22 | 2020-11-06 | 西安交通大学 | Intelligent fault diagnosis method based on depth-to-reactance-domain self-adaption |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9366451B2 (en) * | 2010-12-24 | 2016-06-14 | Commonwealth Scientific And Industrial Research Organisation | System and method for the detection of faults in a multi-variable system utilizing both a model for normal operation and a model for faulty operation |
| CA2991725A1 (en) * | 2018-01-10 | 2019-07-10 | Psi Technologies Inc. | Method and apparatus to predict failure and control vibrations in a subsurface artificial lift system |
| US11401927B2 (en) * | 2020-05-28 | 2022-08-02 | American Jereh International Corporation | Status monitoring and failure diagnosis system for plunger pump |
-
2024
- 2024-01-16 CN CN202410057857.0A patent/CN117570039B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103341396A (en) * | 2013-07-28 | 2013-10-09 | 洛阳理工学院 | Dual temperature measurement and control protective device for main bearing of cement mill |
| CN104514731A (en) * | 2013-09-27 | 2015-04-15 | 天津市滨海鼎鑫机械制造有限公司 | Intelligent submerged pump controller |
| CN106089754A (en) * | 2016-07-04 | 2016-11-09 | 宁波巨神制泵实业有限公司 | Immersible pump real-time fault diagnosis based on " the Internet+" and method for maintaining |
| CN107843287A (en) * | 2017-10-26 | 2018-03-27 | 苏州数言信息技术有限公司 | Integrated sensor device and the environment event recognition methods based on it |
| CN109635483A (en) * | 2018-12-24 | 2019-04-16 | 上海商然数据服务有限公司 | A kind of motor and failure of pump hypothesis analysis system based on electromagnetic detection |
| CN111340238A (en) * | 2020-03-12 | 2020-06-26 | 中南大学 | Fault diagnosis method, device, equipment and storage medium of industrial system |
| CN111365251A (en) * | 2020-03-27 | 2020-07-03 | 北京天泽智云科技有限公司 | Intelligent diagnosis method for faults of centrifugal pump unit |
| CN111898634A (en) * | 2020-06-22 | 2020-11-06 | 西安交通大学 | Intelligent fault diagnosis method based on depth-to-reactance-domain self-adaption |
Non-Patent Citations (2)
| Title |
|---|
| 基于深度学习的故障诊断方法综述;文成林;吕菲亚;;电子与信息学报;20200115(第01期);第234-248页 * |
| 大型潜水电泵状态监测与诊断系统的研究;夏志昌;李春晋;;水利科技与经济;20090830(第08期);第728-730页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117570039A (en) | 2024-02-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117570039B (en) | Mining flameproof submersible sand discharge electric pump fault monitoring system and method | |
| US10520397B2 (en) | Methods and apparatuses for defect diagnosis in a mechanical system | |
| CN110647133A (en) | Rail transit equipment state detection maintenance method and system | |
| US20200284694A1 (en) | Machine Monitoring | |
| CN102944416A (en) | Multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades | |
| CN103487250B (en) | Based on the coal mine equipment anticipatory maintenance method of two-dimensional projections | |
| CN110398384A (en) | The machine health monitoring of rotating machinery | |
| CN118194150A (en) | Remote monitoring and fault prediction system for water chiller | |
| CN118503636A (en) | Power distribution equipment health state assessment method and system | |
| US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
| CN117171657A (en) | Wind power generation equipment fault diagnosis method and device, electronic equipment and storage medium | |
| CN113761233A (en) | Method and device for routing inspection of hydraulic power plant, computer equipment and storage medium | |
| WO2020168379A1 (en) | Power monitoring | |
| CN117782198B (en) | Highway electromechanical equipment operation monitoring method and system based on cloud edge architecture | |
| CN117688464A (en) | Hidden danger analysis method and system based on multi-source sensor data | |
| CN117668646A (en) | Speed reducer running state detection method, system, equipment and medium based on multi-feature fusion | |
| CN108444592A (en) | Wireless vibration monitoring and fault diagnosis system | |
| KR102602840B1 (en) | Training dataset generating system and training dataset generating method using the same | |
| US20230260332A1 (en) | Prognostic and health monitoring system for a device with a rotating component | |
| CN116434368A (en) | Multifunctional handheld inspection instrument and inspection method | |
| CN104236613A (en) | Portable monitoring, diagnosing and on-site verifying system for sensing devices on highway network | |
| CN109271948B (en) | Multi-channel remote vibration signal monitoring system and method based on clustering algorithm | |
| CN117589444B (en) | Wind driven generator gear box fault diagnosis method based on federal learning | |
| Paul et al. | Residual Signature Analysis for Wind Turbine Failure Prognosis | |
| CN219800139U (en) | Transformer fault monitoring device |
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 |