CN109613428A - It is a kind of can be as system and its application in motor device fault detection method - Google Patents

It is a kind of can be as system and its application in motor device fault detection method Download PDF

Info

Publication number
CN109613428A
CN109613428A CN201811519270.8A CN201811519270A CN109613428A CN 109613428 A CN109613428 A CN 109613428A CN 201811519270 A CN201811519270 A CN 201811519270A CN 109613428 A CN109613428 A CN 109613428A
Authority
CN
China
Prior art keywords
data
acquisition
equipment
motor
power
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.)
Pending
Application number
CN201811519270.8A
Other languages
Chinese (zh)
Inventor
唐承佩
王善庆
谭杜康
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Remittance Mdt Infotech Ltd
Original Assignee
Guangzhou Remittance Mdt Infotech Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Guangzhou Remittance Mdt Infotech Ltd filed Critical Guangzhou Remittance Mdt Infotech Ltd
Priority to CN201811519270.8A priority Critical patent/CN109613428A/en
Publication of CN109613428A publication Critical patent/CN109613428A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R31/00Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
    • G01R31/34Testing dynamo-electric machines
    • G01R31/343Testing dynamo-electric machines in operation

Abstract

The invention discloses it is a kind of can be as system and its application in motor device fault detection method, this can include several acquisition device ends, several edge calculations equipment and a cloud server as system, and wherein each edge calculations equipment is communicated to connect at least one acquisition device end;Cloud server is connect with each edge of table computing device communication.The motor device fault detection method using it is described can a kind of sparse self-encoding encoder algorithm of improvement as system and based on deep learning equipment fault sorting algorithm.The present invention can accurately predict the incipient fault that motor device may occur in the process of running, reach and prevent early, plan as a whole the gain effect repaired;It is simultaneously that operand is big, calculating process is complicated, traditional cloud computing method demanding to hardware device is replaced by that operand is small, arithmetic speed is fast, the lower edge calculations method of hardware requirement, improve operation efficiency.

Description

It is a kind of can be as system and its application in motor device fault detection method
Technical field
The present invention relates to fault detection method field, specifically a kind of motor device fault detection method.
Background technique
Core driving device of the induction machine as motor device, plays a crucial role in modern industry.It is special It is not in factory and enterprises such as large-scale pipeline productions, the driving motor in industrial machinery is most important to production automation.Therefore, it drives It is that the sound development gone on smoothly with enterprise is particularly necessary and closes for production work that dynamic motor, which keeps normal operating status, Key.Rotor eccentricity and rotor off-axis are two kinds of main most common failures of induction machine.Rotor is generating centrifugal force Meanwhile the bias for being harmful to equipment life can be generated.And the limited rotor stick of service life is due to overloading and safeguarding improper be easy It produces fracture.Once these failures occur, the entire production line will likely break down and even damage, and cause huge economic loss Even casualties.According to statistics, in China, about 200,000 motors break down every year, and maintenance cost is more than 2,000,000,000 yuan.Just Really the failure of motor is diagnosed in time, find failure in time before generating more serious damage and is overhauled, is Essential link in industrial production activities.
Traditional motor device fault diagnosis technology is mainly based upon intrusive sensor to analyze equipment.However Many common conventional motors equipment do not have onboard data acquisition instrument to obtain new diagnostic data.If upgrading these machines It will be a challenge very serious, because usually not reserved space adds intrusive sensor inside machine.In addition, When device fails, intrusive sensor also vulnerable to damage and is detached from, and the interruption of diagnostic data upload will lead to and set The interruption and erroneous judgement of standby fault diagnosis system operation.In addition, existing most of diagnostic methods are all rule of thumb to believe feature Number carry out artificial selection classification.When using inappropriate characteristic signal, diagnostic data cannot be fully utilized.Meanwhile especially It is the different types of motor device for being operated in dynamic operation environment, the versatility of train classification models is also and accuracy It is also restrained.
Summary of the invention
The present invention to achieve the above object, takes following technical scheme to be achieved:
It is of the present invention to refer to as system in equipment work operating, pass through electric current, voltage, power and the function to equipment The detection of rate factor electrical energy parameter and the current fluctuation of equipment can detect the operating status of analytical equipment, by one section of equipment The data modeling of time, can analytical equipment overall operation life cycle.It is described to be used to mention because of factor of equipment failure as system Preceding early warning is reminded.
It is a kind of can be as system, including several acquisition device ends, several edge calculations equipment and a cloud service Device, each of several described edge calculations equipment edge calculations equipment all at least one acquisition device end communication link It connects, the cloud server is connect with each edge of table computing device communication, and the acquisition device end is used for target object Electrical energy parameter carry out data acquisition and the data of the acquisition be transmitted to edge calculations equipment, the edge calculations equipment is used In the data for receiving and pre-processing the acquisition device end transmission, and pretreated data are transmitted to the cloud service Device.
Further, the electrical energy parameter includes electric current, voltage, power and power factor.
Further, the acquisition device end includes measurement module, data processing module, data transmission module, described Measurement module includes stating current measurement module, voltage measurement module, power measurement module and power-factor measurement module, is used respectively In measure and acquisition target object electric current, voltage, power and with the time series data of power factor, acquire data frequency be 3 seconds/ It is secondary;The data processing module is used to the collected analog signal data of the measurement module being converted to digital signal data; The data transmission module has the function by wirelessly or non-wirelessly network mode sending and receiving data.
Further, the edge calculations equipment includes data transmission module, data computation module;The data calculate mould Block pre-processes the digital signal data from the acquisition device end, and data are transmitted to cloud clothes by treated Business device;The data transmission module has the function by wirelessly or non-wirelessly network mode sending and receiving data.
The invention also includes using it is above-mentioned can as system carry out motor device fault detection method, the method includes with Lower step:
S1: the acquisition device end is electrically connected with motor device to be detected;
S2: the acquisition device end acquires the real-time current of the measurement equipment to be checked, voltage, power and power factor Digital signal data, and it is transmitted through the network to edge calculations equipment;
S3: the edge calculations equipment passes through the digital signal number to the real-time current, voltage, power and power factor According to calculating separately the statistics feature of real-time current, voltage, power and power factor, and obtained data will be calculated and be transferred to Cloud server;
S4: the equipment fault discrimination model based on deep learning algorithm is established by cloud server, uses cloud server The established discrimination model of received data training;
S5: using the discrimination model and cloud server received data in real time to the operating status of motor device into Row differentiates, when differentiating result is that motor device runs well, does not do subsequent processing;When there may be events for differentiation to motor device When barrier, fault alarm is issued to manager, arrestment is powered off in time and runs and carry out subsequent maintenance.
Further, the step S3 statistics feature includes temporal signatures: mean value, variance, the coefficient of variation, and frequency domain is special Sign: low frequency power, high frequency power, low high frequency power ratio, general power, nonlinear characteristic: approximate entropy, fuzzy entropy, Sample Entropy.
Further, the step S4 deep learning algorithm specifically: the equipment for improving sparse self-encoding encoder algorithm (SAE) Failure modes algorithm.
Compared with prior art, the present invention has following gain effect:
It is of the present invention to be intended to as system by the way that multiple acquisition device ends and edge calculations router is arranged, by operation Amount is big, calculating process is complicated, traditional cloud computing method demanding to hardware device be replaced by operand is small, arithmetic speed is fast, The lower edge calculations method of hardware requirement.Meanwhile a variety of electrical energy parameters are acquired to each motor device, to improve conventional method Accuracy, based on different motor device units operation and electrical energy parameter feature diversity and particularity, with solve due to list The low problem of accuracy caused by the limitation of one parameter.Acquire simultaneously device end with non-intrusion type, acquisition it is accurate, can The advantages that connecting internet cloud.The present invention is by using can be as system and the motor device fault detection side based on deep learning Method can effectively improve the accuracy of prediction, the incipient fault that accurately pre- measurement equipment may occur, and reach and prevent early, plan as a whole to repair The gain effect copied.
Detailed description of the invention
Fig. 1 is that the present invention can be as the structural schematic block diagram of system;
Specific embodiment
The present invention will be further described with reference to the examples below, but it should be recognized that embodiment is not to this hair Bright claimed range is construed as limiting.
In the present embodiment, it is a kind of can be as system, including several acquisition device ends, several edge calculations equipment, one Platform cloud server and several motor devices;Each of several edge calculations equipment edge calculations equipment all with At least one acquisition device end is connected by WIFI network, and wherein each acquisition device end is electrically connected with a motor device It connects;The cloud server calculates equipment with each edge of table and is connect by WIFI network.It can be as system and with deep using this Spend the motor device fault detection method of learning algorithm are as follows: acquisition device end is to the motor device to be detected being electrically connected Real-time current, voltage, power and power factor electrical energy parameter are acquired, and obtain the digital signal of electrical energy parameter, and will be digital Signal is transmitted to edge calculations equipment by WIFI network, edge calculations equipment by received from being attached thereto The digital signal of acquisition device end calculates the statistics feature of every group of signal, including temporal signatures: mean value, variance, variation lines Number, frequency domain character: low frequency power, high frequency power, low high frequency power ratio, general power, nonlinear characteristic: approximate entropy, fuzzy entropy, sample This entropy, and result is transferred to cloud server by WIFI network in real time.It is established by cloud server dilute based on improving The equipment fault discrimination model for dredging the equipment fault sorting algorithm of self-encoding encoder algorithm (SAE), with cloud server, institute is received The established discrimination model of data training, is finally passed through using the discrimination model and cloud server received data in real time The operating status of motor device is differentiated, when differentiating result is that equipment runs well, does not do subsequent processing;When differentiation is arrived Equipment issues fault alarm to manager, powers off arrestment in time and run and carry out subsequent maintenance there may be when failure.Base It is constructed in the equipment fault sorting algorithm for improving sparse self-encoding encoder algorithm (SAE) and training deep neural network includes following step It is rapid:
S1: building deep-neural-network, including input layer, hidden layer and output layer;Deep-neural-network is carried out initial Change setting, comprising: setting input layer, hidden layer, the number of plies of each layer of output layer and neuromere points select iterative algorithm for gradient Descent algorithm, and the cost function of iterative algorithm is set, the initial value of step-up error limit value and maximum number of iterations, setting test Accuracy threshold values;
The cost function formula isM is sample size in formula, and footmark i is indicated Sample serial number, parameter θ are the weight vector generated at random of being counted according to every layer of neuromere, hθ(x(i)) indicate pre- with parameter θ and x The prediction y value come is measured, y indicates the standard y value of former training sample;
S2: electrical energy parameter data can be run as the equipment that system acquisition device end collects according to described, comprising setting It is standby to operate normally and multiple classifications of equipment different faults, each classification include multiple and different samples, data sample according to The ratio of 1:1 is divided into training group and test group;
S3: by the set according to stamping behavior class label value and being sent into deep neural network, cost function J is calculated The numerical value of (θ) constantly updates the value of weight vector θ, value of the iteration up to cost function J (θ) by gradient descent algorithm Reach maximum number of iterations less than error limit described in step S1 or the number of iterations, iteration is completed;
S4: testing deep neural network obtained in step S3 using the test group data, is obtaining classification just True sample number calculates test accuracy, its calculation formula is: the correct sample number/total number of samples of test accuracy=classification;
S5: if test accuracy is less than the test accuracy threshold values, the error limit and greatest iteration time are adjusted Number, repeats step S3 to S5;Otherwise training is completed.
Specifically, error limit described in step S1 is set as 10-4, error limit can adjusted again according to test result Whole, the maximum number of iterations is arranged 400 times, and the test accuracy threshold values is set as 95%, the operation of equipment described in step S2 Classification includes: normal operation, rotor eccentricity failure and rotor off-axis failure;
In the present embodiment, database described in S2 can be run using described as the equipment that system acquisition device end collects Electrical energy parameter data set category attribute, rotor off-axis and rotor eccentricity are defined as equipment fault, no rotor off-axis or rotor Bias is defined as equipment normal operation.Building and initialization deep-neural-network, sample different classes of in uploaded data point For training group and test group data, ratio 1:1, cost function formula is
With set according to being trained by the method for above-mentioned steps S3 using to deep neural network, as cost function J The value of (θ) is less than 10-4Or the number of iterations, when reaching 400 times, iteration is completed, these three equipment states are arrived in deep neural network study Classification;Deep neural network is tested with test group data again, tests its standard to the identification of this three kinds of equipment state classifications Exactness trains test sample if deep neural network can correctly identify and test accuracy is more than or equal to 95% It completes;Otherwise, weight vector θ and error limit are modified, continues to be trained by step S3, until training is completed.In this system, Collected new equipment data and the statistics feature being calculated periodically can be uploaded to cloud by the edge calculations equipment Server differentiates that network is detected and sentenced in real time to equipment state by trained depth nerve for cloud server It is disconnected.If cloud server detects equipment, there may be failures, sound an alarm notify manager immediately, and in time to event Hinder the operation of device powers down arrestment, to carry out subsequent maintenance.
In conclusion the present invention can in real time be monitored the operating status of equipment and health status, and can it is accurate, The promptly incipient fault that discovering device is likely to occur, and sounded an alarm in time to manager, to device powers down, after facilitating progress Continuous overhaul of the equipments.Meanwhile acquisition device end has many advantages, such as that non-intrusion type, acquisition is accurate, can connect internet cloud.We Method combination cloud server and edge calculations, while constructing equipment fault discrimination model using deep neural network and can effectively mention The accuracy of height prediction, the failure that accurately pre- measurement equipment may potentially occur reach and prevent early, plan as a whole the gain repaired Effect.
It is provided for the embodiments of the invention technical solution above to be described in detail, specific case used herein The principle and embodiment of the embodiment of the present invention are expounded, the explanation of above embodiments is only applicable to help to understand this The principle of inventive embodiments;At the same time, for those skilled in the art, according to an embodiment of the present invention, in specific embodiment party There will be changes in formula and application range, in conclusion the contents of this specification are not to be construed as limiting the invention.

Claims (7)

1. a kind of can be as system, it is characterised in that: including several acquisition device ends, several edge calculations equipment and one Cloud server, each of several described edge calculations equipment edge calculations equipment are all whole at least one acquisition equipment End communication connection, the cloud server connect with each edge of table computing device communication, and the acquisition device end is used for pair The electrical energy parameter of target object carries out data acquisition and the data of the acquisition is transmitted to edge calculations equipment, the edge meter The data that equipment is used to receive and pre-process the acquisition device end transmission are calculated, and pretreated data are transmitted to the cloud Hold server.
2. according to claim 1 can be as system, it is characterised in that: the electrical energy parameter include electric current, voltage, power and Power factor.
3. according to claim 1 can be as system, it is characterised in that: the acquisition device end includes measurement module, number According to processing module, data transmission module;The measurement module includes stating current measurement module, voltage measurement module, power measurement Module and power-factor measurement module, be respectively used to measure and acquire target object electric current, voltage, power and with power factor Time series data, acquisition data frequency are 3 seconds/time;The data processing module is used for the collected simulation of the measurement module Signal data is converted to digital signal data;The data transmission module, which has, passes through wirelessly or non-wirelessly network mode sending and receiving data Function.
4. according to it is according to claim 1 can be as system, it is characterised in that: the edge calculations equipment include data transmission Module, data computation module;The data computation module carries out the digital signal data from the acquisition device end pre- Processing, and data are transmitted to cloud server by treated;The data transmission module has through wirelessly or non-wirelessly network The function of mode sending and receiving data.
5. a kind of can exist using claim 1-4 is described in any item as the motor device fault detection method of system, feature In: it the described method comprises the following steps:
S1: the acquisition device end is electrically connected with motor device to be detected;
S2: the acquisition device end acquires the number of the real-time current of the measurement equipment to be checked, voltage, power and power factor Signal data, and it is transmitted through the network to edge calculations equipment;
S3: the edge calculations equipment passes through the digital signal data to the real-time current, voltage, power and power factor point Not Ji Suan real-time current, voltage, power and power factor statistics feature, and obtained data will be calculated and be transferred to cloud Server;
S4: the equipment fault discrimination model based on deep learning algorithm is established by cloud server, is connect with cloud server The established discrimination model of data training of receipts;
S5: the operating status of motor device is sentenced in real time using the discrimination model and cloud server received data Not, when differentiating result is that motor device runs well, subsequent processing is not done;When there may be failures for differentiation to motor device When, fault alarm is issued to manager, arrestment is powered off in time and runs and carry out subsequent maintenance.
6. motor device fault detection method according to claim 5, it is characterised in that: the step S4 statistics feature Including temporal signatures: mean value, variance, the coefficient of variation, frequency domain character: low frequency power, high frequency power, low high frequency power ratio, total work Rate, nonlinear characteristic: approximate entropy, fuzzy entropy, Sample Entropy.
7. motor device fault detection method according to claim 5, it is characterised in that: the step S6 deep learning is calculated Method specifically: improve the equipment fault sorting algorithm of sparse self-encoding encoder algorithm (SAE).
CN201811519270.8A 2018-12-12 2018-12-12 It is a kind of can be as system and its application in motor device fault detection method Pending CN109613428A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811519270.8A CN109613428A (en) 2018-12-12 2018-12-12 It is a kind of can be as system and its application in motor device fault detection method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811519270.8A CN109613428A (en) 2018-12-12 2018-12-12 It is a kind of can be as system and its application in motor device fault detection method

Publications (1)

Publication Number Publication Date
CN109613428A true CN109613428A (en) 2019-04-12

Family

ID=66007224

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811519270.8A Pending CN109613428A (en) 2018-12-12 2018-12-12 It is a kind of can be as system and its application in motor device fault detection method

Country Status (1)

Country Link
CN (1) CN109613428A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110008898A (en) * 2019-04-02 2019-07-12 中国计量大学 Industrial equipment data edges processing method based on symbol and convolutional neural networks
CN110197128A (en) * 2019-05-08 2019-09-03 华南理工大学 The recognition of face architecture design method planned as a whole based on edge calculations and cloud
CN110262421A (en) * 2019-06-21 2019-09-20 深圳市美兆环境股份有限公司 The control method and device of production equipment
CN110749462A (en) * 2019-07-19 2020-02-04 华瑞新智科技(北京)有限公司 Industrial equipment fault detection method and system based on edge calculation
CN112015153A (en) * 2020-09-09 2020-12-01 江南大学 System and method for detecting abnormity of sterile filling production line

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160247153A1 (en) * 2015-02-20 2016-08-25 Innovative Global Systems, Llc Automated at-the-pump system and method for managing vehicle fuel purchases
CN106443447A (en) * 2016-09-26 2017-02-22 南京航空航天大学 An aero-generator fault feature extraction method based on iSDAE
CN107895176A (en) * 2017-11-13 2018-04-10 国网湖南省电力有限公司 A kind of mist computing system and method towards water power group of planes wide area monitoring, diagnosing
CN107957551A (en) * 2017-12-12 2018-04-24 南京信息工程大学 Stacking noise reduction own coding Method of Motor Fault Diagnosis based on vibration and current signal
CN108591104A (en) * 2018-04-18 2018-09-28 广东寰球智能科技有限公司 A kind of Research on Fan Fault Forecasting based on cloud platform and health management system arranged, method
CN108828409A (en) * 2018-08-03 2018-11-16 南方电网科学研究院有限责任公司 A kind of fault detection system based on edge calculations

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160247153A1 (en) * 2015-02-20 2016-08-25 Innovative Global Systems, Llc Automated at-the-pump system and method for managing vehicle fuel purchases
CN106443447A (en) * 2016-09-26 2017-02-22 南京航空航天大学 An aero-generator fault feature extraction method based on iSDAE
CN107895176A (en) * 2017-11-13 2018-04-10 国网湖南省电力有限公司 A kind of mist computing system and method towards water power group of planes wide area monitoring, diagnosing
CN107957551A (en) * 2017-12-12 2018-04-24 南京信息工程大学 Stacking noise reduction own coding Method of Motor Fault Diagnosis based on vibration and current signal
CN108591104A (en) * 2018-04-18 2018-09-28 广东寰球智能科技有限公司 A kind of Research on Fan Fault Forecasting based on cloud platform and health management system arranged, method
CN108828409A (en) * 2018-08-03 2018-11-16 南方电网科学研究院有限责任公司 A kind of fault detection system based on edge calculations

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
WENJUN SUN等: "A sparse auto-encoder-based deep neural network approach for induction motor faults classification", 《MEASUREMENT》 *
周洪发: "基于云平台的电机设备远程运维系统设计", 《电机与控制应用》 *

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110008898A (en) * 2019-04-02 2019-07-12 中国计量大学 Industrial equipment data edges processing method based on symbol and convolutional neural networks
CN110197128A (en) * 2019-05-08 2019-09-03 华南理工大学 The recognition of face architecture design method planned as a whole based on edge calculations and cloud
CN110262421A (en) * 2019-06-21 2019-09-20 深圳市美兆环境股份有限公司 The control method and device of production equipment
CN110749462A (en) * 2019-07-19 2020-02-04 华瑞新智科技(北京)有限公司 Industrial equipment fault detection method and system based on edge calculation
CN110749462B (en) * 2019-07-19 2021-05-07 华瑞新智科技(北京)有限公司 Industrial equipment fault detection method and system based on edge calculation
CN112015153A (en) * 2020-09-09 2020-12-01 江南大学 System and method for detecting abnormity of sterile filling production line
CN112015153B (en) * 2020-09-09 2021-06-22 江南大学 System and method for detecting abnormity of sterile filling production line

Similar Documents

Publication Publication Date Title
CN109613428A (en) It is a kind of can be as system and its application in motor device fault detection method
CN107003663B (en) The monitoring of device with movable part
CN103754718B (en) Lift running safety monitoring system and method thereof
CN107247230B (en) Rotary motor state monitoring method based on support vector machine and data driving
CN104573850A (en) Method for evaluating state of thermal power plant equipment
CN107357275B (en) Non-gaussian industrial process fault detection method and system
JP2004523735A (en) Inference signal generator for instrumented equipment and processes
CN106407589B (en) Fan state evaluation and prediction method and system
CN106006344B (en) Staircase On-line Fault early warning system and method for diagnosing faults
CN1862278A (en) Method and system for predicting remaining life for motors featuring on-line insulation condition monitor
EP3441947B1 (en) System and method for remaining useful life determination
CN107016235A (en) The equipment running status health degree appraisal procedure adaptively merged based on multiple features
CN108921303A (en) The Fault diagnosis and forecast maintaining method of industrial motor
CN105138770B (en) Space product Reliablility simulation appraisal procedure based on indirect characteristic quantities
CN107844067A (en) A kind of gate of hydropower station on-line condition monitoring control method and monitoring system
CN106441843B (en) A kind of rotating machinery fault method for waveform identification
Smarsly et al. A decentralized approach towards autonomous fault detection in wireless structural health monitoring systems
CN110376522A (en) A kind of Method of Motor Fault Diagnosis of the deep learning network of data fusion
CN106762343A (en) The diagnostic method of the hydraulic generator set thrust bearing failure based on online data
CN107036808A (en) Gearbox of wind turbine combined failure diagnostic method based on SVMs probability Estimation
Qin et al. Remaining useful life prediction for rotating machinery based on optimal degradation indicator
CN109061387A (en) A kind of power grid abnormality method of discrimination based on annulus theorem
CN109990803A (en) The method, apparatus of method, apparatus and the sensor processing of detection system exception
CN108982106A (en) A kind of effective ways of quick detection of complex system dynamics mutation
CN107330264A (en) A kind of verification method of bridge monitoring data reliability

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