CN105873111A - Soft and hard fault diagnosis and self restoration method suitable for health monitoring - Google Patents
Soft and hard fault diagnosis and self restoration method suitable for health monitoring Download PDFInfo
- Publication number
- CN105873111A CN105873111A CN201610405559.1A CN201610405559A CN105873111A CN 105873111 A CN105873111 A CN 105873111A CN 201610405559 A CN201610405559 A CN 201610405559A CN 105873111 A CN105873111 A CN 105873111A
- Authority
- CN
- China
- Prior art keywords
- node
- network
- monitoring
- fault
- data
- 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
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/02—Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
- A61B5/0205—Simultaneously evaluating both cardiovascular conditions and different types of body conditions, e.g. heart and respiratory condition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Cardiology (AREA)
- Physiology (AREA)
- Heart & Thoracic Surgery (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Pulmonology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Physics & Mathematics (AREA)
- Pathology (AREA)
- Computing Systems (AREA)
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
The invention discloses a soft and hard fault diagnosis and self restoration method suitable for health monitoring. According to the method, a route link is formed; the head node of the link is a cluster head; the second node is a spare cluster head, and the like; a method provided by M.A-B.Abdo is used; two kinds of feature parameters respectively used for node fault detection and structure damage recognition are extracted from original collection data; the interference of the fault node on the structure damage is avoided. A self restoration wireless sensor network structure system is planned to use a mixed grading topology structure; besides the original sensor node, a monitoring node used for monitoring the network state is considered to be arranged in a network. The method has the advantages that the cooperative communication mechanism is applied to the multi-stream transmission problem of a wireless network for realizing the network maximum throughput; the stability and the reliability of the network are ensured.
Description
Technical field
The present invention relates to diagnosis and the self-repair method of a kind of soft or hard fault being suitable for health monitoring, belong to wireless sensor network technology field.
Background technology
Monitoring structural health conditions (Structural Health Monitoring, SHM) system uses the concept of Intelligent material structure, utilize integrated sensing/driving element networks in the structure, the online information that acquisition is relevant to structural health in real time, binding signal, information processing method and material structure mechanical modeling method, extract characteristic parameter, identify the structural damage (such as material crack, hole, corrosion etc.) of material, it is achieved structural health self diagnosis.The research of monitoring structural health conditions is a forward position research direction relating to multiple field of scientific studies such as mechanics, machinery, communication, network, has been directed to the research of structural health monitoring technology in the heavy construction structures such as aviation field, bridge, engineering construction.If the U.S. is under the subsidy of USAF, carry out the explorative research research of structural health monitoring technology for the aircraft such as F-18, F-22, JSF and DC-X2, X-33.The health monitoring for bridge of 350 sensing passages is arranged in Hong Kong on Tsing Ma Bridge.
Traditional monitoring structural health conditions is set up mostly on the basis of wired sensor gathers, and wired monitoring system and method also exists that lead-in wire is many and the problem such as quantity of information transmission is big, and needed for safeguarding, man power and material is the hugest.For supplying a gap, recent domestic scholar proposes based on wireless sensor network (Wireless Sensor Networks, WSNs) structural healthy monitoring system, by being deployed in the wireless sensing node of monitored area, perception in real time and the relevant information of collection monitoring object, carry out collaborative process and network transmission to information.WSNs has rapid deployment, organizes themselves into net, stronger anti-ruins and the advantages such as collaborative work ability, is the hot fields of Chinese scholars research.And intelligent health monitoring system of based on radio sensing network has obtained concern and the attention of the most increasing research institution, achieve numerous application achievements in fields such as aviation, bridge, building and seismic structural monitorings.
When using wireless sensor network that the heavy construction structures such as aviation, bridge, engineering construction are monitored monitoring, system is normally operated in rugged environment, its node outer can occur various fault owing to being exposed to, directly contribute the mistake of measured value, even cause some afunction of WSN or even whole network paralysis.Additionally, in the specific field of health monitoring, the application of radio sensing network has a following characteristics:
(1) sensor is distributed in measurand key position, once node deployment, would not move, it is not necessary to consider node mobility.
(2) different with other applications of WSN, the sensor node in monitoring structural health conditions works under fixed frequency, and the original data volume that each node gathers is the biggest.Therefore it is crucial for gathering transducing signal the most accurately, and network needs real time reaction fast, reliable communications, completes to gather and forward in the shortest time.
(3) for completing health monitoring task, some key monitoring point needs to gather multiple transducing signal (such as Lamb wave, light, strain, displacement, acceleration, temperature, pressure etc.), and therefore WSN should possess the function of multi-parameters sampling.
(4) in health monitoring, non-destructive tests must have been cooperated, even if one of them nodes break down also results in the drastically decline of non-destructive tests precision jointly by the multiple sensor node of surrounding.
Above feature shows, in monitoring structural health conditions, has higher requirement wireless sensor node and network；I.e. system must be able to real-time, accurate, stable, reliable offer for health monitoring and gathers data, and fault directly affects the result of monitoring.WSN node is possessed the requirement of multi-parameters sampling, also increases the probability that fault occurs.Therefore, in order to improve the reliability of monitoring structural health conditions, node, the diagnosing malfunction of network and reparation are very important promptly and accurately.
At present, the method for traditional sensors fault detection and diagnosis mainly has three major types, physically-redundant method, method based on model and method based on neutral net.Although the nodal fault diagnostics in wireless sensor network is limited by radio communication, Energy Efficient etc., it is slightly different on diagnostic method in detection, but still can use for reference the diagnostic techniques of traditional sensors.The node failure of wireless sensor network is divided into two classes: hard fault and soft fault.Hard fault, refer to that a certain module of sensor node breaks down so that can not with other node communication (such as due to node communication module failure, node energy exhaust, node motion and cannot communicate departing from what the reason such as communication range of whole network caused)；Soft fault, although referring to that sensor node breaks down, but still can work on and with other node communication (software and hardware of communication module all normal and there is routing iinformation), but the data of node institute's perception or transmission are incorrect, or sensor node instantaneous generation communication failure.In the nodal fault diagnostics method of existing wireless sensor network, document proposes node failure recognizer based on Bayesian network, can distinguish fault types different in network.J.L.Gao et al. proposes malfunctioning node marking algorithm, by the average ratio with adjacent node relatively, carrys out the state of Judging fault node；X.Luo et al. proposes the fault-toleranr technique with Energy Efficient, for the fault detect of radio sensing network；Document is for the appearance of fault in network, it is proposed that the confirmation algorithm of autonomous aggregation node, reduces the node failure impact on overall network performance.J.R.Chen et al. proposes the DFD algorithm of the method diagnosis node state utilizing adjacent node to test mutually.Document proposes attribute in wireless sensor network program and violates the diagnostic method of type error.
In terms of the wireless sensor network fault diagnosis of structure-oriented health monitoring, Chinese scholars research in this respect at the early-stage, the document of this respect is less.Document, for monitoring structural health conditions based on vibration, gives the mathematical model of sensor node fault and structural damage, is given up by malfunctioning node from network, it is to avoid the interference of node failure；X.Liu et al. proposes the fault tolerant technique of monitoring structural health conditions based on WSN, it is achieved the fault detect of node；Document proposes the self-healing property research of wireless sensing system based on bridge structural health monitoring, proposes, when relaying node failure, to give up this node, realize network self-healing function in standby via node mode.
To sum up, the most existing research is both for greatly the diagnosis of node soft fault, and the process being not involved with hard fault also fails to relate to the reparation of fault.Hard fault once occurs, and owing to the hardware of wireless sensor node is all fixing, unless node is removed from sensor network and redesigned the software and hardware system of node, otherwise node can not reconfigure.The most furthermore, existing soft fault diagnosis algorithm is often based on the exchange of internodal data, and big due to the data volume that gathers between node and exchange in monitoring structural health conditions, thus cannot be suitable for.The thinking replaced is to carry out fault diagnosis after the data compression to node.Under above two situation, deploying once sensor network and the most again can not change configuration, existing node cannot carry out selfreparing and via Self-reconfiguration when losing efficacy, and can only give up failure node.Therefore, it is necessary to the fault for the sensor network in this field proposes effective hardware and software failure diagnostic method and selfreparing mechanism.
(1) present Research of the hard fault of wireless sensor network in SHM
In actual biosphere, via Self-reconfiguration and self-repair function generally exist, if wireless sensor network all has bionic function from network node to network struture system, possesses the function (such as biological immune ability) that similar biology is had, can via Self-reconfiguration and selfreparing, then great for improving the robustness of wireless sensor network and security sense.And the research for the fault diagnosis of the wireless sense network of monitoring structural health conditions is of practical significance very much, the research of this specific application area is the most at the early-stage, and the research to the fault diagnosis in this field has the biggest space and value.
Wireless sensor network to be realized all has bionic function from network node to network struture system, possesses biological immune ability, can use for reference existing bionic hardware technology.So-called bionic hardware (BHW, Bio-inspired Hardware) is to realize the electronic circuit self reconstruct real-time in system, such that it is able to have hardware self-adapting, self-organizing, selfreparing feature as biology by evolutionary mechanism.Currently, the research of bionic hardware technology has been carried out more than 10 years, and its basic thought is to use the restructing device mimic biology evolutionary mechanism able to programme such as on-site programmable gate array FPGA s and field programmable analog array FPAAs to realize Bio-simulated Evolution hardware circuit.In recent years, the scholar of the states such as the U.S., Germany, Britain had studied and had been applied in radio sensing network by bionic hardware, and the research of bionical sensing network based on selfreparing is increasingly becoming an important research direction of sensor network.
(2) present Research of the data compression of wireless sensor network in SHM
In monitoring structural health conditions, traditional soft fault diagnosis method, it is limited to the exchange of internodal substantial amounts of data.Solution is the data of node to be compressed, current compression method includes: 2006, Lehigh university of U.S. building and Y.F.Zhang. and J.Li of environment engineering, propose and damage data compression algorithm based on what Lifting Wavelet changed, for eliminating the temporal correlation that node is gathered, it is thus achieved that good compression effectiveness.In the same year, Y.F.Zhang and J.Li et al. is in order to realize process and the research of seismic response data, it is proposed that a kind of data compression method based on ARX model (Auto-Regressive with eXogenous input model)；2006, the K.K.Chintalapudi of American South University of California proposes linear predictive coding (Linear Predictive Coding, LPC) data compression scheme, the method is lossless data compression, uses prediction algorithm based on auto regressive moving average (ARMA) model that the data gathered are carried out lossless compress.2007, Y.F.Zhang. improve further on the basis of ARX model with J.Li, use autoregression (Auto-Regressive, AR) as model structure, use method of instrumental variable (instrumental variables method, IV) calculate Prediction Parameters, and propose compression algorithm based on linear prediction, it is achieved the lossless compress to data.On the other hand, in structure monitoring field based on wireless sensor network, N.Xu et al. it is also proposed wavelet data compression method based on local node in wireless senser, solves the problem that in structure monitoring, wireless sensor data transmission bandwidth limits.J.P.Lynch et al. have studied the volume of transmitted data of the wireless senser using Huffman coding minimizing monitoring structural health conditions.But above-described data compression method is all belonging to traditional compression algorithm, the most first obtain complete collection data, be then compressed data processing.At present, D.Donoho proposes a kind of new compression sampling technology, referred to as compressed sensing (Compressive sensing, CS), can the data of direct collect and process form.The compression of the narrow band signal in the monitoring of the method very suitable structures, although the research ground zero in the wireless sensor network field in structure-oriented monitoring, but there is good application prospect.
Above present Research is for the invention provides thinking, and we will be on the basis of data compression and bionic hardware, for the fault diagnosis of the wireless sensor network in monitoring structural health conditions field, it is provided that effective hardware and software failure diagnostic method and selfreparing mechanism.
The limitation of existing research work is as described in " present Research ", and currently the fault diagnosis for wireless sensor network has had a lot of fruitful research, also has scholar to attempt being applied to Artificial Immunology Mechanism the fault diagnosis of radio sensing network.But owing to monitoring structural health conditions based on WSN exists the particularity of its application so that the fault diagnosis in this field remains and fails to obtain system solution at many critical problems, needs to be studied further.Particular problem is as follows:
(1) extraction of fault diagnosis interior joint interchange of data.Existing fault diagnosis technologies etc., utilize the spatial simlanty of neighbors to carry out data interchange and realize the fault diagnosis of node.But in monitoring structural health conditions based on WSN, node gathers substantial amounts of original sample with fixed frequency, and substantial amounts of data interchange can exhaust the energy of network every time.Effective data compression method is crucial problem to be solved, and from original sample, extraction can identify that the characteristic parameter of abnormal data and normal damage data is the emphasis required study in addition.
(2) fault diagnosis and self-repair method is integrated.Existing method for diagnosing faults is both for greatly the diagnosis of individual node soft fault, and is not involved with the process of hard fault.And the wireless sense network fault diagnosis of structure-oriented health monitoring, due to its field, particularly urgent for the diagnosis of hardware and software failure and the demand of selfreparing, need in-depth fault diagnosis and the research repairing integrated approach further.
(3) power of test that multiple sensors break down simultaneously.Being different from other applications, in health monitoring, non-destructive tests must have been cooperated jointly by the multiple sensor node of surrounding, even if the precision that one of them nodes break down also results in loss identification drastically declines.Multisensor break down simultaneously even localized network fault diagnosis with repair need deeper into research.
In sum, the present invention is by the specific application area of the monitoring structural health conditions for radio sensing network, on the basis of invention group early-stage Study, sum up and use for reference existing fault diagnosis technology, propose to study soft, the diagnosis of hard fault of a kind of wireless sensor network being suitable for health monitoring and selfreparing is theoretical and method.This research is for improving selfreparing and the fault-tolerant ability of the wireless sensor network reply anomalous event such as network attack, node failure, and the robustness and the safety that improve wireless sensor network in SHM are significant.
Summary of the invention
Goal of the invention: (1) solves the identification problem of the fault message under many abnormal datas overlapping: this is the premise realizing fault diagnosis, the purpose of fault diagnosis is to judge the health status of material to accurately extract the damage data of structural material, and the gauger of malfunctioning node is huge on material damage identification impact, therefore need to analyze essential distinctions and the contact of two kinds of overlapping measured values, find and solve abnormal data to the interference of normal damage data and recognition methods.
(2) compression problem of the data exchange of network failure lower sensor node is solved: this is the effective ways expanding traditional soft fault diagnosis.Effective data compression and acquisition method, can make traditional soft fault diagnosis method be extended to monitoring structural health conditions field.It is thus desirable to analyze and study a kind of feasible data compression acquisition method.
(3) solving the Construct question of selfreparing trigger condition under multinode fault: this is the strong approach realizing fault diagnosis selfreparing, when multiple sensor nodes break down simultaneously, different trigger conditions should realize different selfreparing mechanism.I.e. to node selfreparing, network selfreparing, and the determination of the trigger condition (threshold limit value) of network rerouting.
Fault detection problem in wireless sensor network under invention research structure health monitoring, through the summary of newest research results domestic and international to this area, binding target, the research contents of the present invention is as shown in Figure 1.
The research of routing mechanism during efficient real in 1:SHM: require in monitoring structural health conditions that wireless sensor network must be able to real-time, accurate, stable, reliably for health monitoring offer collection data.It is therefore desirable to routing mechanism during efficient real in SHM is studied.Concrete research includes:
(1) research of multiple stream transmission in SHM.In monitoring structural health conditions based on wireless sensor network, the reliability and stability of network transmission are the bases of various application.But owing to being limited by bandwidth, through-put power in being wirelessly transferred, and the impact of signal fadeout (Signal Fading) effect, the performance of wireless system can be substantially reduced, including power system capacity, efficiency of transmission, service quality and energy efficiency etc..Solution can use the collaboration communication mechanism of multiple stream transmission, can realize the maximum throughput of network, it is ensured that the stability of network and reliability.In SHM, the problem of multiple stream transmission is one of basic content of studying of the present invention.
(2) research of Clustering Routing in monitoring structural health conditions.The alternative method of spare cluster head during research leader cluster node fault；
(3) routing mechanism is to structural healthy monitoring system and the performance evaluation of the fault diagnosis of network.
The abnormal data of the 2:SHM interior joint interference analysis to structural damage measured value: as " project verification foundation " carries, the structural damage of material refers to the change of the aspects such as the crackle of material, hole, the identification of damage and location must have been worked in coordination with by the multiple sensor node of surrounding, and this is the important feature of monitoring structural health conditions；When node produces measurement error because sensing module breaks down, the precision of Damage Assessment Method location can be severely influenced.Now temporarily should reject to avoid interference by this malfunctioning node, carry out non-destructive tests location the most again, so detecting that malfunctioning node becomes crucial timely.
Existing WSN node failure detection means mostly needs to gather data by exchange between node and detects fault.But in SHM, as " project verification foundation " put forward " feature two ", node every time can with fixed frequency gather substantial amounts of original sample, substantial amounts of data interchange can exhaust the energy of network.So for the fault of sensing element, existing WSN node failure detection means is in monitoring structural health conditions and the most applicable.
Therefore, from acquired original sample, extraction can identify abnormal data and the characteristic parameter of normal damage data, it is achieved node failure detects, and is the precondition of this research.
The soft fault diagnosis of wireless sensor node based on compressed sensing in 3:SHM: in monitoring structural health conditions based on wireless sensor network, although node is fixing without considering its mobility, it might even be possible to manual configuration route.But in view of motility and the robustness of this field nodal fault diagnostics, use clustering routing mechanism that fault management can be made to complete in being distributed to respective bunch of region.It is proposed that be applicable to the soft fault diagnosis method of the node based on compressed sensing (containing bunch head) of monitoring structural health conditions.Concrete research includes:
(1) node gathers the Compression Study of data；The reconstruction property of compression algorithm is to monitoring structural health conditions and the impact analysis of network fault diagnosis；
(2) method for diagnosing faults of node.When in research node generation instantaneous communication soft fault or structural material, not damaged occurs, the method for diagnosing faults of ordinary node and routing algorithm.Intend with reference to and improve existing neighbors collaborative process way to adapt to the specific area of the present invention, it is preferred that emphasis is the type of exchange data when improving neighbors cooperation.
The hard fault diagnosis mechanism of sensor node and the design of bionical selfreparing node in 4:SHM: for feature and the particular demands of monitoring structural health conditions of radio sensing network, bionic hardware is applied on the node of sensor reviewed one's lessons by oneself, it is necessary to meet the features such as low cost, volume be little.The advantage the most how utilizing bionic hardware, proposes to be applicable to the selfreparing mechanism of wireless sensor node hard fault, is one of the emphasis of this research.Concrete research includes:
(1) selfreparing node software and hardware architecture framework；
(2) selection of self-repair module and method to set up, studies for sensing elements such as the strain of SHM, Lamb wave；
(3) design of selfreparing sensor node and realization, hard fault diagnosis and the via Self-reconfiguration flow process of node；
(4) performance test of selfreparing sensor node, including: the performances such as the power consumption of node, analog digital conversion, transmission range, signal spectrum.
Technical solution of the present invention is as follows:
The present invention is directed to the feature of the specific application area of monitoring structural health conditions, it is provided that a kind of soft, diagnosis of hard fault being suitable for health monitoring and selfreparing theory and method.When studying efficient real on the basis of routing mechanism, for the diagnosis of soft fault, by the support of compressive sensing theory, propose to use neighbours' cooperation and the method for clustering routing, ordinary node and leader cluster node are carried out soft defect detection；For hard fault, theoretical according to bionical selfreparing, study wireless sensor network network fault diagnosis in the case of node failure, subnetwork inefficacy and external node invasion and selfreparing is theoretical and method.Selfreparing and the fault-tolerant ability of the wireless sensor network reply anomalous event such as network attack, node failure is improved by the studies above.
The diagnosis of the soft or hard fault of health monitoring and self-repair method comprise step in detail below:
Preliminary scene setting:
Step 1) it route during efficient real in SHM: based on the collaboration communication mechanism of multiple stream transmission, designing distributed clustering routing algorithm, form routing link, the first node of link is a bunch head, and secondary nodal point is that spare cluster is first-class.
Step 2) in SHM the interference of abnormal data process: combine the feature extracting method of the structure monitoring vibration signal of invention group early-stage Study, the method intending using M.A-B.Abdo to propose, go out to be respectively used to node failure detection and two kinds of characteristic parameters of Damage Assessment Method from acquired original extracting data, after the feature extraction obtaining natural frequency, according to two kinds of measured value localities and theory of overall importance, intend using sample statistics algorithm, or conventional node cooperation fault diagnosis algorithm, it is achieved node failure exceptional value and the differentiation of structural damage measured value and extraction.
The monitoring vibration signal obtained from sensor, uses Gabor Computed order tracking and Viterbi maximum-likelihood decoding algorithm to carry out natural frequency extraction, uses sample statistics algorithm, or conventional node cooperation fault diagnosis algorithm comes recognition node fault value and structural damage value.
Step 3) diagnosis mechanism in SHM: in monitoring structural health conditions, during node sample, use compression sampling based on compressed sensing, to adapt to the exchange of internodal data.For the measured value of sensor node, use one and orthogonal basis Ψ ∈ RN × NIncoherent matrix Φ ∈ RM × N(M ＜ ＜ N), projects on a lower dimensional space, it is achieved the compression of node sample signal by high dimensional signal.
Step 4) hardware structure: in self-repairing wireless sensing device node hardware framework, selecting programming device as Embedded solution, main functional modules is connected as module by crossing bionic hardware FPAAs.Using bionic hardware, field programmable analog array FPAAs realizes the signal link of the sensing module in node, and designs fault diagnosis functions and the selfreparing control function of Inductive links in the signal processing module of node.
Step 5) software architecture: it is compared provide diagnostic result with the threshold value set by abnormal signal diagnotor after binary data in A/D translation register is converted into decimal data by data acquisition driver；Master control programme sends drive command according to diagnostic result to FPAA driver；First this program reads the FPAA configuration file resided in outside flash storage according to drive command, then configures FPAA dynamically so that it is complete sensing and the reconstruct of redundant layer signal link thereof.
Step 6) performance optimization: self-repairing wireless sensing device network struture system uses mixing-classifying topological structure, arranges the monitor node being used for monitoring network state in a network.
Beneficial effect
Compared with the fault diagnosis research of existing wireless sensor network, characteristic is with innovation:
1) in monitoring structural health conditions, soft, the hard fault of wireless sensor network is diagnosed and repair, identify the abnormal data of malfunctioning node and normal damage data, improve robustness and the accuracy of system monitoring.
2) bionic hardware is applied in the sensing element of sensor node realize fault diagnosis and the selfreparing of node, it is considered to the engineering demand of actual SHM, multiple sensor simultaneous faults can be solved and cause the critical problem of network failure.
3) use compression sampling and the mechanism of multiple stream transmission, improve fault diagnosis and the performance of data transmission.
Accompanying drawing explanation
The research contents of Fig. 1 present invention.
Fig. 2 invents the total technology path intending taking.
Fig. 3 is the soft fault diagnosis mechanism of wireless sensor node
Fig. 4 intends the self-repairing wireless sensing device node hardware framework taked.
The self-repairing wireless sensing device node software framework that Fig. 5 drafts.
Detailed description of the invention
Technical scheme is illustrated below in conjunction with accompanying drawing.
Invent the soft and hardware diagnosing malfunction for the wireless sensor network in monitoring structural health conditions field and reparation, the research design of selfreparing node of bionic hardware emphatically, the abnormal data of malfunctioning node and the differentiation of normal data, soft fault diagnosis based on compressed sensing, node and the content such as the fault diagnosis of network and repair mechanism, be illustrated in figure 1 research contents of the present invention.Invention uses the research approach that theory analysis and experimental verification combine, first obtain sensor node or the fault type of network in SHM, malfunctioning node number and threshold limit value are compared, thus node or subnetwork are repaired respectively, whole study route is as shown in Figure 2, wherein parameter ψ is designated the persistent period of hard fault for predicate node soft fault, φ 1 represents the threshold limit value of the malfunctioning node number needing again clustering routing in SHM, and φ 2 is the threshold limit value of the malfunctioning node number that network failure occurs in SHM.
(1): the research of routing mechanism during efficient real in SHM
Sensor node for SHM is fixing without considering ambulant situation, when the routing mechanism of SHM and soft fault diagnosis algorithm should possess stable, real-time, feature that volume of transmitted data is little.Therefore on routing mechanism, consider multiple stream transmission.
Our thinking is the multiple stream transmission problem that collaboration communication mechanism is applied to wireless network, to improve its transmission performance.The technology path intending using is: utilize the method such as linear programming, integer program that problem is carried out formalized description；The method using stipulations proves that subproblem is that NP is difficult；Use the technology such as the saturated coupling of bipartite graph, Dijkstra method, branch and bound method, dynamic programming to design effective algorithm and agreement, and utilize the mathematical measures such as binomial relevant nature to carry out the theoretical performance of parser.
On the basis of the collaboration communication mechanism of above-mentioned multiple stream transmission, consider the actual demand of monitoring structural health conditions, inspired by distributed level Molecule cluster method and thought, the scheme that we intend taking is: design distributed clustering routing algorithm, form routing link, the first node of link is a bunch head, and secondary nodal point is that spare cluster is first-class.
(2): in SHM, the interference of structural damage measured value is processed by the abnormal data of malfunctioning node
Node failure directly affects the accuracy of non-destructive tests to the interference of structural damage, and normal structural damage measured value brings interference often to the fault diagnosis of sensor network, and formation is obscured.Thus extract from acquired original sample and can identify abnormal data and the characteristic parameter of normal damage data, it is the key of fault diagnosis.
The inspiring of mathematical model by SHM interior joint fault and structural damage, feature extracting method in conjunction with the structure monitoring vibration signal of invention group early-stage Study, the method that invention is intended using M.A-B.Abdo to propose, go out to be respectively used to node failure detection and two kinds of characteristic parameters of Damage Assessment Method from acquired original extracting data, it is to avoid the malfunctioning node interference to structural damage.The technical thought intending using is as follows:
Monitoring structural health conditions interior joint fault measuring value and structural damage measured value belong to weak signal, and interfere, mutually obscure.Both it is unfavorable for the fault diagnosis of network, also have impact on non-destructive tests in monitoring structural health conditions.Theory according to document: " in monitoring structural health conditions, the Main Means of damage identification and location is that the structural vibration feature (natural frequency) by signal realizes, and the natural frequency of the natural frequency of malfunctioning node measured value and healthy node measurement value exists obvious difference；Former is local, and latter is overall.”
Above-mentioned theory is for distinguishing malfunctioning node measured value and normal damage measurement value, it is achieved fault diagnosis provides important evidence.Therefore, we intend utilizing the Gabor coefficient C representing structure vibration signals time-frequency characteristicsm,n, use Gabor Computed order tracking method, obtain discrete time point and the time frequency grid face of discrete Frequency point composition, use the optimal frequency path between Viterbi algorithm hunting time point, thus the frequecy characteristic realizing sensor measurement signal extracts.
After the feature extraction obtaining natural frequency, according to two kinds of measured value localities and theory of overall importance, intend using sample statistics algorithm, or conventional node cooperation fault diagnosis algorithm, it is achieved node failure exceptional value and the differentiation of structural damage measured value and extraction.The scheme tentatively drafted is as shown in Figure 3.
(3): the soft fault diagnosis mechanism of wireless sensor node based on compressed sensing in SHM
The weighted median method for diagnosing faults combining invention group early-stage Study in soft fault diagnosis algorithm realizes the fault detect of ordinary node.The situation when program occurs mainly for not damaged in instantaneous communication soft fault or structural material, routing link once generates and keeps the fixing triggering until φ 1 threshold limit value.The technical thought intending using is as follows:
There is the dependency in time and space in view of sensor node, for the measured value x of certain sensor node in a networki, its fault diagnosis can compare differentiation with the perception value of neighbors around.For certain, there is M neighbors and treat diagnosis node.Its weighted median may be defined as:
Wherein,For weighted median, xj(j=1,2...M) is the perception measured value of neighbors, λj(j=1,2...M) it is the weights of each neighbors.The Distinguishing diagnosis of node soft fault (abnormality sensing measured value) can be realized by following fault diagnosis function.
But owing to, in monitoring structural health conditions, node can gather substantial amounts of original sample every time, above-mentioned weighted median fault distinguishing cannot adapt to substantial amounts of data exchange between node.The resolving ideas intending using is to using compression sampling based on compressed sensing during node sample, to adapt to the exchange of internodal data.For the measured value of sensor node, use one and orthogonal basis Ψ ∈ RN × NIncoherent matrix Φ ∈ RM × N(M ＜ ＜ N), projects on a lower dimensional space, it is achieved the compression of node sample signal by high dimensional signal.Its core formula is:
Y=Φ x=Φ Ψ α=Θ α (4)
Wherein, Φ is M × N matrix, referred to as calculation matrix；Θ=Φ Ψ is the matrix of M × N, referred to as observing matrix.Finally by solving lpThe method of norm minimum, it is achieved the reconstruct of compressed signal.
(4): the hard fault diagnosis mechanism of wireless sensor node and the design of selfreparing node in SHM
1) self-repairing wireless sensing device node hardware framework
Research is intended the self-repairing wireless sensing device node hardware framework of employing as shown in Figure 4.
For the characteristic of multi-parameters sampling in monitoring structural health conditions, for without loss of generality, we intend studying the typical self-repair function straining wireless sensor node, and the design principle of other kinds of sensor node is identical.In view of the application demand of the Practical Project of monitoring structural health conditions, the node of self-repairing wireless sensing device must is fulfilled for feature low in energy consumption, that low cost, volume are little.Can be selected for lower-cost programming device as Embedded solution, main functional modules changes and was mutually connected in the past, is connected as module cell by bionic hardware FPAAs.Strain wireless sensor node shown in Fig. 4 includes three parts: sensing module, signal processing module, radio receiving transmitting module.Using bionic hardware, field programmable analog array FPAAs realizes the signal link of the sensing module in node, and designs fault diagnosis functions and the selfreparing control function of Inductive links in the signal processing module of node.Wherein, the redundancy section in self-repair module both can use the hardware circuit identical with former sensing module to realize, it would however also be possible to employ the functional unit that programmable array is constituted realizes.
2) self-repairing wireless sensing device node software framework
Research is intended the self-repairing wireless sensing device node software part framework of employing as shown in Figure 5.It is compared provide diagnostic result with the threshold value set by abnormal signal diagnotor after binary data in A/D translation register is converted into decimal data by data acquisition driver；Master control programme sends drive command according to diagnostic result to FPAA driver；First this program reads the FPAA configuration file resided in outside flash storage according to drive command, then configures FPAA dynamically so that it is complete sensing and the reconstruct of redundant layer signal link thereof.
3) performance of self-repairing wireless sensing network considers
Self-repairing wireless sensing device network struture system is intended using mixing-classifying topological structure, in addition to original sensor node, it is considered to arrange the monitor node being used for monitoring network state in a network.The systematic parameter considered is needed to include the many kinds of parameters such as network function, reliability, power consumption, repair time during reparation.
Claims (1)
1. the diagnosis of the soft or hard fault being suitable to health monitoring and self-repair method, it is characterised in that bag
Include following steps:
1) it route during efficient real in SHM: based on the collaboration communication mechanism of multiple stream transmission, design point
Cloth Clustering Routing, forms routing link, and the first node of link is a bunch head, and secondary nodal point is spare cluster
First-class；
2) in SHM, the interference of abnormal data processes: the feature extracting method of integrated structure monitoring vibration signal,
Go out to be respectively used to node failure detection and two kinds of features of Damage Assessment Method from acquired original extracting data
Parameter, after the feature extraction obtaining natural frequency, uses sample statistics algorithm, or conventional node cooperation event
Barrier diagnosis algorithm, it is achieved node failure exceptional value and the differentiation of structural damage measured value and extraction；
3) diagnosis mechanism in SHM: in monitoring structural health conditions, uses during node sample based on compression sense
The compression sampling known, to adapt to the exchange of internodal data；
4) Hardware-software architecture: in self-repairing wireless sensing device node hardware framework, selects Programmable
Part, as Embedded solution, is connected as module by crossing bionic hardware FPAAs；Reviewing one's lessons by oneself
In multiple wireless sensor node software architecture, first program reads according to drive command and resides on outside Flash
FPAA configuration file in memorizer, then configures FPAA dynamically so that it is complete sensing and
The reconstruct of redundant layer signal link.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610405559.1A CN105873111A (en) | 2016-06-08 | 2016-06-08 | Soft and hard fault diagnosis and self restoration method suitable for health monitoring |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610405559.1A CN105873111A (en) | 2016-06-08 | 2016-06-08 | Soft and hard fault diagnosis and self restoration method suitable for health monitoring |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105873111A true CN105873111A (en) | 2016-08-17 |
Family
ID=56677411
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610405559.1A Pending CN105873111A (en) | 2016-06-08 | 2016-06-08 | Soft and hard fault diagnosis and self restoration method suitable for health monitoring |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105873111A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108173620A (en) * | 2016-12-08 | 2018-06-15 | 南京海道普数据技术有限公司 | WSN abnormal datas based on compression network coding find network system realization |
| CN109057863A (en) * | 2018-07-31 | 2018-12-21 | 郑州智谷工业技术有限公司 | A kind of underground life detecting device and method based on Evolvable Hardware |
| CN110266527A (en) * | 2019-06-11 | 2019-09-20 | 同济大学 | Sensor node failure modes alarm method and device based on spatial coherence |
| CN116627953A (en) * | 2023-05-24 | 2023-08-22 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714190A (en) * | 2009-11-20 | 2010-05-26 | 重庆大学 | Role diagnostic system for bridge structural state |
| CN103974367A (en) * | 2014-05-21 | 2014-08-06 | 哈尔滨工程大学 | Error-tolerance and multi-path optimization method based on HEED algorithm in wireless sensor network |
-
2016
- 2016-06-08 CN CN201610405559.1A patent/CN105873111A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714190A (en) * | 2009-11-20 | 2010-05-26 | 重庆大学 | Role diagnostic system for bridge structural state |
| CN103974367A (en) * | 2014-05-21 | 2014-08-06 | 哈尔滨工程大学 | Error-tolerance and multi-path optimization method based on HEED algorithm in wireless sensor network |
Non-Patent Citations (2)
| Title |
|---|
| M.A.-B.ABDO等：: ""A numerical study of structural damage detection using changes in the rotation of mode shapes"", 《JOURNAL OF SOUND AND VIBRATION》 * |
| 季赛等：: ""面向无线传感结构健康监测的压缩感知方法研究"", 《传感技术学报》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108173620A (en) * | 2016-12-08 | 2018-06-15 | 南京海道普数据技术有限公司 | WSN abnormal datas based on compression network coding find network system realization |
| CN108173620B (en) * | 2016-12-08 | 2020-12-29 | 南京海道普数据技术有限公司 | Implementation method of WSN abnormal data discovery system based on compressed network coding |
| CN109057863A (en) * | 2018-07-31 | 2018-12-21 | 郑州智谷工业技术有限公司 | A kind of underground life detecting device and method based on Evolvable Hardware |
| CN110266527A (en) * | 2019-06-11 | 2019-09-20 | 同济大学 | Sensor node failure modes alarm method and device based on spatial coherence |
| CN116627953A (en) * | 2023-05-24 | 2023-08-22 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
| CN116627953B (en) * | 2023-05-24 | 2023-10-27 | 首都师范大学 | Method for repairing loss of groundwater level monitoring data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Mariniello et al. | Structural damage detection and localization using decision tree ensemble and vibration data | |
| Fu et al. | Sensor fault management techniques for wireless smart sensor networks in structural health monitoring | |
| CN105515184B (en) | Multisensor many reference amounts distribution synergic monitoring system based on wireless sensor network | |
| CN105873111A (en) | Soft and hard fault diagnosis and self restoration method suitable for health monitoring | |
| CN109872003B (en) | Object state prediction method, object state prediction system, computer device, and storage medium | |
| CN103246265B (en) | Electromechanical equipment detection maintaining method | |
| CN102749585A (en) | Multi-layer SVM (support vector machine) based storage battery on-line monitoring method | |
| CN108491860A (en) | Shale gas pressure break equips Life cycle reliability in time dynamic assessment method and device | |
| CN109640335B (en) | Wireless sensor fault diagnosis method based on convolutional neural network | |
| Yu et al. | Node fault diagnosis in WSN based on RS and SVM | |
| CN106056273A (en) | Redundant electric pump body failure reliability monitoring method based on fault tree | |
| Bhuiyan et al. | Towards cyber-physical systems design for structural health monitoring: Hurdles and opportunities | |
| CN108199899A (en) | A kind of wireless sensor network fault detection method, apparatus and system | |
| CN102033789A (en) | Reliability analysis method for embedded safety-critical system | |
| CN104880217A (en) | Fault sensor information reconstruction method based on measured value association degree | |
| Risodkar et al. | A survey: Structural health monitoring of bridge using WSN | |
| Tahat et al. | An evaluation of machine learning algorithms in an experimental structural health monitoring system incorporating lora iot connectivity | |
| US20140032079A1 (en) | Diagnostic and control system and method | |
| Zhang et al. | Sensor fault diagnosis and location for small and medium-scale wireless sensor networks | |
| CN110427689B (en) | Method for monitoring and diagnosing gas turbine unit group based on new information technology | |
| Arjannikov et al. | Using markov chains to model sensor network reliability | |
| Lo et al. | Reference-free detection of spike faults in wireless sensor networks | |
| CN103235206A (en) | Transformer fault diagnosis method | |
| Wang et al. | A PHM architecture of maglev train based on the distributed hierarchical structure | |
| CN115936663A (en) | Maintenance method and device for power system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160817 |
|
| WD01 | Invention patent application deemed withdrawn after publication |