CN116506469B - Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning - Google Patents
Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning Download PDFInfo
- Publication number
- CN116506469B CN116506469B CN202310745111.4A CN202310745111A CN116506469B CN 116506469 B CN116506469 B CN 116506469B CN 202310745111 A CN202310745111 A CN 202310745111A CN 116506469 B CN116506469 B CN 116506469B
- Authority
- CN
- China
- Prior art keywords
- federal
- training
- model parameters
- rfid sensor
- bridge
- 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
- 238000012544 monitoring process Methods 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000012549 training Methods 0.000 claims abstract description 71
- 230000006870 function Effects 0.000 claims abstract description 18
- 239000003990 capacitor Substances 0.000 claims description 36
- 238000012806 monitoring device Methods 0.000 claims description 8
- 230000004927 fusion Effects 0.000 claims description 7
- 239000003381 stabilizer Substances 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 2
- 238000009826 distribution Methods 0.000 claims description 2
- 230000006641 stabilisation Effects 0.000 claims description 2
- 238000011105 stabilization Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 5
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 229910052744 lithium Inorganic materials 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000919 ceramic Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003862 health status Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000011150 reinforced concrete Substances 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02N—ELECTRIC MACHINES NOT OTHERWISE PROVIDED FOR
- H02N2/00—Electric machines in general using piezoelectric effect, electrostriction or magnetostriction
- H02N2/18—Electric machines in general using piezoelectric effect, electrostriction or magnetostriction producing electrical output from mechanical input, e.g. generators
- H02N2/186—Vibration harvesters
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01H—MEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
- G01H17/00—Measuring mechanical vibrations or ultrasonic, sonic or infrasonic waves, not provided for in the preceding groups
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
- G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
- G06K7/10009—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation sensing by radiation using wavelengths larger than 0.1 mm, e.g. radio-waves or microwaves
- G06K7/10297—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation sensing by radiation using wavelengths larger than 0.1 mm, e.g. radio-waves or microwaves arrangements for handling protocols designed for non-contact record carriers such as RFIDs NFCs, e.g. ISO/IEC 14443 and 18092
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q9/00—Arrangements in telecontrol or telemetry systems for selectively calling a substation from a main station, in which substation desired apparatus is selected for applying a control signal thereto or for obtaining measured values therefrom
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2209/00—Arrangements in telecontrol or telemetry systems
- H04Q2209/70—Arrangements in the main station, i.e. central controller
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q2209/00—Arrangements in telecontrol or telemetry systems
- H04Q2209/80—Arrangements in the sub-station, i.e. sensing device
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- Toxicology (AREA)
- Software Systems (AREA)
- Computer Security & Cryptography (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Electromagnetism (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention provides a bridge state monitoring method based on vibration energy taking RFID sensors and federation learning, wherein a central cloud server pushes initialized global model parameters to each RFID sensor participating in federation training, the RFID sensors train according to bridge vibration data acquired locally to obtain respective local model parameters, then the central cloud server fuses and updates the local model parameters according to a federation average algorithm to obtain updated global model parameters and pushes the updated global model parameters again until the local models of all the RFID sensors reach the minimum of loss functions to obtain final global model parameters, then the RFID sensors train the final global model parameters by utilizing local personalized data, data processing pressure of the central cloud server can be reduced, monitoring pertinence is improved, the generalization capability is better, finally, the bridge state is monitored through a global personalized model, and the accuracy of bridge state monitoring is effectively improved.
Description
Technical Field
The invention relates to the technical field of highway facility detection, in particular to a bridge state monitoring method based on a vibration energy-taking RFID sensor and federal learning.
Background
The bridge is one of the vital links in the modern transportation system, the health status of the bridge plays a vital role in the social production and life, most of the bridges are currently made of reinforced concrete, the loss and damage conditions are inevitably caused in the frequent passing of vehicles, the real-time status monitoring of the bridge is beneficial to strengthening the traffic safety, and the stability of the social production and life is guaranteed.
The existing bridge state monitoring methods mainly depend on manual appearance inspection and regular field test, and the methods belong to traditional detection methods, are monitoring methods for post analysis, cannot identify potential faults in time, and cannot meet the development trend of current state maintenance. At present, although some means for digitally monitoring bridge states exist, the problems of high data processing pressure, poor monitoring pertinence and low monitoring accuracy of a cloud server exist.
Disclosure of Invention
The invention aims to provide a bridge state monitoring method based on a vibration energy-taking RFID sensor and federal learning, which aims to solve the problems of high data processing pressure, poor monitoring pertinence and low monitoring accuracy of a cloud server in the prior art.
The bridge state monitoring device comprises a piezoelectric energy collector, an energy management circuit and an RFID sensor which are electrically connected in sequence, wherein the piezoelectric energy collector is used for collecting bridge vibration energy and converting the bridge vibration energy into unstable direct current, the energy management circuit is used for stabilizing the unstable direct current to obtain stable direct current, and the stable direct current is used for supplying power to the RFID sensor;
the method comprises the following steps:
the central cloud server randomly generates an initialized global model parameterAnd will->Pushing the RFID sensors to each RFID sensor participating in federal training;
each RFID sensor participating in federal training receivesAfter that, training +_ according to the respective locally acquired bridge vibration data>Obtaining respective local model parameters, synchronously training all RFID sensors participating in federal training, uploading the respective local model parameters to a central cloud server, and carrying out fusion updating on all the local model parameters by the central cloud server according to a federal average algorithm to obtain updated global model parameters->；
The central cloud server updates the global model parametersAnd issuing the local model to all the RFID sensors participating in the federal training again to perform the next round of local training, and iterating the training until the local model of all the RFID sensors participating in the federal training reaches the minimum loss function, wherein for the r-th RFID sensor participating in the federal training, the conditional expression is satisfied:
wherein ,local model parameter representing the r-th RFID sensor involved in federal training, +.>Loss function representing the local model of the r-th RFID sensor involved in federal training, +.>Representing learning rate, b representing the number of samples in a batch of training samples, +.>Indicating a gradient drop;
after the local models of all the RFID sensors participating in the federal training reach the minimum loss function, uploading the local model parameters obtained by the last iteration training to a central cloud server by each RFID sensor participating in the federal training, and carrying out fusion updating by the central cloud server according to a federal average algorithm to obtain final global model parameters;
the central cloud server transmits the final global model parameters to all RFID sensors participating in federal training;
each RFID sensor participating in federal training trains the final global model parameters by utilizing local personalized data to obtain a local personalized output model of each RFID sensor participating in federal training;
splicing the hidden layer corresponding to the final global model parameters and the output layer of the local personalized output model of each RFID sensor participating in federal training to obtain a global personalized model;
and monitoring the bridge state through the global personalized model to obtain a monitoring result.
According to the bridge state monitoring method based on the vibration energy taking RFID sensor and the federal learning, the central cloud server pushes the initialized global model parameters to each RFID sensor participating in federal training, the RFID sensors train according to the locally acquired bridge vibration data to obtain the local model parameters, then the central cloud server fuses and updates the locally acquired global model parameters according to the federal average algorithm, the locally acquired global model parameters are pushed again until the local models of all the RFID sensors reach the minimum loss function, then the central cloud server obtains the final global model parameters, then the RFID sensors train the final global model parameters by utilizing the local personalized data to obtain the local personalized output models of the RFID sensors participating in federal training, the data processing pressure of the central cloud server is reduced, the monitoring pertinence is improved, the generalization capability is better, finally the hidden layers corresponding to the final global model parameters and the output layers of the locally acquired local output models of the RFID sensors participating in federal training are spliced to obtain the global model parameters, the personalized global model parameters are obtained, the global model is used for monitoring the bridge state is effectively monitored by the global model, and the bridge state monitoring state is greatly applied to the bridge state.
In addition, the piezoelectric energy collector and the energy management circuit are adopted to convert vibration energy generated by the vehicle passing through the bridge into electric energy, so that the defect that the battery needs to be replaced frequently when the conventional RFID sensor uses the lithium battery is avoided.
Drawings
FIG. 1 is a block diagram of a bridge condition monitoring device;
fig. 2 is a schematic circuit diagram of an energy management circuit.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The embodiment of the invention provides a bridge state monitoring method based on a vibration energy-taking RFID sensor and federal learning, which is applied to a bridge state monitoring device.
Referring to fig. 1 to 2, the bridge status monitoring device includes a piezoelectric energy collector 10, an energy management circuit 20 and an RFID sensor 30 electrically connected in sequence, where the piezoelectric energy collector 10 is configured to collect bridge vibration energy and convert the bridge vibration energy into unstable direct current, and the piezoelectric energy collector 10 may use piezoelectric ceramics. The energy management circuit 20 is configured to perform voltage stabilization on the unstable dc power to obtain a stable dc power, and power the RFID sensor 30 through the stable dc power.
The energy management circuit 20 includes a full wave bridge rectifier circuit, an under-voltage latch circuit, a super capacitor C4 and a low-voltage-drop regulator, a second resistor R2, a first capacitor C1 and a fifth capacitor C5.
The piezoelectric energy collector 10, the full-wave bridge rectifier circuit, the under-voltage locking circuit, the super capacitor C4, the low-voltage-drop voltage stabilizer and the RFID sensor 30 are electrically connected in sequence.
The full-wave bridge rectifier circuit includes first diode D1, second diode D2, third diode D3, fourth diode D4 that head links to each other in proper order, the one end of piezoelectric energy harvester 10 is connected first diode D1 with between the second diode D2, the other end of piezoelectric energy harvester 10 is connected third diode D3 with between the fourth diode D4.
The under-voltage locking circuit comprises a TPS62120 chip, a first resistor R1, a first inductor L1, a second capacitor C2 and a third capacitor C3, wherein the VIN pin of the TPS62120 chip is connected between the second diode D2 and the fourth diode D4.
The low-voltage drop voltage stabilizer adopts an LT3009 chip, and an IN pin of the LT3009 chip is electrically connected with an VOUT pin of the TPS62120 chip.
One end of the first resistor R1 is grounded, the other end of the first resistor R1 is electrically connected with the FB pin of the TPS62120 chip, one end of the first inductor L1 is electrically connected with the SW pin of the TPS62120 chip, the other end of the first inductor L1 is electrically connected with the VOUT pin of the TPS62120 chip and the IN pin and SHDN pin of the LT3009 chip, one end of the second capacitor C2 is electrically connected with the VIN pin of the TPS62120 chip, the other end of the second capacitor C2 is grounded, one end of the third capacitor C3 is electrically connected with the SGND pin and VOUT pin of the TPS62120 chip, and the other end of the third capacitor C3 is electrically connected with the FB pin of the TPS62120 chip and the first resistor R1.
One end of the super capacitor C4 is electrically connected with the IN pin of the LT3009 chip and the VOUT pin of the TPS62120 chip respectively, and the other end of the super capacitor C4 is grounded.
The second resistor R2 and the first capacitor C1 are connected in parallel between the full-wave bridge rectifier circuit and the VIN pin of the TPS62120 chip.
One end of the fifth capacitor C5 is electrically connected to the OUT pin of the LT3009 chip and the RFID sensor 30, and the other end of the fifth capacitor C5 is grounded.
The under-voltage locking circuit is used for controlling charge and discharge of the super capacitor C4, the second capacitor C2, the third capacitor C3, the first resistor R1 and the first inductor L1 are used for providing working voltage and pin input signals for the TPS62120 chip, and the super capacitor C4 is used for storing energy. The electric energy obtained by the piezoelectric energy collector 10 is rectified by a full-wave bridge rectifier circuit, and then is output to the RFID sensor 30 through an under-voltage latch circuit and a low-voltage-drop voltage stabilizer LT3009 chip to stabilize the direct-current voltage. The second resistor R2 and the first capacitor C1 are used for reducing fluctuation of the output voltage of the full-wave bridge rectifier circuit so as to meet the input requirement of the under-voltage latch-up circuit. The fifth capacitor C5 is connected to the RFID sensor 30 at both ends for powering the RFID sensor 30.
Based on the bridge state monitoring device, the bridge state monitoring method based on the vibration energy-taking RFID sensor and federal learning comprises the following steps of S101-S108:
s101, the central cloud server randomly generates an initialized global model parameterAnd will->Push to each RFID sensor involved in federal training.
S102, each RFID sensor participating in federal training receivesAfter that, training +_ according to the respective locally acquired bridge vibration data>Obtaining respective local model parameters, synchronously training all RFID sensors participating in federal training, uploading the respective local model parameters to a central cloud server, and carrying out fusion updating on all the local model parameters by the central cloud server according to a federal average algorithm to obtain updated global model parameters->。
Wherein the updated global model parametersThe expression of (2) is:
where R represents the total number of RFID sensors involved in federal training.
S103, the central cloud server updates the updated global model parametersAnd transmitting the information to all the RFID sensors participating in the federal training again for carrying out the next round of local training, and iterating the training until the local model of all the RFID sensors participating in the federal trainingA loss function minimization is achieved, wherein for the r-th RFID sensor participating in federal training, the conditional expression is satisfied:
wherein ,local model parameter representing the r-th RFID sensor involved in federal training, +.>Loss function representing the local model of the r-th RFID sensor involved in federal training, +.>Representing learning rate, b representing the number of samples in a batch of training samples, +.>Indicating a gradient drop.
In this embodiment, the loss function of the local model of the r-th RFID sensor involved in federal trainingExpressed in terms of L2 norms, the specific expression is as follows:
wherein Yt represents the bridge state true value,representing bridge status values calculated by the global model.
And S104, after the local models of all the RFID sensors participating in the federal training reach the minimum loss function, uploading the local model parameters obtained by the last iteration training to a central cloud server by each RFID sensor participating in the federal training, and carrying out fusion updating by the central cloud server according to a federal average algorithm to obtain final global model parameters.
And S105, the central cloud server transmits the final global model parameters to all RFID sensors participating in federal training.
S106, each RFID sensor participating in the federal training trains the final global model parameters by using local personalized data to obtain a local personalized output model of each RFID sensor participating in the federal training.
The local personalized data at least comprises business district distribution data, workday data, holiday data and weather data.
Step S106 satisfies the following conditional expression:
wherein ,output value representing global personalization model, +.>Representing Sigmoid activation function,/->Representing hidden layers corresponding to final global model parameters,/->An output layer representing a locally personalized output model of each of the federally trained RFID sensors.
And S107, splicing the hidden layer corresponding to the final global model parameters and the output layer of the local personalized output model of each RFID sensor participating in federal training to obtain a global personalized model.
Wherein the global personalization model satisfies the following conditional expression:
wherein ,a set of parameters representing a global personalization model that minimizes the loss function.
S108, monitoring the bridge state through the global personalized model to obtain a monitoring result.
In summary, according to the bridge state monitoring method based on vibration energy-taking RFID sensors and federation learning provided by the invention, after the central cloud server pushes initialized global model parameters to each RFID sensor participating in federation training, the RFID sensors train according to respective locally acquired bridge vibration data to obtain respective local model parameters, then the central cloud server fuses and updates the respective local model parameters according to a federation average algorithm to obtain updated global model parameters and pushes the updated global model parameters again until the local models of all the RFID sensors reach the minimum of loss functions, then the central cloud server obtains final global model parameters, then the RFID sensors train the final global model parameters by utilizing local personalized data to obtain local personalized output models of each RFID sensor participating in federation training, the data processing pressure of the central cloud server is reduced, the monitoring pertinence is improved, finally the hidden layers corresponding to the final global model parameters and the output layers of the local output models of each RFID sensor participating in federation training are subjected to global splicing operation, the global personalized bridge is obtained, the global personalized bridge state is effectively monitored by the global state monitoring method is improved, and the bridge state monitoring method is suitable for the bridge state monitoring field.
In addition, the piezoelectric energy collector and the energy management circuit are adopted to convert vibration energy generated by the vehicle passing through the bridge into electric energy, so that the defect that the battery needs to be replaced frequently when the conventional RFID sensor uses the lithium battery is avoided.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
While embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that: many changes, modifications, substitutions and variations may be made to the embodiments without departing from the spirit and principles of the invention, the scope of which is defined by the claims and their equivalents.
Claims (9)
1. The bridge state monitoring method based on the vibration energy-taking RFID sensor and the federal learning is characterized by being applied to a bridge state monitoring device, wherein the bridge state monitoring device comprises a piezoelectric energy collector, an energy management circuit and an RFID sensor which are electrically connected in sequence, the piezoelectric energy collector is used for collecting bridge vibration energy and converting the bridge vibration energy into unstable direct current, the energy management circuit is used for conducting voltage stabilization treatment on the unstable direct current to obtain stable direct current, and the stable direct current is used for supplying power to the RFID sensor;
the method comprises the following steps:
the central cloud server randomly generates an initialized global model parameterAnd will->Pushing the RFID sensors to each RFID sensor participating in federal training;
each RFID sensor participating in federal training receivesAfter that, training according to the locally acquired bridge vibration dataObtaining respective local model parameters, synchronously training all RFID sensors participating in federal training, uploading the respective local model parameters to a central cloud server, and carrying out fusion updating on all the local model parameters by the central cloud server according to a federal average algorithm to obtain updated global model parameters->；
The central cloud server updates the global model parametersAnd issuing the local model to all the RFID sensors participating in the federal training again to perform the next round of local training, and iterating the training until the local model of all the RFID sensors participating in the federal training reaches the minimum loss function, wherein for the r-th RFID sensor participating in the federal training, the conditional expression is satisfied:
wherein ,local model parameter representing the r-th RFID sensor involved in federal training, +.>Loss function representing the local model of the r-th RFID sensor involved in federal training, +.>Representing learning rate, b representing the number of samples in a batch of training samples, +.>Indicating a gradient drop;
after the local models of all the RFID sensors participating in the federal training reach the minimum loss function, uploading the local model parameters obtained by the last iteration training to a central cloud server by each RFID sensor participating in the federal training, and carrying out fusion updating by the central cloud server according to a federal average algorithm to obtain final global model parameters;
the central cloud server transmits the final global model parameters to all RFID sensors participating in federal training;
each RFID sensor participating in federal training trains the final global model parameters by utilizing local personalized data to obtain a local personalized output model of each RFID sensor participating in federal training;
splicing the hidden layer corresponding to the final global model parameters and the output layer of the local personalized output model of each RFID sensor participating in federal training to obtain a global personalized model;
and monitoring the bridge state through the global personalized model to obtain a monitoring result.
2. The bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning according to claim 1, wherein the central cloud server performs fusion update on all local model parameters according to federal average algorithm to obtain updated global model parametersIn the step (a), updated global model parameters +.>The expression of (2) is:
where R represents the total number of RFID sensors involved in federal training.
3. The bridge state monitoring method based on vibration energy taking RFID sensor and federal learning according to claim 1, wherein the loss function of the local model of the r-th RFID sensor participating in federal trainingExpressed in terms of L2 norms, the specific expression is as follows:
wherein Yt represents the bridge state true value,representing bridge status values calculated by the global model.
4. The bridge status monitoring method based on vibration energy-taking RFID sensors and federal learning according to claim 3, wherein in the step of performing a splicing operation on an hidden layer corresponding to the final global model parameter and an output layer of a local personalized output model of each RFID sensor participating in federal training to obtain a global personalized model, the following conditional expression is satisfied:
wherein ,output value representing global personalization model, +.>Representing Sigmoid activation function,/->Representing hidden layers corresponding to final global model parameters,/->An output layer representing a locally personalized output model of each of the federally trained RFID sensors.
5. The bridge state monitoring method based on vibration energy-taking RFID sensors and federal learning according to claim 4, wherein the global personalized model satisfies the following conditional expression:
wherein ,parameters of the global personalization model that minimize the loss function are represented.
6. The bridge status monitoring method based on vibration energy-taking RFID sensors and federal learning according to claim 1, wherein the local personalized data at least includes business district distribution data, workday data, holiday data, weather data.
7. The bridge state monitoring method based on the vibration energy-taking RFID sensor and the federal learning according to claim 1, wherein the energy management circuit comprises a full-wave bridge rectifier circuit, an under-voltage lockout circuit, a super capacitor and a low-voltage drop voltage stabilizer;
the piezoelectric energy collector, the full-wave bridge rectifier circuit, the under-voltage locking circuit, the super capacitor, the low-voltage-drop voltage stabilizer and the RFID sensor are electrically connected in sequence.
8. The bridge state monitoring method based on the vibration energy taking RFID sensor and the federal learning according to claim 7, wherein the full-wave bridge rectifier circuit comprises a first diode, a second diode, a third diode and a fourth diode which are sequentially connected in an end-to-end mode, one end of the piezoelectric energy collector is connected between the first diode and the second diode, and the other end of the piezoelectric energy collector is connected between the third diode and the fourth diode;
the under-voltage locking circuit comprises a TPS62120 chip, a first resistor, a first inductor, a second capacitor and a third capacitor, wherein the VIN pin of the TPS62120 chip is connected between the second diode and the fourth diode;
the low-voltage drop voltage stabilizer adopts an LT3009 chip, and an IN pin of the LT3009 chip is electrically connected with an VOUT pin of the TPS62120 chip;
one end of the first resistor is grounded, the other end of the first resistor is electrically connected with the FB pin of the TPS62120 chip, one end of the first inductor is electrically connected with the SW pin of the TPS62120 chip, the other end of the first inductor is electrically connected with the VOUT pin of the TPS62120 chip and the IN pin and the SHDN pin of the LT3009 chip respectively, one end of the second capacitor is electrically connected with the VIN pin of the TPS62120 chip, the other end of the second capacitor is grounded, one end of the third capacitor is electrically connected with the SGND pin and the VOUT pin of the TPS62120 chip respectively, and the other end of the third capacitor is electrically connected with the FB pin and the first resistor of the TPS62120 chip respectively;
one end of the super capacitor is electrically connected with the IN pin of the LT3009 chip and the VOUT pin of the TPS62120 chip respectively, and the other end of the super capacitor is grounded.
9. The bridge status monitoring method based on vibration energy-taking RFID sensor and federal learning of claim 8, wherein the bridge status monitoring device further comprises a second resistor, a first capacitor, and a fifth capacitor;
the second resistor and the first capacitor are connected in parallel between the full-wave bridge rectifier circuit and the VIN pin of the TPS62120 chip;
one end of the fifth capacitor is electrically connected with the OUT pin of the LT3009 chip and the RFID sensor respectively, and the other end of the fifth capacitor is grounded.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310745111.4A CN116506469B (en) | 2023-06-25 | 2023-06-25 | Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310745111.4A CN116506469B (en) | 2023-06-25 | 2023-06-25 | Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116506469A CN116506469A (en) | 2023-07-28 |
| CN116506469B true CN116506469B (en) | 2023-09-05 |
Family
ID=87325015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310745111.4A Active CN116506469B (en) | 2023-06-25 | 2023-06-25 | Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116506469B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112488398A (en) * | 2020-12-03 | 2021-03-12 | 广东电力通信科技有限公司 | Electricity utilization management method and system based on MEC edge intelligent gateway |
| CN114530028A (en) * | 2022-02-14 | 2022-05-24 | 大连理工大学 | Campus student intelligent bracelet monitoring system and method based on LoRa communication and federal learning |
| CN115376031A (en) * | 2022-10-24 | 2022-11-22 | 江西省科学院能源研究所 | Road unmanned aerial vehicle routing inspection data processing method based on federal adaptive learning |
| WO2023029944A1 (en) * | 2021-08-30 | 2023-03-09 | 华为云计算技术有限公司 | Federated learning method and device |
-
2023
- 2023-06-25 CN CN202310745111.4A patent/CN116506469B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112488398A (en) * | 2020-12-03 | 2021-03-12 | 广东电力通信科技有限公司 | Electricity utilization management method and system based on MEC edge intelligent gateway |
| WO2023029944A1 (en) * | 2021-08-30 | 2023-03-09 | 华为云计算技术有限公司 | Federated learning method and device |
| CN114530028A (en) * | 2022-02-14 | 2022-05-24 | 大连理工大学 | Campus student intelligent bracelet monitoring system and method based on LoRa communication and federal learning |
| CN115376031A (en) * | 2022-10-24 | 2022-11-22 | 江西省科学院能源研究所 | Road unmanned aerial vehicle routing inspection data processing method based on federal adaptive learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116506469A (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111537938B (en) | Error short-time prediction method for electronic transformer based on intelligent algorithm | |
| US20100295673A1 (en) | Cloud computing for monitoring an above-ground oil production facility | |
| CN107730117B (en) | Cable maintenance early warning method and system based on heterogeneous data comprehensive analysis | |
| DE102020117609B4 (en) | Processing of status data of a battery for aging estimation | |
| CN113049035B (en) | Transformer state monitoring system based on Internet of things | |
| CN116826933B (en) | Knowledge-graph-based hybrid energy storage battery power supply backstepping control method and system | |
| CN105891698B (en) | A kind of Boost circuit multi-parameter on-line identification method | |
| CN116506469B (en) | Bridge state monitoring method based on vibration energy-taking RFID sensor and federal learning | |
| CN106651658A (en) | Non-intruding type dwelling electrical load decomposition method based on finite-state machine | |
| CN111555776A (en) | Fusion sensing and joint diagnosis method, system and device for power transmission line | |
| CN109782051B (en) | Load current abnormity early warning device, system and method based on Internet of things | |
| CN113379005B (en) | Intelligent energy management system and method for power grid power equipment | |
| CN107295553A (en) | A kind of base station maintenance method and device | |
| CN116664095B (en) | Security inspection prediction method based on deep learning | |
| CN105811398A (en) | Monitoring data-based determination method and system of maximum load coincidence factor | |
| CN115603464B (en) | Intelligent generation management system for power grid operation ticket based on digital twin | |
| JP4181689B2 (en) | Remote automatic diagnosis system for power consumption | |
| CN204098865U (en) | A kind of intelligent monitoring oil exploration equipment system | |
| CN116739550A (en) | Intelligent auxiliary decision-making method and system for rush repair | |
| CN114977162A (en) | Charging monitoring method and system for electric automobile | |
| CN206181419U (en) | Lamp controller | |
| CN106936216A (en) | A kind of panorama monitors analysis process system on-line | |
| CN204481696U (en) | A kind of power supply circuits of power collection systems | |
| CN111478340B (en) | Distribution network line reactive compensation condition analysis method | |
| CN113690998A (en) | Multi-mode energy management system and method |
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 |