CN115146230A

CN115146230A - Ancient building health monitoring system, method and equipment

Info

Publication number: CN115146230A
Application number: CN202210796382.8A
Authority: CN
Inventors: 冯超; 王自法; 吴禄源; 位栋梁; 苗鹏宇; 王祥祺; 候笑焜; 高曹珀; 廖吉安; 党浩天
Original assignee: Henan University
Current assignee: Henan University
Priority date: 2022-07-06
Filing date: 2022-07-06
Publication date: 2022-10-04

Abstract

The invention provides a system, a method and equipment for monitoring the health of an ancient building. The system comprises a sensing subsystem, a monitoring subsystem and a monitoring subsystem, wherein the sensing subsystem acquires environmental monitoring data, component monitoring data and structure monitoring data of a target historic building through a sensor; the data acquisition and transmission subsystem acquires the monitoring data acquired by the sensing subsystem and performs preprocessing; the data storage and management subsystem receives the preprocessed monitoring data, performs attribute storage according to different attributes of the monitoring data, and performs secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters; and the damage positioning and early warning subsystem identifies the structural damage by using the CNN convolutional neural network and outputs a health monitoring result. In this way, the damage of the structure can be positioned by using the structural damage identification method of the CNN convolutional neural network through the collected various monitoring information, and the early warning function is realized according to the damage early warning standard, so that the health condition of the ancient building is comprehensively, comprehensively and accurately evaluated.

Description

Ancient building health monitoring system, method and equipment

Technical Field

The present invention relates generally to the field of construction engineering and, more particularly, to an ancient building health monitoring system, method and apparatus.

Background

The ancient architecture is a crystal of the architectural civilization and a carrier of the historical information of China, and the ancient architecture is a precious cultural heritage and nonrenewable cultural resource. Due to the invasion of wind, frost, rain and dew, natural disasters and the influence of environmental factors, a series of problems such as corrosion, weathering, cracks, inclination and the like can occur to the ancient architecture, and in addition, due to human factors, the ancient architecture can be seriously damaged, so that the ancient architecture is necessary to be protected and monitored by using the modern information technology.

However, the traditional historic building health monitoring mainly adopts manual detection as a main part, and because the manual detection has strong subjectivity and low acquisition frequency, a large amount of manpower and material resources are required to be invested to acquire data on site, and higher requirements are provided for the capability of workers. And usually, only local detection can be carried out, the positions of diseases and structural damages of the historic building can not be found in time, the collected data is poor in continuity and visualization effect, the data is complex and complicated, the analysis is difficult and serious, and real-time on-line monitoring and structural damage early warning can not be carried out, so that the further development of the historic building health monitoring is restricted. In the face of analysis and calculation of massive monitoring data information with various complex forms, the problems of low efficiency, long period and low accuracy of the traditional manual analysis method are particularly highlighted.

Disclosure of Invention

According to an embodiment of the invention, an ancient building health monitoring scheme is provided. According to the scheme, the structure is damaged and positioned by using various collected monitoring information and a structural damage identification method of the CNN convolutional neural network, and the early warning function is realized according to the damage early warning standard, so that the health condition of the ancient building is comprehensively, comprehensively and accurately evaluated.

In a first aspect of the invention, an ancient building health monitoring system is provided. The system comprises:

the sensing subsystem is used for acquiring environment monitoring data, component monitoring data and structure monitoring data of the target historic building through a sensor;

the data acquisition and transmission subsystem is used for acquiring the monitoring data acquired by the sensing subsystem, preprocessing the monitoring data and transmitting the preprocessed monitoring data to the data storage and management subsystem;

the data storage and management subsystem is used for receiving the preprocessed monitoring data, performing attribute storage according to different attributes of the monitoring data, and performing secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters;

and the damage positioning and early warning subsystem is used for taking the time-frequency domain parameters as the input of the CNN convolutional neural network, identifying structural damage by using the CNN convolutional neural network and outputting a health monitoring result.

Further, the identifying the structural damage by using the CNN convolutional neural network includes:

constructing the CNN convolutional neural network, taking different positions of ancient architectural structure damage as classification labels of the CNN convolutional neural network, and predicting the structure damage position by adopting a Softmax function through an output layer of the CNN convolutional neural network; calculating deviation values of the probability distribution by using a Cross Entropy Cross-Entrophy loss function;

inputting the time-frequency domain parameters subjected to secondary processing as structural damage characteristic data into the CNN convolutional neural network for model training to obtain a CNN prediction model;

and inputting structural damage characteristic data to be identified and corresponding classification labels into the CNN prediction model, obtaining a 1 x n probability distribution matrix through dimension reduction processing, calculating a maximum expected value according to the probability distribution matrix, and determining the structural damage position.

Further, the calculating the deviation value of the probability distribution by using the Cross Entropy Cross-control loss function comprises:

wherein m is the number of samples; k is the number of categories of the classification label; p (x) _ij ) Is a sample x _i True probability of class j; q (theta, x) _ij ) For the model parameter theta, the sample x _i The probability of prediction as class j.

Further, the obtaining a 1 × n probability distribution matrix through dimension reduction processing, calculating a maximum expected value according to the probability distribution matrix, and determining the structure damage position includes:

the sample set D is: { (X) ⁽¹⁾ ，y ⁽¹⁾ ，(X ⁽²⁾ ，y ⁽²⁾ )，…，(X ^(m) ，y ^(m) ) H, input characteristics are

Damage status definition class label y ⁽ⁱ⁾ E {0,1, \8230;, k }, assuming the function is:

wherein the content of the first and second substances, ^P {y ⁽ⁱ⁾ ＝j|X ⁽ⁱ⁾ for each sample, estimating the probability of the label to which it belongs; k is the number of categories of the classification label;

for the values after normalizing the probability distribution, this makes the sum of the probabilities 1; because of

The parameter matrix θ is thus expressed as follows:

for each sample, the probability of the label to which it belongs is estimated as:

and obtaining the probability corresponding to each label, namely the probability distribution matrix.

Furthermore, the data acquisition and transmission subsystem is also used for diagnosing the working state of the sensor in real time, and generating an alarm signal when the state of the sensor is abnormal.

Further, still include:

the historic building health monitoring platform is based on a browser/server (B/S) framework and used for carrying out system management on the historic building health monitoring system, carrying out equipment management on the sensor, carrying out data management on monitoring data, carrying out early warning management on a damage positioning and early warning subsystem and carrying out report management on a health detection result.

In a second aspect of the invention, a method of monitoring the health of an ancient building is provided. The method comprises the following steps:

acquiring environmental monitoring data, component monitoring data and structure monitoring data of a target historic building;

preprocessing the acquired monitoring data to obtain preprocessed monitoring data;

performing attribute-based storage according to different attributes of the monitoring data, and performing secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters;

and taking the time-frequency domain parameters as the input of the CNN convolutional neural network, identifying the structural damage by using the CNN convolutional neural network, and outputting a health monitoring result.

Further, the structural damage identification by using the CNN convolutional neural network includes:

constructing the CNN convolutional neural network, taking different positions of the ancient building structure damage as classification labels of the CNN convolutional neural network, and predicting the structure damage position by adopting a Softmax function in an output layer of the CNN convolutional neural network; calculating a deviation value of the probability distribution by using a Cross Entropy Cross-Encopy loss function;

In a third aspect of the invention, an electronic device is provided. The electronic device at least one processor; and a memory communicatively coupled to the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of the second aspect of the invention.

It should be understood that the statements herein reciting aspects are not intended to limit the critical or essential features of any embodiment of the invention, nor are they intended to limit the scope of the invention. Other features of the present invention will become apparent from the following description.

Drawings

The above and other features, advantages and aspects of various embodiments of the present invention will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. In the drawings, like or similar reference characters denote like or similar elements, and wherein:

FIG. 1 shows a schematic view of an ancient building health monitoring system according to an embodiment of the invention;

FIG. 2 illustrates a flow diagram of a method of monitoring the health of an ancient building according to an embodiment of the invention;

FIG. 3 illustrates a block diagram of an exemplary electronic device capable of implementing embodiments of the present invention;

of these, 300 is an electronic device, 301 is a CPU, 302 is a ROM, 303 is a RAM, 304 is a bus, 305 is an I/O interface, 306 is an input unit, 307 is an output unit, 308 is a storage unit, and 309 is a communication unit.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.

In addition, the term "and/or" herein is only one kind of association relationship describing an associated object, and means that there may be three kinds of relationships, for example, a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. In addition, the character "/" herein generally indicates that the former and latter associated objects are in an "or" relationship.

According to the invention, through various collected monitoring information, a structural damage identification method of the CNN convolutional neural network is used for carrying out damage positioning on the structure, and an early warning function is realized according to a damage early warning standard, so that the health condition of the ancient building is comprehensively, comprehensively and accurately evaluated.

Fig. 1 shows a schematic view of an ancient building health monitoring system according to an embodiment of the invention.

The system comprises:

and the sensing subsystem 110 is used for acquiring environment monitoring data, component monitoring data and structure monitoring data of the target historic building through the sensors.

Specifically, the sensor comprises sensing elements or measuring instruments required by environment monitoring, component monitoring and structure monitoring, and is used for optimally arranging the sensor according to the structure dynamics principle and comprehensively sensing the local performance and the overall performance of the historic building structure, and the part is the front end and the foundation of the historic building health monitoring system.

The environment monitoring sensing elements comprise a temperature and humidity sensor, a carbon dioxide concentration sensor, a PM2.5 concentration sensor, an illumination intensity sensor, a wind speed and wind pressure sensor and the like. The environment monitoring data comprises data such as temperature and humidity, carbon dioxide concentration, PM2.5 concentration, illumination intensity, wind speed and wind pressure and the like.

The component monitoring and sensing elements comprise stress strain sensors, displacement sensors, deformation sensors and the like. The component monitoring data comprises strain, displacement, deformation and other data of the structure.

The structure monitoring sensing elements comprise an inclination sensor, a settlement sensor, a crack sensor, an MEMS acceleration sensor and the like. The structure monitoring data comprises data such as gradient, settlement degree, cracking degree and acceleration.

In this embodiment, the vent should be avoided when temperature and humidity sensor arranges, can arrange in the region that receives the humiture change to influence greatly, to the indoor outdoor humiture synchronous monitoring of ancient building to carry out contrastive analysis. The wind speed and wind pressure sensor can be firmly installed on the surface of an ancient building structure by adopting a fixed support, monitors the wind speed and the wind direction of the surrounding environment of the ancient building in real time, and has certain strength. The inclination sensor is mainly installed on main bearing force members such as beams, columns and the like which have obvious deformation in the historic building. The crack sensor is preferably arranged at the position of a crack with obvious cracks or cracks with development tendency and large strain in the ancient building components, so that the change process of the crack during monitoring can be continuously measured and recorded. The MEMS acceleration sensor is suitable to be arranged on the surface of the historic building member, is installed by adopting an M4 screw, has certain stability and rigidity, and is mainly used for monitoring the vibration law change of the historic building member.

As an embodiment of the present invention, after the sensing subsystem 110 collects the monitoring data, it waits for the data collection and transmission subsystem 120 to collect the monitoring data.

And the data acquisition and transmission subsystem 120 is configured to acquire the monitoring data acquired by the sensing subsystem, perform preprocessing, and transmit the preprocessed monitoring data to the data storage and management subsystem 130.

Specifically, the data acquisition and transmission subsystem is connected with the sensing subsystem, and the data acquisition subsystem realizes the work of receiving, recording, processing, sending and the like of various sensor signals in a real-time, timing, full-time and triggering mode.

Further, the data acquisition and transmission subsystem 120 has stability, durability and reliability, the data acquisition system is fully automated and can operate continuously under the unattended condition, and the acquired data can be shared and transmitted remotely. The acquisition mode is a distributed or mixed acquisition mode, and environment monitoring information such as temperature, humidity, carbon dioxide concentration, illumination intensity, wind speed and wind pressure and the like is acquired in a full time period; the structural vibration information is preferably acquired in a timing acquisition mode, and parameters such as sampling frequency, sampling time interval, threshold triggering and the like can be set. The strain, displacement, deformation and other component information of the structure can be acquired in a full time period.

Specifically, the data collecting and transmitting subsystem 120 converts analog signals such as temperature, humidity, light, stress strain, displacement and the like collected by various sensors into digital signals. Due to the noise interference and the abnormal condition of equipment, the original data collected by the collection system can generate abnormity and noise, and if a large amount of original monitoring data cannot be properly processed and analyzed, the state evaluation of the structure can be adversely affected, so that the monitoring data needs to be preprocessed and analyzed.

In the embodiment, in the preprocessing process, abnormal data in the original monitoring data are removed, missing data are filled up by an interpolation method, and data smoothing processing is considered, so that more real and higher-quality data can be obtained.

In this embodiment, the lagrangian interpolation method is used to fill in the missing data and the weighted average method is used to smooth the data. Through to data preprocessing back, can be more audio-visual observe the analysis to the monitoring data, know the deformation condition of ancient building component to improve the state assessment of structure and the accuracy of early warning.

In the embodiment, the data exchange and transmission of the monitoring system are completed by adopting a wireless communication technology (4G, NB-IOT/5G) in combination with the deployment conditions of various sensors, and the data acquired on site are transmitted to the data storage and processing subsystem for further analysis and processing.

As an embodiment of the invention, the data transmission subsystem is networked in a star structure, and the acquired data is transmitted to the data storage and processing subsystem in a wireless transmission (4G, NB-IOT/5G) mode.

Specifically, the data acquisition and transmission subsystem 120 converts analog signals of temperature, humidity, light, stress strain, displacement and the like acquired by various sensors into digital signals, and preprocesses acquired real-time monitoring data, wherein the data preprocessing includes signal filtering, signal amplification, analog-to-digital conversion, abnormal values, missing values, data smoothing and the like, and mainly through simple mathematical statistics, the calculation result can be used as an input primary early warning value. The data remote transmission adopts wireless communication technology (4G, NB-IOT/5G) to transmit the data collected on site to the data storage and processing subsystem for further analysis and processing.

As an embodiment of the present invention, the data collection and transmission subsystem 120 is further configured to perform real-time diagnosis on the working state of the sensor, and generate an alarm signal when the state of the sensor is abnormal.

Specifically, the data collection and transmission subsystem 120 should diagnose the health condition of the sensor in real time, and provide an alarm function when the sensor has failure, signal interruption, signal abnormality, or other fault information.

And the data storage and management subsystem 130 is configured to receive the preprocessed monitoring data, perform attribute-based storage according to different attributes of the monitoring data, and perform secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters.

The data storage and management subsystem 130 is the data hub and control center of the entire historic building health monitoring system. In a conventional structural health monitoring system, a data center of the system is often a physical server to provide services such as data storage, and the physical server is easily limited by environmental conditions, cannot realize data sharing, and even can cause data loss in severe cases. In the system, the cloud server is used as a data center, the cloud server is high in cost performance, high in response speed and flexible, stable and shared, and a cloud database is shared no matter at a web end or a mobile end, so that the whole monitoring system is connected.

As an embodiment of the present invention, the data storage and management subsystem 130 is composed of a cloud server, data management and data backup software, and the like. The cloud server classifies, stratifies, modularizes and stores the environmental monitoring data, the structural monitoring data and the structural monitoring data of the historic building. And the data management system carries out secondary processing on the monitoring data, namely secondary processing on the acquired acceleration data and stress-strain data to obtain time-frequency domain parameters. The system has the functions of data conversion, maintenance, online backup and recovery and the like, and can perform operations of increasing, deleting, modifying, checking and the like on data in the database.

The data storage and management subsystem 130 may perform several functions:

(1) Receiving and storing the monitoring data, classifying, layering, partitioning into modules, storing in a partition structure and carrying out secondary processing on the monitoring data. The secondary processing is to perform mathematical statistics, data compression or signal processing on the data, thereby identifying parameters of the time domain and the frequency domain.

(2) And a communication interface is provided, corresponding data stored in the database can be inquired at any time and any place, and a control instruction sent by a web end can be stored and fed back to the sensor node, so that close contact with the sensor node is kept.

(3) The method can quickly display, control, maintain, backup and restore the data on line, and can add, delete, modify, check and the like the existing data in the database. Different management authorities are given to different users, various types of data are stored in different modes, and a protection function is provided for the data.

(4) And a data resource library of the structure state assessment and early warning related information is established, and important data support is provided for the health state analysis and early warning of the ancient building components.

And the damage positioning and early warning subsystem 140 is used for taking the time-frequency domain parameters as input of the CNN convolutional neural network, identifying structural damage by using the CNN convolutional neural network, and outputting a health monitoring result.

Specifically, the identifying the structural damage by using the CNN convolutional neural network includes:

1) Constructing the CNN convolutional neural network, taking different positions of the ancient building structure damage as classification labels of the CNN convolutional neural network, and predicting the structure damage position by adopting a Softmax function in an output layer of the CNN convolutional neural network;

2) Calculating a deviation value of the probability distribution by using a Cross Entropy Cross-Encopy loss function; inputting the time-frequency domain parameters subjected to secondary processing as structural damage characteristic data into the CNN convolutional neural network for model training to obtain a CNN prediction model;

3) And inputting structural damage characteristic data to be identified and corresponding classification labels into the CNN prediction model, obtaining a 1 x n probability distribution matrix through dimension reduction processing, calculating a maximum expected value according to the probability distribution matrix, and determining the structural damage position.

In this embodiment, the calculating a deviation value of the probability distribution by using the Cross Entropy Cross-entry loss function includes:

wherein m is the number of samples; k is the number of categories of the classification label; p (x) _ij ) Is a sample x _i True probability of class j; q (theta, x) _ij ) For the model parameter theta, the sample x _i Probability of prediction as class j.

In this embodiment, the obtaining a 1 × n probability distribution matrix through dimension reduction processing, calculating a maximum expected value according to the probability distribution matrix, and determining the structural damage position includes:

the sample set D is: { ( ^X(1) ，y ⁽¹⁾ )，(X ⁽²⁾ ，y ⁽²⁾ )，…，(X ^(m) ，y ^(m) ) Is characterized by the input

wherein, P { y ⁽ⁱ⁾ ＝j|X ⁽ⁱ⁾ Estimating for each sample the probability of its belonging label; k is the number of categories of the classification label;

for the values after normalizing the probability distribution, this makes the sum of the probabilities 1; because of the fact that

The parameter matrix θ is thus expressed as follows:

for each sample, the probability of the label to which the sample belongs is estimated as:

As an embodiment of the present invention, the possible damage positions of the structure coexist in k, which is used as the classification label category number of the random forest model, and n feature parameters are used for input, so that the sample set D is:

{(X ⁽¹⁾ ，y ⁽¹⁾ ，(X ⁽²⁾ ，y ⁽²⁾ )，…，(X ^(m) ，y ^(m) )}

wherein the input features are

The label is y ⁽ⁱ⁾ ∈{0，1，…，k}。

The probability distribution of different classification labels belonging to each sample data is calculated by adopting a Softmax function, and the formula is simplified and expressed as follows:

wherein in the above formula, S _i Is the value V of each element in the array V _i Output value after the Softmax pooling of (a).

Is an index of the element(s),

for all element indices and the output values thus obtained for the multiple classes are converted to a range of [0,1 ]]And a probability distribution of 1.

The Cross-Entropy loss function enables the predicted probability distribution to approach the real probability distribution as much as possible by reducing the difference of the two probability distributions, completes the updating of the hyper-parameter in the back propagation process of the neural network, and obtains the best model performance when the calculated loss value is minimum, and the calculation formula is as follows:

wherein m is the number of samples; k is the number of categories classified; p (x) _ij ) Is a sample x _i True probability of class j; q (theta, x) _ij ) For the model parameter is theta, sample x _i Probability of prediction as class j.

Further, after the training of the damage identification model is finished, the structural damage characteristic parameters and the corresponding classification labels are input, and the specific calculation principle is as follows:

assume that sample set D is: { (X) ⁽¹⁾ ，y ⁽¹⁾ )，(X ⁽²⁾ ，y ⁽²⁾ )，…，(X ^(m) ，y ^(m) )}

Its input characteristics are

The label is y ⁽ⁱ⁾ E {0,1, \8230;, k }, the specific hypothetical function is:

wherein P { y [ ] ⁽ⁱ⁾ ＝j|X ⁽ⁱ⁾ Estimating for each sample the probability of its belonging label;

for the values after normalizing the probability distribution, this makes the sum of the probabilities 1; wherein

Therefore, θ is a parameter matrix of all parameters of the model, i.e. kx (n + 1), and the parameter matrix can be expressed as follows:

finally, it can be found that for each sample, the probability of the label to which it belongs is estimated as:

therefore, the classification weight of different classification label values of the input structural damage characteristic parameters can be calculated; and determining a classification label corresponding to the input data according to the classification weight so as to diagnose the damage position of the structure and realize an early warning function according to a damage early warning standard.

The structural damage identification method based on the CNN convolutional neural network can solve the problem of how to improve the structural damage diagnosis precision. The method comprises the steps of inputting time-frequency domain parameters after secondary processing as convolutional layers, using different damage positions of an ancient building structure as CNN classification labels, completing feature self-extraction work by using a deep learning network, integrating feature graphs extracted by an upper network and auxiliary data such as temperature and humidity values at different time points as full-connection layer input, and finally calculating the maximum expected value of damage at each position by using a Sorftmax function so as to obtain classification weight. And determining the classification label corresponding to the input data according to the classification weight, thereby determining the damage position of the historic building structure.

In some embodiments, the historic building health monitoring system further comprises a historic building health monitoring platform. The historic building health monitoring platform is based on a browser/server (B/S) framework and used for carrying out system management on the historic building health monitoring system, carrying out equipment management on the sensor, carrying out data management on monitoring data, carrying out early warning management on a damage positioning and early warning subsystem and carrying out report management on a health detection result. The historic building health monitoring platform adopts a development mode with a front end and a rear end separated. The B/S framework realizes the core service of the monitoring system at the cloud server end, and only few services are realized at the front end. And the method does not need to specially install a client, can be directly placed on a wide area network, has wide application range, can interact with a cloud server only by using a web browser, and has strong sharing performance.

In some optional implementations of this embodiment, the historic building health monitoring platform is divided into the following modules:

(1) The system management module is divided into user management and authority management, and the user management can realize normal registration and login of a user on the health monitoring platform, and realize addition, modification, deletion and the like of the user. The authority management is mainly to distribute the operation authority of the system, and can be divided into a system administrator and a common user, and different roles correspond to different system operation authorities.

(2) The equipment management module is divided into equipment position and equipment information, and relies on the WebGIS technology by embedding the WebGIS technology

And the geographic information system can display the specific positioning information of the target monitoring point on the electronic map. Various sensor devices are managed in a unified mode, sensor information is displayed in a list mode, a user can check the working state of the sensor in real time through the system, the devices are controlled remotely, and device parameters are adjusted. When the equipment is in abnormal conditions such as failure and damage, the monitoring system sends alarm information and informs a manager to process.

(3) The data display module is divided into real-time data, historical data and data statistics, the real-time monitoring data are displayed in a line graph mode, a bar graph mode and the like, the historical monitoring data are inquired in a graph mode, a data list mode and the like, characteristic value data such as a maximum value, a minimum value, an average value, a standard deviation and the like can be displayed, and the functions of time period selection and the like are achieved. And the data statistical analysis realizes secondary processing on the monitoring data, comprises basic statistical analysis, correlation, time-frequency domain analysis and the like, and is combined with the actual situation of the ancient building member to carry out safety evaluation on the member.

(4) The early warning management module is divided into threshold value configuration and early warning processing, various collected monitoring data are analyzed through a convolutional neural network, and the threshold value is configured according to the damage early warning standard. After the early warning mechanism is triggered, relevant managers can be informed in a multi-dimensional mode through voice calls, mobile phone app push, short message notification, mail notification and the like so as to be processed in time.

(5) The report management module comprises report generation and report downloading, the historic building health monitoring platform can carry out classified storage and management according to monitoring data in different time periods, and reports and visual charts are generated on line. The platform can provide different types of reports, such as daily statistics reports, monthly statistics reports, quarterly statistics reports and the like, and a user can download corresponding file reports according to requirements.

The invention applies the technologies of Internet of things, cloud computing, wireless communication, artificial intelligence and the like, changes the traditional intermittent and offline monitoring into full-automatic real-time online monitoring, and makes the whole monitoring system develop towards the direction of automation, networking, digitization and intelligence.

The structural damage identification method using the CNN convolutional neural network has high efficiency, high precision and high robustness, and can extract the damage characteristics of the structure and automatically process monitoring data, thereby improving the damage diagnosis precision of the structure.

The cloud server is used for classifying, layering, modularizing and storing the environmental monitoring data, the structure monitoring data and the component monitoring data of the historic building, and the cloud server has the functions of data conversion, maintenance, online backup and recovery and the like. The historic building health monitoring platform based on the B/S framework can realize functions of online data display, historical data query, GIS positioning, data report and the like, so that the health condition of the historic building is comprehensively, comprehensively and accurately evaluated.

The above is a description of system embodiments, and the following is a further description of the solution of the present invention by way of method embodiments.

As shown in fig. 2, the ancient building health monitoring method comprises the following steps:

s201, acquiring environmental monitoring data, component monitoring data and structure monitoring data of a target historic building;

s202, preprocessing the acquired monitoring data to obtain preprocessed monitoring data;

s203, performing attribute-based storage according to different attributes of the monitoring data, and performing secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters;

and S204, taking the time-frequency domain parameters as input of the CNN convolutional neural network, identifying structural damage by using the CNN convolutional neural network, and outputting a health monitoring result.

It can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working process of the described module may refer to the corresponding process in the foregoing method embodiment, and is not described herein again.

In the technical scheme of the invention, the acquisition, storage, application and the like of the personal information of the related user all accord with the regulations of related laws and regulations without violating the good customs of the public order.

According to an embodiment of the invention, the invention further provides an electronic device.

FIG. 3 shows a schematic block diagram of an electronic device 300 that may be used to implement an embodiment of the invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

The device 300 comprises a computing unit 301 which may perform various suitable actions and processes according to a computer program stored in a Read Only Memory (ROM) 302 or a computer program loaded from a storage unit 308 into a Random Access Memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the device 300 can also be stored. The calculation unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input/output (I/O) interface 305 is also connected to bus 304.

Various components in device 300 are connected to I/O interface 305, including: an input unit 306 such as a keyboard, a mouse, or the like; an output unit 307 such as various types of displays, speakers, and the like; a storage unit 308 such as a magnetic disk, optical disk, or the like; and a communication unit 309 such as a network card, modem, wireless communication transceiver, etc. The communication unit 309 allows the device 300 to exchange information/data with other devices via a computer network such as the internet and/or various telecommunication networks.

The computing unit 301 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various dedicated Artificial Intelligence (AI) computing chips, various computing units running machine learning model algorithms, a Digital Signal Processor (DSP), and any suitable processor, controller, microcontroller, and so forth. The calculation unit 301 executes the respective methods and processes described above, such as the methods S201 to S204. For example, in some embodiments, methods S201-S204 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 308. In some embodiments, part or all of the computer program may be loaded and/or installed onto device 300 via ROM 302 and/or communication unit 309. When the computer program is loaded into RAM 303 and executed by the computing unit 301, one or more steps of methods S201-S204 described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to perform the methods S201-S204 in any other suitable manner (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuitry, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), system on a chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs that are executable and/or interpretable on a programmable system including at least one programmable processor, which may be special or general purpose, receiving data and instructions from, and transmitting data and instructions to, a storage system, at least one input device, and at least one output device.

Program code for implementing the methods of the present invention may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions/operations specified in the flowchart and/or block diagram to be performed. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present invention, a machine-readable medium may be a tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which a user can provide input to the computer. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic, speech, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a back-end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), and the Internet.

The computer system may include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server may be a cloud server, a server of a distributed system, or a server combining a blockchain.

It should be understood that various forms of the flows shown above may be used, with steps reordered, added, or deleted. For example, the steps described in the present invention may be executed in parallel, sequentially, or in different orders, and are not limited herein as long as the desired results of the technical solution of the present invention can be achieved.

The above-described embodiments should not be construed as limiting the scope of the invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may be made in accordance with design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An ancient building health monitoring system, comprising:

the sensing subsystem is used for acquiring environmental monitoring data, component monitoring data and structure monitoring data of the target historic building through a sensor;

2. The historic building health monitoring system of claim 1, wherein the structural damage identification using the CNN convolutional neural network comprises:

3. The historic building health monitoring system of claim 2, wherein the calculating of the deviation value for the probability distribution using a Cross Entropy Cross-entry loss function comprises:

wherein m is the number of samples; k is the number of categories of the classification label; p (x) _ij ) Is a sample x _i True probability of class j; q (theta, x) _ij ) For the model parameter is theta, sample x _i Probability of prediction as class j.

4. The historic building health monitoring system of claim 2, wherein the determining the location of the structural damage by obtaining a 1 x n probability distribution matrix through dimension reduction processing and calculating the maximum expected value from the probability distribution matrix comprises:

the sample set D is:

the input features are

Damage status definition class tag y ⁽ⁱ⁾ E {0,1, \8230;, k }, assuming the function is:

is divided into probabilityThe normalized values are distributed, which makes the sum of the probabilities 1; because of the fact that

The parameter matrix θ is expressed as follows:

5. The historic building health monitoring system of claim 1, wherein the data acquisition and transmission subsystem is further configured to diagnose the working state of the sensor in real time and generate an alarm signal when the sensor is out of order.

6. The historic building health monitoring system of claim 1, further comprising:

7. A historic building health monitoring method is characterized by comprising the following steps:

performing attribute-based storage according to different attributes of monitoring data, and performing secondary processing on the preprocessed monitoring data to obtain time-frequency domain parameters;

8. The historic building health monitoring method of claim 7, wherein the identification of structural damage using the CNN convolutional neural network comprises:

constructing the CNN convolutional neural network, taking different positions of ancient architectural structure damage as classification labels of the CNN convolutional neural network, and predicting the structure damage position by adopting a Softmax function through an output layer of the CNN convolutional neural network; calculating a deviation value of the probability distribution by using a Cross Entropy Cross-Encopy loss function;

inputting the secondarily processed time-frequency domain parameters serving as structural damage characteristic data into the CNN convolutional neural network for model training to obtain a CNN prediction model;

9. An electronic device comprising at least one processor; and

a memory communicatively coupled to the at least one processor; it is characterized in that the preparation method is characterized in that,

the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of claims 7-8.