Intelligent detection and disposal method for large-displacement modular expansion and contraction diseases

Abstract

The invention relates to the technical field of disease detection and disposal methods, in particular to an intelligent detection and disposal method for large-displacement modular expansion and contraction diseases, and the method comprises the steps: obtaining static data and dynamic data of a large-displacement modular expansion and contraction device; wherein the static data comprises equipment specification, product model and manufacturer information, and the dynamic data comprises temperature, humidity, displacement, tensile stress and speed information; constructing a data analysis model by adopting an artificial intelligence algorithm based on the static data and the dynamic data; determining the health condition, predictive maintenance time and fault diagnosis information of the large-displacement modular telescopic device according to the output result of the data analysis model; generating an intelligent detection report containing health conditions, predictive maintenance time and fault diagnosis information; and based on an intelligent detection report, a disposal scheme of the large-displacement modular telescopic device is formulated, so that the reliability and the service life of the equipment are remarkably improved, and the maintenance cost is greatly reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of disease detection and treatment methods, and in particular to an intelligent detection and treatment method for large displacement modulus telescopic diseases. Background Art

[0002] As a key component of large infrastructure such as bridges and elevated roads, the health of large-displacement modular telescopic devices directly affects the safety and service life of the overall structure. With the rapid development of infrastructure construction, the demand for intelligent detection and maintenance of large-displacement modular telescopic devices is becoming increasingly urgent.

[0003] Traditional telescopic device inspection methods mainly rely on manual regular inspections, which have many limitations. First, the frequency of manual inspections is limited by human resources, making it difficult to achieve continuous monitoring and easily missing sudden changes in the equipment status. Second, the inspection results often rely on the experience and subjective judgment of the inspectors, which have certain errors and uncertainties. Furthermore, manual inspections are difficult to penetrate into the equipment, and the ability to discover and diagnose some hidden faults is limited.

[0004] In recent years, with the development of sensor technology and the Internet of Things, some automated detection systems have begun to be used in the monitoring of telescopic devices. These systems have achieved real-time monitoring of parameters such as temperature, humidity, and displacement by installing various sensors. However, these systems still have some shortcomings. For example, most systems simply collect and display data, lacking the ability to deeply analyze and intelligently interpret the data. At the same time, these systems usually operate in isolation, making it difficult to share data and conduct collaborative analysis with other systems, resulting in inefficient information utilization.

[0005] In addition, existing detection systems generally lack predictive maintenance capabilities. They can only reflect the current status of the equipment and cannot predict problems that may arise in the future, which makes maintenance work still remain at the passive response stage. For key components such as large displacement modular telescopic devices that bear high loads and high frequency of use, relying solely on passive maintenance is far from enough.

[0006] In terms of handling, existing technologies also have obvious shortcomings. Most systems can only issue simple alarm information and lack targeted handling suggestions and remote control capabilities. This results in a large amount of manual intervention after the problem is discovered, delaying the troubleshooting time and increasing maintenance costs.

[0007] Finally, existing systems generally lack the ability to self-learn and continuously optimize. As the equipment is used for a longer time, its performance characteristics may change, and fixed detection models are difficult to adapt to such dynamic changes, which can easily lead to a decrease in detection accuracy. Summary of the invention

[0008] In response to the above problems, the present invention proposes an intelligent detection and treatment method for large displacement modular telescopic diseases. The method aims to achieve all-round, intelligent and predictive management of large displacement modular telescopic devices, thereby improving the reliability of the equipment, extending its service life, reducing maintenance costs, and providing strong guarantees for the safe operation of infrastructure.

[0009] The present invention proposes an intelligent detection and treatment method for large displacement modulus expansion and contraction diseases, including:

[0010] The acquisition steps include:

[0011] Obtain static data and dynamic data of the large displacement modular telescopic device; wherein the static data includes equipment specifications, product model and manufacturer information, and the dynamic data includes temperature, humidity, displacement, tensile stress and speed information;

[0012] Processing steps include:

[0013] Based on the static data and dynamic data, an artificial intelligence algorithm is used to construct a data analysis model;

[0014] Determine the health status, predictive maintenance time and fault diagnosis information of the large displacement modular telescopic device according to the output results of the data analysis model;

[0015] Output steps include:

[0016] generating an intelligent inspection report including the health status, predictive maintenance time and fault diagnosis information;

[0017] Based on the intelligent detection report, a disposal plan for the large-displacement modular telescopic device is formulated.

[0018] Preferably, the obtaining step specifically includes:

[0019] Collect the dynamic data by means of a plurality of sensors installed on the large displacement modular telescopic device;

[0020] The static data is obtained from a device management system.

[0021] Preferably, the artificial intelligence algorithm used in the processing step includes a deep learning algorithm, and the deep learning algorithm is at least one of a convolutional neural network CNN, a recurrent neural network RNN ​​or a long short-term memory network LSTM.

[0022] Preferably, the processing step further comprises:

[0023] Preprocessing the static data and dynamic data, including data cleaning, outlier processing and data standardization;

[0024] Extract features from the preprocessed data to obtain feature data;

[0025] The data analysis model is trained based on the feature data.

[0026] Preferably, the processing step further comprises:

[0027] Based on historical data and currently collected dynamic data, predict the future state of the large displacement modular telescopic device;

[0028] Determine the optimal time for predictive maintenance based on the predicted future state.

[0029] Preferably, the disposal scheme in the output step includes:

[0030] When an abnormality is detected, an alarm message is generated and sent to the management personnel;

[0031] Provide specific maintenance suggestions based on fault diagnosis information;

[0032] Develop equipment maintenance plans based on predictive maintenance time.

[0033] As a preference, the remote control step is also included:

[0034] generating a control instruction based on the intelligent detection report;

[0035] The control instructions are sent to the control system of the large displacement modular telescopic device to achieve remote adjustment and control.

[0036] Preferably, the method further comprises the step of data storage:

[0037] Storing the static data, dynamic data, intelligent detection report and treatment plan in a database;

[0038] Back up and update stored data regularly.

[0039] Preferably, the processing step further comprises:

[0040] Dynamically adjust the working parameters of the large displacement modular telescopic device based on traffic flow data and environmental factors;

[0041] Optimize the equipment's operating strategy based on the adjusted working parameters.

[0042] As a preferred embodiment, the model updating step is also included:

[0043] Collect actual operation data and maintenance records of large displacement modular telescopic devices;

[0044] Based on the actual operation data and maintenance records, the data analysis model is regularly updated to improve the accuracy and adaptability of the model.

[0045] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0046] The method of the present invention has achieved technical breakthroughs in many aspects through the organic combination of innovative technologies such as multi-dimensional data collection, deep learning analysis, intelligent decision support, adaptive control and continuous optimization. First, the method establishes a comprehensive data collection system, which not only includes traditional physical parameters, but also covers the static information and environmental factors of the equipment, laying a solid data foundation for subsequent intelligent analysis. Secondly, by introducing deep learning algorithms, especially LSTM networks for time series data, the method can mine deep features and laws from massive data, greatly improving the accuracy of fault diagnosis and prediction.

[0047] More importantly, the method of the present invention achieves a shift from passive response to active prevention. By establishing a predictive maintenance model, the method can predict possible problems in advance and give optimal maintenance time recommendations, thereby avoiding emergency shutdowns and high maintenance costs caused by sudden failures. At the same time, the method also has intelligent decision-making support capabilities, and can automatically generate targeted disposal plans based on equipment status and prediction results, greatly reducing the need for manual decision-making.

[0048] In terms of control, the method of the present invention introduces an adaptive control mechanism, which can adjust the working parameters of the equipment in real time according to the traffic flow, environmental factors, etc., so that the telescopic device always maintains the best working state. This not only improves the adaptability of the equipment, but also extends its service life.

[0049] In addition, the method of the present invention also has the ability of continuous learning and self-optimization. Through online learning technology, the detection model can continuously absorb new data and experience, automatically update and optimize, so as to maintain long-term high accuracy and strong adaptability. This solves the problem that traditional fixed models are difficult to adapt to performance changes during long-term use of equipment.

[0050] In general, the intelligent detection and disposal method provided by the present invention realizes the all-round intelligent management of large-displacement modular telescopic devices through the organic integration of multiple innovative technologies. This method not only significantly improves the reliability and service life of the equipment, but also greatly reduces the maintenance cost, providing strong technical support for the safe operation of large infrastructure such as bridges. It represents an important development direction of telescopic device management technology and is expected to generate huge economic and social benefits in practical applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1It is a detailed logic block diagram of the acquisition steps of the present invention.

[0052] Figure 2 A detailed logic block diagram of the processing steps of the present invention.

[0053] Figure 3 It is a logic block diagram of the output step of the present invention.

[0054] Figure 4 It is a logic block diagram of the remote control step of the present invention.

[0055] Figure 5 It is a logic block diagram of the data storage step of the present invention. DETAILED DESCRIPTION

[0056] Please refer to Figure 1-5 The present invention discloses an intelligent detection and treatment method for large displacement modular telescopic diseases. The method collects and analyzes multi-dimensional data of large displacement modular telescopic devices to achieve intelligent diagnosis and predictive maintenance of the device status, thereby improving the reliability and management efficiency of the equipment.

[0057] The method of the present invention comprises an acquisition step, a processing step and an output step.

[0058] In the acquisition step, the method acquires static data and dynamic data of the large displacement modular telescopic device. The static data includes equipment specifications, product models and manufacturer information, which are essential for understanding the basic characteristics of the equipment. The dynamic data includes temperature, humidity, displacement, tensile stress and speed information, which are real-time data that can reflect the current operating status of the equipment.

[0059] For example, static data may include parameters such as the maximum displacement of the telescopic device (such as ±500mm) and the designed service life (such as 30 years). Dynamic data may include the current ambient temperature (such as -20℃ to 60℃), relative humidity (such as 0-100%), real-time displacement (such as ±450mm), tensile stress (such as 0-500MPa) and displacement speed (such as 0-10mm / s). Changes in ambient temperature will affect the physical properties of the material, such as elastic modulus and thermal expansion coefficient, and thus affect the performance of the telescopic device. The temperature range is usually -20℃ to 60℃.

[0060] Relative humidity (0-100%) affects the corrosion rate of metal parts and the aging rate of seals, which in turn affects the life and safety of the equipment. Real-time displacement (such as ±450mm) is a key indicator to measure the working status of the telescopic device. Excessive displacement may cause equipment damage or failure. Tensile stress (such as 0-500MPa) reflects the load conditions that the telescopic device bears during operation. Excessive stress may cause material fatigue or fracture. Displacement speed (such as 0-10mm / s) indicates the movement speed of the telescopic device. Excessive speed may cause impact and vibration, affecting the stability and safety of the equipment.

[0061] In the processing step, the method uses an artificial intelligence algorithm to build a data analysis model based on the acquired static data and dynamic data. The artificial intelligence algorithm here can be a deep learning algorithm, such as a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM).

[0062] LSTM (Long Short-Term Memory) is a special recurrent neural network that is particularly suitable for processing and predicting time series data. It solves the gradient vanishing problem in traditional RNN by introducing "memory units" and can effectively capture long-term dependencies. The core idea of ​​LSTM is to control the flow of information through three gating mechanisms (forget gate, input gate, and output gate) to achieve effective modeling of time series data. Through memory units and gating mechanisms, LSTM can effectively capture long-term dependencies in time series, which is very important for predicting the future state of the telescopic device.

[0063] LSTM can handle complex nonlinear relationships and is suitable for multivariate time series data, such as temperature, humidity, displacement, stress, and speed. LSTM is highly robust to noise and can maintain high prediction accuracy in the presence of noise.

[0064] The LSTM model can effectively capture the characteristics of the state of the large displacement modulus telescopic device changing over time, so as to more accurately predict the future state of the equipment. Specifically, the model can predict key parameters such as future displacement, stress, and temperature based on historical data and real-time data, helping maintenance personnel to detect potential problems in advance and take preventive measures.

[0065] Through this model, the time-varying characteristics of the state of the large displacement modular telescopic device can be effectively captured, thereby more accurately predicting the future state of the equipment.

[0066] Based on the output of the data analysis model, the method determines the health status, predictive maintenance time and fault diagnosis information of the large displacement modular telescopic device. For example, the health status can be represented by a score of 0-100, where 0 means that the device has failed and 100 means that the device is in the best condition. The predictive maintenance time can be accurate to the day, such as "maintenance is recommended within 30 days". Fault diagnosis information may include specific fault types and severity, such as "minor displacement abnormality, observation is recommended" or "serious stress exceeds the standard, immediate maintenance is required".

[0067] In the output step, the method generates an intelligent detection report including the above health status, predictive maintenance time and fault diagnosis information. This report can be a structured data file or a visual chart report. Based on this intelligent detection report, the method further formulates a disposal plan for the large displacement modular telescopic device.

[0068] The acquisition step of the present invention specifically includes collecting dynamic data through multiple sensors installed on the large-displacement modular telescopic device, and acquiring static data from the equipment management system.

[0069] In a preferred embodiment, these sensors may include temperature sensors, humidity sensors, displacement sensors, stress sensors, and speed sensors. The temperature sensor may be a thermocouple or a thermistor with an accuracy of ±0.1°C. The humidity sensor may be a capacitive sensor with an accuracy of ±2%RH. The displacement sensor may be a linear variable differential transformer (LVDT) with a measurement range of ±500mm and an accuracy of 0.1%FS. The stress sensor may be a resistance strain gauge with a measurement range of 0-500MPa and an accuracy of 0.1%FS. The speed sensor may be a Hall effect sensor with a measurement range of 0-10mm / s and an accuracy of 0.5%FS.

[0070] These sensors transmit data to the data acquisition unit in real time through the wireless sensor network. At the same time, static data stored in the equipment management system, such as equipment specifications, product models and manufacturer information, will also be automatically acquired. This fusion of multi-source data lays a solid foundation for subsequent intelligent analysis.

[0071] The artificial intelligence algorithm used in the processing steps of the present invention includes a deep learning algorithm, which can specifically be at least one of a convolutional neural network (CNN), a recurrent neural network (RNN), or a long short-term memory network (LSTM).

[0072] In one embodiment of the present invention, an LSTM network can be used to process time series data. A key advantage of the LSTM network is that it can effectively capture long-term dependencies, which is particularly important for analyzing the long-term performance changes of large displacement modulus telescopic devices.

[0073] For example, a network structure can be constructed that includes an input layer, an LSTM layer, and an output layer. The input layer receives standardized multidimensional sensor data, the LSTM layer contains 128 neurons, and the output layer outputs the health score, predictive maintenance time, and fault type probability.

[0074] The network training process can use the Adam optimizer, with the learning rate set to 0.001, the batch size to 64, and the number of training rounds to 100. In order to prevent overfitting, a Dropout layer can be added after the LSTM layer, with the dropout rate set to 0.5.

[0075] In this way, a powerful predictive model can be built that can not only accurately assess the current state of the equipment, but also predict possible problems in the future, thus achieving true predictive maintenance.

[0076] In general, the intelligent detection and disposal method provided by the present invention realizes the all-round intelligent management of the large displacement modular telescopic device through multi-dimensional data collection, deep learning analysis and intelligent decision support. This not only improves the reliability and service life of the equipment, but also greatly reduces the maintenance cost, providing a strong guarantee for the safe operation of large infrastructure such as bridges. In a preferred embodiment of the present invention, the processing step also includes preprocessing static data and dynamic data, including data cleaning, outlier processing and data standardization; extracting features from the preprocessed data to obtain feature data; and training a data analysis model based on the feature data.

[0077] In the data preprocessing stage, data cleaning is first performed to remove possible noise and redundant information. For example, for temperature data, the moving average filter method can be used to eliminate short-term fluctuations, and the window size can be set to 5-10 sampling points. The triple standard deviation method is used for outlier processing, that is, data that exceeds the range of ±3 times the standard deviation of the mean is regarded as an outlier and is removed or replaced. Data standardization uses the Z-score standardization method. In this way, data of different dimensions can be unified to the same scale, which is conducive to subsequent feature extraction and model training.

[0078] Feature extraction is a key step in data processing. This method uses a feature extraction strategy that combines time domain and frequency domain. In the time domain, statistical features such as mean, standard deviation, peak, skewness and kurtosis are extracted; in the frequency domain, spectrum features are extracted by fast Fourier transform (FFT). Taking displacement data as an example, its frequency domain features can be calculated by the following formula:

[0079]

[0080] Among them, x(n) is the time domain signal, X(k) is the frequency domain signal, and N is the number of sampling points. By analyzing the amplitude and phase information of X(k), the periodic variation characteristics of the large displacement modulus telescopic device can be captured, which is crucial for predicting the long-term performance of the device.

[0081] Based on the extracted feature data, this method uses cross-validation to train the data analysis model. Specifically, the data set is divided into training set, validation set and test set in a ratio of 8:1:1. During the training process, early stopping is used to prevent overfitting, that is, when the loss function on the validation set does not decrease for 5 consecutive epochs, the training is stopped. This method can effectively control the training time and computing resource consumption while ensuring the performance of the model.

[0082] The processing steps of the present invention also include predicting the future state of the large displacement modular telescopic device based on historical data and currently collected dynamic data; and determining the optimal predictive maintenance time according to the predicted future state.

[0083] In this step, the method uses time series prediction technology to estimate the future state of the equipment. Specifically, a long short-term memory network (LSTM) is used to build a prediction model, through which the method can predict the equipment status in the next 7 days, 30 days, and 90 days. Based on the prediction results, combined with the health threshold of the equipment (for example, a health score below 70 points is considered to require maintenance), the optimal predictive maintenance time is determined. This method can maximize the use time of the equipment while ensuring the safety of the equipment, thereby optimizing the maintenance cost.

[0084] The disposal scheme in the output step of the present invention includes: when an abnormality is detected, an alarm message is generated and sent to the management personnel; specific maintenance suggestions are provided according to the fault diagnosis information; and an equipment maintenance plan is formulated based on the predictive maintenance time.

[0085] In actual application, this method sets up a multi-level alarm mechanism. For example, when the equipment health score is between 60 and 70 points, the system will issue a yellow warning, suggesting that managers pay attention to the equipment status; when the score is below 60 points, the system will issue a red alarm, requiring immediate inspection of the equipment. The alarm information is delivered to the managers in a timely manner through SMS, email or mobile application push.

[0086] Fault diagnosis information is generated based on the output of the deep learning model and combined with the expert knowledge base. For example, if the model detects abnormal displacement, the system will recommend checking whether the sliding parts of the telescopic device are stuck; if abnormal stress is detected, it may recommend checking whether the fixing bolts are loose. These specific maintenance suggestions can greatly improve maintenance efficiency and reduce misjudgments and missed judgments.

[0087] The predictive maintenance plan is generated based on the predicted status of the equipment and the optimization of maintenance resources. This method uses a genetic algorithm to solve this multi-objective optimization problem. The objective function includes minimizing maintenance costs, maximizing equipment availability, and balancing maintenance resource load. In this way, an optimal maintenance plan that balances multiple objectives can be obtained.

[0088] For example, the present invention also includes a remote control step: generating control instructions based on the intelligent detection report; sending the control instructions to the control system of the large-displacement modular telescopic device to achieve remote adjustment and control.

[0089] In this step, the method first generates a control strategy based on the intelligent detection report. For example, if the displacement of the telescopic device is detected to be close to the limit, the system will generate an instruction to adjust the limit device; if the stress distribution is detected to be uneven, an instruction to adjust the prestress may be generated. These control instructions are sent to the on-site control system via a secure network protocol (such as HTTPS).

[0090] After receiving the command, the control system will first perform a safety verification to ensure the legality and executability of the command. After the verification is passed, the control system will perform corresponding operations, such as adjusting the motor output, changing the hydraulic system pressure, etc., so as to achieve precise control of the large displacement modulus telescopic device.

[0091] The entire remote control process has a real-time feedback mechanism, and the control effect will be immediately reflected in the intelligent detection report, forming a closed-loop control system. This remote control capability greatly improves the efficiency and flexibility of equipment management, especially for installation scenarios in remote locations or harsh environments.

[0092] Through the above detailed description, it can be seen that the intelligent detection and disposal method for large-displacement modular telescopic diseases provided by the present invention is comprehensive, intelligent and practical. This method can not only accurately diagnose the current state, but also predict future trends and provide targeted disposal solutions, thereby realizing the intelligent management of the entire life cycle of large-displacement modular telescopic devices. This has important practical significance for improving the safety and reliability of infrastructure and reducing maintenance costs. In another embodiment of the present invention, the method also includes a data storage step: storing static data, dynamic data, intelligent detection reports and disposal solutions in a database; and regularly backing up and updating the stored data.

[0093] In the data storage step, this method uses a distributed database system to store massive amounts of structured and unstructured data. Static data and dynamic data are stored in different data tables and associated through unique device IDs. Intelligent detection reports and disposal plans are stored in JSON format to facilitate subsequent retrieval and analysis.

[0094] Preferably, this method uses a time series database (such as InfluxDB) to store dynamic data. The data model of the time series database can be expressed as:

[0095] measurement <tag key =tag value >[,tag key =tag value ]field key

[0096] =field value [,field key =field value ][timestamp]

[0097] Among them, measurement represents the measurement index (such as temperature, humidity, etc.), tag is used for indexing, field stores the actual value, and timestamp is the timestamp. This data model can efficiently process time series data and support fast data writing and query.

[0098] To ensure the security and reliability of data, this method implements a regular backup strategy. Specifically, incremental backups are performed daily and full backups are performed weekly. The backup data is encrypted and stored on an off-site server using the AES-256 encryption algorithm, and key management uses a distributed key management system. This multiple protection mechanism can effectively prevent data loss and unauthorized access.

[0099] The data update strategy adopts a combination of regular updates and triggered updates. Static data is updated once a month, and dynamic data is updated in real time. When a major change in the equipment is detected (such as a major overhaul or replacement of a key component), a comprehensive data update will be triggered to ensure that the information in the database is always up-to-date and accurate.

[0100] The processing steps of the present invention also include: dynamically adjusting the working parameters of the large-displacement modular telescopic device based on vehicle flow data and environmental factors; and optimizing the operation strategy of the equipment according to the adjusted working parameters.

[0101] In this step, the method introduces the consideration of external environmental factors to achieve more intelligent and adaptable equipment management. Traffic flow data can be obtained through vehicle detectors installed on bridges or roads, and environmental factors include meteorological data such as temperature, humidity, and wind speed, which can be obtained from nearby weather stations or meteorological sensors installed on site.

[0102] This method uses multi-factor regression analysis to establish a relationship model between vehicle flow, environmental factors and telescopic device working parameters. The model can be expressed as:

[0103] Y=β 0 +β 1 X 1 +β 2 X 2 +...+β n X n +∈

[0104] Among them, Y is the working parameter (such as prestressing size, position of limit device, etc.), X 1 ,X 2 ,...,X n are influencing factors (such as traffic volume, temperature, humidity, etc.), β 0 ,β 1 ,...,β n is the regression coefficient, ∈ is the random error term. The regression coefficient is estimated by the least squares method, and we get:

[0105]

[0106] Among them, X is the independent variable matrix and Y is the dependent variable vector.

[0107] Based on this model, the method can dynamically calculate the optimal working parameters according to real-time traffic flow and environmental data. For example, in the case of heavy traffic and high temperature, the system may increase the prestress to improve the stiffness and stability of the telescopic device.

[0108] The optimized operation strategy is implemented by a fuzzy controller. The input of the fuzzy controller is a fuzzy set of vehicle flow and environmental factors, and the output is the adjustment of the working parameters. The general form of the fuzzy rule is:

[0109] IF (traffic volume is large) AND (temperature is high) THEN (prestress is increased)

[0110] In this way, the method realizes the intelligent adaptive control of the large displacement modular telescopic device, which significantly improves the adaptability and reliability of the equipment.

[0111] The present invention also includes a model updating step: collecting actual operation data and maintenance records of the large-displacement modular telescopic device; based on the actual operation data and maintenance records, regularly updating the data analysis model to improve the accuracy and adaptability of the model.

[0112] In this step, this method uses online learning technology to enable the model to continuously update and optimize itself as new data is accumulated. Specifically, an incremental learning algorithm, such as Stochastic Gradient Descent (SGD), is used to update the model parameters.

[0113] For the neural network model, the parameter update formula can be expressed as:

[0114]

[0115] Among them, θ t is the model parameter at time t, η is the learning rate, L(θ t ,x t ,y t ) is the loss function, x t and t are the input features and true labels, respectively.

[0116] Preferably, this method uses an adaptive learning rate algorithm, such as Adam (Adaptive Moment Estimation), to dynamically adjust the learning rate. The parameter update formula of the Adam algorithm is:

[0117] m t =β 1 m t-1 +(1-β 1 ) t

[0118]

[0119] Among them, m t and v t They are the first-order moment estimate and the second-order moment estimate of the gradient, β 1 and β 2 is the decay rate, and ∈ is a small constant used to prevent the denominator from being zero.

[0120] In order to prevent catastrophic forgetting during model updating, this method also introduces the Elastic Weight Consolidation (EWC) technique. EWC protects important parameters by adding a regularization term to the loss function:

[0121]

[0122] Among them, L B (θ) is the loss function of the new task, F i is the Fisher information matrix, is the optimal parameter for the old task, and λ is a trade-off coefficient.

[0123] Through this continuous learning and updating mechanism, this method can ensure that the data analysis model always maintains high accuracy and strong adaptability, and can accurately reflect the latest status and change trend of the large displacement modulus telescopic device. This not only improves the accuracy of intelligent detection, but also can adapt to new failure modes and performance changes that may occur in the long-term use of the equipment.

[0124] In general, the intelligent detection and treatment method for large-displacement modular telescopic diseases provided by the present invention realizes the all-round intelligent management of large-displacement modular telescopic devices through innovative technologies such as multi-dimensional data collection, deep learning analysis, intelligent decision support, adaptive control and continuous optimization. This method not only greatly improves the reliability and service life of the equipment, but also significantly reduces the maintenance cost, providing strong technical support for the safe operation of large infrastructure such as bridges.

[0125] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent detection and treatment method for large displacement modulus expansion and contraction diseases, characterized in that: include: The acquisition steps include: Obtain static data and dynamic data of the large displacement modular telescopic device; wherein the static data includes equipment specifications, product model and manufacturer information, and the dynamic data includes temperature, humidity, displacement, tensile stress and speed information; Processing steps include: Based on the static data and dynamic data, an artificial intelligence algorithm is used to construct a data analysis model; Determine the health status, predictive maintenance time and fault diagnosis information of the large displacement modular telescopic device according to the output results of the data analysis model; Output steps include: generating an intelligent inspection report including the health status, predictive maintenance time and fault diagnosis information; Based on the intelligent detection report, a disposal plan for the large-displacement modular telescopic device is formulated.

2. The method according to claim 1, characterized in that The acquisition step specifically includes: Collect the dynamic data by means of a plurality of sensors installed on the large displacement modular telescopic device; The static data is obtained from a device management system.

3. The method according to claim 1, characterized in that The artificial intelligence algorithm used in the processing step includes a deep learning algorithm, and the deep learning algorithm is at least one of a convolutional neural network CNN, a recurrent neural network RNN, or a long short-term memory network LSTM.

4. The method according to claim 1, characterized in that: The processing steps also include: Preprocessing the static data and dynamic data, including data cleaning, outlier processing and data standardization; Extract features from the preprocessed data to obtain feature data; The data analysis model is trained based on the feature data.

5. The method according to claim 1, characterized in that: The processing steps also include: Based on historical data and currently collected dynamic data, predict the future state of the large displacement modular telescopic device; Determine the optimal time for predictive maintenance based on the predicted future state.

6. The method according to claim 1, characterized in that The disposal scheme in the output step includes: When an abnormality is detected, an alarm message is generated and sent to the management personnel; Provide specific maintenance suggestions based on fault diagnosis information; Develop equipment maintenance plans based on predictive maintenance time.

7. The method according to claim 1, characterized in that Also includes remote control steps: generating a control instruction based on the intelligent detection report; The control instructions are sent to the control system of the large displacement modular telescopic device to achieve remote adjustment and control.

8. The method according to claim 1, characterized in that It also includes the data storage step: Storing the static data, dynamic data, intelligent detection report and treatment plan in a database; Back up and update stored data regularly.

9. The method according to claim 1, characterized in that: The processing steps also include: Dynamically adjust the working parameters of the large displacement modular telescopic device based on traffic flow data and environmental factors; Optimize the equipment's operating strategy based on the adjusted working parameters.

10. The method according to claim 1, characterized in that It also includes a model update step: Collect actual operation data and maintenance records of large displacement modular telescopic devices; Based on the actual operation data and maintenance records, the data analysis model is regularly updated to improve the accuracy and adaptability of the model.

Images